{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9773719730051607,
  "eval_steps": 3000,
  "global_step": 30000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.321103572845459,
      "epoch": 0.0004962286621675268,
      "grad_norm": 1.03125,
      "learning_rate": 2e-06,
      "loss": 6.9502,
      "mean_token_accuracy": 0.08894372060894966,
      "num_tokens": 11135.0,
      "step": 5
    },
    {
      "entropy": 7.291898345947265,
      "epoch": 0.0009924573243350536,
      "grad_norm": 0.9453125,
      "learning_rate": 4.5e-06,
      "loss": 6.9889,
      "mean_token_accuracy": 0.08681938499212265,
      "num_tokens": 22299.0,
      "step": 10
    },
    {
      "entropy": 7.309418964385986,
      "epoch": 0.0014886859865025804,
      "grad_norm": 1.1953125,
      "learning_rate": 7e-06,
      "loss": 6.9968,
      "mean_token_accuracy": 0.08861658647656441,
      "num_tokens": 33670.0,
      "step": 15
    },
    {
      "entropy": 7.30535888671875,
      "epoch": 0.001984914648670107,
      "grad_norm": 1.109375,
      "learning_rate": 9.5e-06,
      "loss": 6.9494,
      "mean_token_accuracy": 0.08504866808652878,
      "num_tokens": 44240.0,
      "step": 20
    },
    {
      "entropy": 7.317498207092285,
      "epoch": 0.002481143310837634,
      "grad_norm": 1.0234375,
      "learning_rate": 1.2e-05,
      "loss": 6.8974,
      "mean_token_accuracy": 0.09154646843671799,
      "num_tokens": 54168.0,
      "step": 25
    },
    {
      "entropy": 7.3441589832305905,
      "epoch": 0.0029773719730051607,
      "grad_norm": 0.84375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.9,
      "mean_token_accuracy": 0.09077088311314582,
      "num_tokens": 64704.0,
      "step": 30
    },
    {
      "entropy": 7.337980651855469,
      "epoch": 0.0034736006351726877,
      "grad_norm": 0.8515625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.9297,
      "mean_token_accuracy": 0.08911860883235931,
      "num_tokens": 75766.0,
      "step": 35
    },
    {
      "entropy": 7.370907640457153,
      "epoch": 0.003969829297340214,
      "grad_norm": 0.87109375,
      "learning_rate": 1.95e-05,
      "loss": 6.9438,
      "mean_token_accuracy": 0.08970932587981224,
      "num_tokens": 88074.0,
      "step": 40
    },
    {
      "entropy": 7.34942741394043,
      "epoch": 0.004466057959507741,
      "grad_norm": 0.88671875,
      "learning_rate": 2.2e-05,
      "loss": 6.85,
      "mean_token_accuracy": 0.09375255852937699,
      "num_tokens": 99313.0,
      "step": 45
    },
    {
      "entropy": 7.417236328125,
      "epoch": 0.004962286621675268,
      "grad_norm": 0.87890625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 7.0061,
      "mean_token_accuracy": 0.09381884783506393,
      "num_tokens": 110819.0,
      "step": 50
    },
    {
      "entropy": 7.343614864349365,
      "epoch": 0.0054585152838427945,
      "grad_norm": 0.93359375,
      "learning_rate": 2.7e-05,
      "loss": 6.8035,
      "mean_token_accuracy": 0.09773246571421623,
      "num_tokens": 121156.0,
      "step": 55
    },
    {
      "entropy": 7.392715406417847,
      "epoch": 0.0059547439460103215,
      "grad_norm": 0.8671875,
      "learning_rate": 2.95e-05,
      "loss": 6.9373,
      "mean_token_accuracy": 0.0945754513144493,
      "num_tokens": 131307.0,
      "step": 60
    },
    {
      "entropy": 7.320441150665284,
      "epoch": 0.0064509726081778485,
      "grad_norm": 0.96484375,
      "learning_rate": 3.2e-05,
      "loss": 6.8968,
      "mean_token_accuracy": 0.09729647785425186,
      "num_tokens": 141812.0,
      "step": 65
    },
    {
      "entropy": 7.360306215286255,
      "epoch": 0.0069472012703453755,
      "grad_norm": 0.9609375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.8448,
      "mean_token_accuracy": 0.08999949172139168,
      "num_tokens": 152425.0,
      "step": 70
    },
    {
      "entropy": 7.37144455909729,
      "epoch": 0.007443429932512902,
      "grad_norm": 0.8984375,
      "learning_rate": 3.7e-05,
      "loss": 6.8837,
      "mean_token_accuracy": 0.09427939280867577,
      "num_tokens": 162562.0,
      "step": 75
    },
    {
      "entropy": 7.373857641220093,
      "epoch": 0.007939658594680429,
      "grad_norm": 0.9140625,
      "learning_rate": 3.95e-05,
      "loss": 6.9141,
      "mean_token_accuracy": 0.09801442995667457,
      "num_tokens": 173322.0,
      "step": 80
    },
    {
      "entropy": 7.314117288589477,
      "epoch": 0.008435887256847956,
      "grad_norm": 0.89453125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.857,
      "mean_token_accuracy": 0.09269940182566642,
      "num_tokens": 183169.0,
      "step": 85
    },
    {
      "entropy": 7.3311868667602536,
      "epoch": 0.008932115919015483,
      "grad_norm": 0.92578125,
      "learning_rate": 4.45e-05,
      "loss": 7.0099,
      "mean_token_accuracy": 0.0901446096599102,
      "num_tokens": 194473.0,
      "step": 90
    },
    {
      "entropy": 7.3603380680084225,
      "epoch": 0.00942834458118301,
      "grad_norm": 0.9609375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.8324,
      "mean_token_accuracy": 0.09027603566646576,
      "num_tokens": 206166.0,
      "step": 95
    },
    {
      "entropy": 7.34685206413269,
      "epoch": 0.009924573243350537,
      "grad_norm": 1.0390625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.8455,
      "mean_token_accuracy": 0.09469634592533112,
      "num_tokens": 215923.0,
      "step": 100
    },
    {
      "entropy": 7.320917844772339,
      "epoch": 0.010420801905518062,
      "grad_norm": 0.83984375,
      "learning_rate": 5.2e-05,
      "loss": 6.9275,
      "mean_token_accuracy": 0.09063069075345993,
      "num_tokens": 227139.0,
      "step": 105
    },
    {
      "entropy": 7.346196937561035,
      "epoch": 0.010917030567685589,
      "grad_norm": 0.92578125,
      "learning_rate": 5.45e-05,
      "loss": 6.8558,
      "mean_token_accuracy": 0.09765214547514915,
      "num_tokens": 237770.0,
      "step": 110
    },
    {
      "entropy": 7.377953720092774,
      "epoch": 0.011413259229853116,
      "grad_norm": 0.84765625,
      "learning_rate": 5.7e-05,
      "loss": 6.936,
      "mean_token_accuracy": 0.09508275687694549,
      "num_tokens": 249241.0,
      "step": 115
    },
    {
      "entropy": 7.348550510406494,
      "epoch": 0.011909487892020643,
      "grad_norm": 0.90234375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.8565,
      "mean_token_accuracy": 0.09396146684885025,
      "num_tokens": 258562.0,
      "step": 120
    },
    {
      "entropy": 7.275180292129517,
      "epoch": 0.01240571655418817,
      "grad_norm": 1.0,
      "learning_rate": 6.2e-05,
      "loss": 6.902,
      "mean_token_accuracy": 0.09861418008804321,
      "num_tokens": 269548.0,
      "step": 125
    },
    {
      "entropy": 7.438327026367188,
      "epoch": 0.012901945216355697,
      "grad_norm": 0.96875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.9914,
      "mean_token_accuracy": 0.09398091435432435,
      "num_tokens": 280965.0,
      "step": 130
    },
    {
      "entropy": 7.4196254253387455,
      "epoch": 0.013398173878523224,
      "grad_norm": 1.1015625,
      "learning_rate": 6.7e-05,
      "loss": 6.8143,
      "mean_token_accuracy": 0.09857342019677162,
      "num_tokens": 290063.0,
      "step": 135
    },
    {
      "entropy": 7.25293459892273,
      "epoch": 0.013894402540690751,
      "grad_norm": 0.984375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.9046,
      "mean_token_accuracy": 0.09994807839393616,
      "num_tokens": 301013.0,
      "step": 140
    },
    {
      "entropy": 7.350448560714722,
      "epoch": 0.014390631202858276,
      "grad_norm": 0.90234375,
      "learning_rate": 7.2e-05,
      "loss": 6.917,
      "mean_token_accuracy": 0.09355304390192032,
      "num_tokens": 313351.0,
      "step": 145
    },
    {
      "entropy": 7.391125869750977,
      "epoch": 0.014886859865025803,
      "grad_norm": 0.8203125,
      "learning_rate": 7.45e-05,
      "loss": 6.9412,
      "mean_token_accuracy": 0.09017230421304703,
      "num_tokens": 324871.0,
      "step": 150
    },
    {
      "entropy": 7.379152727127075,
      "epoch": 0.01538308852719333,
      "grad_norm": 0.8359375,
      "learning_rate": 7.7e-05,
      "loss": 6.8043,
      "mean_token_accuracy": 0.09457008466124535,
      "num_tokens": 335856.0,
      "step": 155
    },
    {
      "entropy": 7.27963695526123,
      "epoch": 0.015879317189360857,
      "grad_norm": 0.90234375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.8058,
      "mean_token_accuracy": 0.08980907499790192,
      "num_tokens": 345410.0,
      "step": 160
    },
    {
      "entropy": 7.367447996139527,
      "epoch": 0.016375545851528384,
      "grad_norm": 0.85546875,
      "learning_rate": 8.2e-05,
      "loss": 6.8948,
      "mean_token_accuracy": 0.09169329479336738,
      "num_tokens": 356370.0,
      "step": 165
    },
    {
      "entropy": 7.369842910766602,
      "epoch": 0.01687177451369591,
      "grad_norm": 1.140625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.9352,
      "mean_token_accuracy": 0.09325620308518409,
      "num_tokens": 368085.0,
      "step": 170
    },
    {
      "entropy": 7.361319303512573,
      "epoch": 0.01736800317586344,
      "grad_norm": 0.9453125,
      "learning_rate": 8.7e-05,
      "loss": 6.8003,
      "mean_token_accuracy": 0.09971174001693725,
      "num_tokens": 378995.0,
      "step": 175
    },
    {
      "entropy": 7.319345235824585,
      "epoch": 0.017864231838030965,
      "grad_norm": 0.91796875,
      "learning_rate": 8.95e-05,
      "loss": 6.909,
      "mean_token_accuracy": 0.09415822923183441,
      "num_tokens": 390727.0,
      "step": 180
    },
    {
      "entropy": 7.4322614669799805,
      "epoch": 0.018360460500198492,
      "grad_norm": 0.91015625,
      "learning_rate": 9.2e-05,
      "loss": 6.849,
      "mean_token_accuracy": 0.09615265130996704,
      "num_tokens": 401875.0,
      "step": 185
    },
    {
      "entropy": 7.230821657180786,
      "epoch": 0.01885668916236602,
      "grad_norm": 0.87890625,
      "learning_rate": 9.45e-05,
      "loss": 6.8835,
      "mean_token_accuracy": 0.0952032633125782,
      "num_tokens": 414959.0,
      "step": 190
    },
    {
      "entropy": 7.3696495532989506,
      "epoch": 0.019352917824533546,
      "grad_norm": 0.88671875,
      "learning_rate": 9.7e-05,
      "loss": 6.8469,
      "mean_token_accuracy": 0.09475988298654556,
      "num_tokens": 425745.0,
      "step": 195
    },
    {
      "entropy": 7.317271423339844,
      "epoch": 0.019849146486701073,
      "grad_norm": 1.0234375,
      "learning_rate": 9.95e-05,
      "loss": 6.7841,
      "mean_token_accuracy": 0.09933056309819221,
      "num_tokens": 435149.0,
      "step": 200
    },
    {
      "entropy": 7.311147165298462,
      "epoch": 0.0203453751488686,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000102,
      "loss": 6.8521,
      "mean_token_accuracy": 0.09915618151426316,
      "num_tokens": 445706.0,
      "step": 205
    },
    {
      "entropy": 7.283332872390747,
      "epoch": 0.020841603811036124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.839,
      "mean_token_accuracy": 0.10072373151779175,
      "num_tokens": 455982.0,
      "step": 210
    },
    {
      "entropy": 7.41149492263794,
      "epoch": 0.02133783247320365,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000107,
      "loss": 6.8673,
      "mean_token_accuracy": 0.09816423431038857,
      "num_tokens": 465462.0,
      "step": 215
    },
    {
      "entropy": 7.340578031539917,
      "epoch": 0.021834061135371178,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001095,
      "loss": 6.808,
      "mean_token_accuracy": 0.09429572075605393,
      "num_tokens": 475770.0,
      "step": 220
    },
    {
      "entropy": 7.281261396408081,
      "epoch": 0.022330289797538705,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000112,
      "loss": 6.766,
      "mean_token_accuracy": 0.09254665300250053,
      "num_tokens": 486345.0,
      "step": 225
    },
    {
      "entropy": 7.2465247631073,
      "epoch": 0.022826518459706232,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001145,
      "loss": 6.8581,
      "mean_token_accuracy": 0.09497587159276008,
      "num_tokens": 496588.0,
      "step": 230
    },
    {
      "entropy": 7.330821084976196,
      "epoch": 0.02332274712187376,
      "grad_norm": 0.984375,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.764,
      "mean_token_accuracy": 0.10548505261540413,
      "num_tokens": 506676.0,
      "step": 235
    },
    {
      "entropy": 7.259016704559326,
      "epoch": 0.023818975784041286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.7581,
      "mean_token_accuracy": 0.09661711901426315,
      "num_tokens": 516866.0,
      "step": 240
    },
    {
      "entropy": 7.343613529205323,
      "epoch": 0.024315204446208813,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000122,
      "loss": 6.9088,
      "mean_token_accuracy": 0.09192305207252502,
      "num_tokens": 529349.0,
      "step": 245
    },
    {
      "entropy": 7.2918651580810545,
      "epoch": 0.02481143310837634,
      "grad_norm": 0.96875,
      "learning_rate": 0.0001245,
      "loss": 6.8356,
      "mean_token_accuracy": 0.0957394741475582,
      "num_tokens": 541310.0,
      "step": 250
    },
    {
      "entropy": 7.347570466995239,
      "epoch": 0.025307661770543867,
      "grad_norm": 0.984375,
      "learning_rate": 0.000127,
      "loss": 6.8669,
      "mean_token_accuracy": 0.09501437321305276,
      "num_tokens": 551897.0,
      "step": 255
    },
    {
      "entropy": 7.286837148666382,
      "epoch": 0.025803890432711394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001295,
      "loss": 6.8722,
      "mean_token_accuracy": 0.09505478516221047,
      "num_tokens": 564667.0,
      "step": 260
    },
    {
      "entropy": 7.2854108810424805,
      "epoch": 0.02630011909487892,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000132,
      "loss": 6.7365,
      "mean_token_accuracy": 0.09921214655041695,
      "num_tokens": 575338.0,
      "step": 265
    },
    {
      "entropy": 7.331110572814941,
      "epoch": 0.026796347757046448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.8525,
      "mean_token_accuracy": 0.09441650062799453,
      "num_tokens": 586169.0,
      "step": 270
    },
    {
      "entropy": 7.240020895004273,
      "epoch": 0.027292576419213975,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.8366,
      "mean_token_accuracy": 0.09875859245657921,
      "num_tokens": 596864.0,
      "step": 275
    },
    {
      "entropy": 7.349655055999756,
      "epoch": 0.027788805081381502,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0001395,
      "loss": 6.8323,
      "mean_token_accuracy": 0.09189202263951302,
      "num_tokens": 608894.0,
      "step": 280
    },
    {
      "entropy": 7.330462837219239,
      "epoch": 0.02828503374354903,
      "grad_norm": 1.09375,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.8435,
      "mean_token_accuracy": 0.09644725024700165,
      "num_tokens": 619863.0,
      "step": 285
    },
    {
      "entropy": 7.33879017829895,
      "epoch": 0.028781262405716553,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0001445,
      "loss": 6.8891,
      "mean_token_accuracy": 0.09643067345023155,
      "num_tokens": 632691.0,
      "step": 290
    },
    {
      "entropy": 7.232556533813477,
      "epoch": 0.02927749106788408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000147,
      "loss": 6.7748,
      "mean_token_accuracy": 0.09985843300819397,
      "num_tokens": 643993.0,
      "step": 295
    },
    {
      "entropy": 7.320186996459961,
      "epoch": 0.029773719730051607,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001495,
      "loss": 6.7693,
      "mean_token_accuracy": 0.09714792221784592,
      "num_tokens": 654275.0,
      "step": 300
    },
    {
      "entropy": 7.211906003952026,
      "epoch": 0.030269948392219134,
      "grad_norm": 1.015625,
      "learning_rate": 0.000152,
      "loss": 6.7464,
      "mean_token_accuracy": 0.09860079064965248,
      "num_tokens": 664694.0,
      "step": 305
    },
    {
      "entropy": 7.303035020828247,
      "epoch": 0.03076617705438666,
      "grad_norm": 1.15625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.7814,
      "mean_token_accuracy": 0.09757039919495583,
      "num_tokens": 674736.0,
      "step": 310
    },
    {
      "entropy": 7.143521499633789,
      "epoch": 0.03126240571655419,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000157,
      "loss": 6.6907,
      "mean_token_accuracy": 0.100661750882864,
      "num_tokens": 687261.0,
      "step": 315
    },
    {
      "entropy": 7.219635915756226,
      "epoch": 0.031758634378721715,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0001595,
      "loss": 6.736,
      "mean_token_accuracy": 0.10075967833399772,
      "num_tokens": 698904.0,
      "step": 320
    },
    {
      "entropy": 7.224134159088135,
      "epoch": 0.032254863040889245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000162,
      "loss": 6.7152,
      "mean_token_accuracy": 0.10156732127070427,
      "num_tokens": 709035.0,
      "step": 325
    },
    {
      "entropy": 7.154531097412109,
      "epoch": 0.03275109170305677,
      "grad_norm": 1.109375,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.7922,
      "mean_token_accuracy": 0.10236325189471245,
      "num_tokens": 719565.0,
      "step": 330
    },
    {
      "entropy": 7.312559509277344,
      "epoch": 0.03324732036522429,
      "grad_norm": 1.09375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8179,
      "mean_token_accuracy": 0.0999167263507843,
      "num_tokens": 730329.0,
      "step": 335
    },
    {
      "entropy": 7.277273511886596,
      "epoch": 0.03374354902739182,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.81,
      "mean_token_accuracy": 0.09728368297219277,
      "num_tokens": 741258.0,
      "step": 340
    },
    {
      "entropy": 7.169039726257324,
      "epoch": 0.034239777689559346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.6991,
      "mean_token_accuracy": 0.09544317200779914,
      "num_tokens": 752636.0,
      "step": 345
    },
    {
      "entropy": 7.274296426773072,
      "epoch": 0.03473600635172688,
      "grad_norm": 1.078125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.7879,
      "mean_token_accuracy": 0.09527788013219833,
      "num_tokens": 764113.0,
      "step": 350
    },
    {
      "entropy": 7.15718035697937,
      "epoch": 0.0352322350138944,
      "grad_norm": 1.109375,
      "learning_rate": 0.000177,
      "loss": 6.6747,
      "mean_token_accuracy": 0.10008381679654121,
      "num_tokens": 772872.0,
      "step": 355
    },
    {
      "entropy": 7.275865650177002,
      "epoch": 0.03572846367606193,
      "grad_norm": 1.125,
      "learning_rate": 0.0001795,
      "loss": 6.6976,
      "mean_token_accuracy": 0.10527255833148956,
      "num_tokens": 782788.0,
      "step": 360
    },
    {
      "entropy": 7.130430555343628,
      "epoch": 0.036224692338229454,
      "grad_norm": 1.0,
      "learning_rate": 0.000182,
      "loss": 6.7768,
      "mean_token_accuracy": 0.09755202606320382,
      "num_tokens": 793859.0,
      "step": 365
    },
    {
      "entropy": 7.306216430664063,
      "epoch": 0.036720921000396985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001845,
      "loss": 6.7439,
      "mean_token_accuracy": 0.09889776185154915,
      "num_tokens": 804906.0,
      "step": 370
    },
    {
      "entropy": 7.191127681732178,
      "epoch": 0.03721714966256451,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000187,
      "loss": 6.7651,
      "mean_token_accuracy": 0.09736918658018112,
      "num_tokens": 815520.0,
      "step": 375
    },
    {
      "entropy": 7.181806135177612,
      "epoch": 0.03771337832473204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001895,
      "loss": 6.7859,
      "mean_token_accuracy": 0.09997142925858497,
      "num_tokens": 826395.0,
      "step": 380
    },
    {
      "entropy": 7.2829760074615475,
      "epoch": 0.03820960698689956,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000192,
      "loss": 6.732,
      "mean_token_accuracy": 0.09944447204470634,
      "num_tokens": 836627.0,
      "step": 385
    },
    {
      "entropy": 7.153994750976563,
      "epoch": 0.03870583564906709,
      "grad_norm": 0.90625,
      "learning_rate": 0.0001945,
      "loss": 6.7706,
      "mean_token_accuracy": 0.09017840772867203,
      "num_tokens": 848250.0,
      "step": 390
    },
    {
      "entropy": 7.316059160232544,
      "epoch": 0.039202064311234616,
      "grad_norm": 0.90625,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7703,
      "mean_token_accuracy": 0.09917276129126548,
      "num_tokens": 859195.0,
      "step": 395
    },
    {
      "entropy": 7.160914516448974,
      "epoch": 0.03969829297340215,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.7453,
      "mean_token_accuracy": 0.10009187385439873,
      "num_tokens": 871955.0,
      "step": 400
    },
    {
      "entropy": 7.1762594223022464,
      "epoch": 0.04019452163556967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000202,
      "loss": 6.7606,
      "mean_token_accuracy": 0.09773519337177276,
      "num_tokens": 882729.0,
      "step": 405
    },
    {
      "entropy": 7.1601005554199215,
      "epoch": 0.0406907502977372,
      "grad_norm": 1.125,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.6661,
      "mean_token_accuracy": 0.09903571754693985,
      "num_tokens": 893183.0,
      "step": 410
    },
    {
      "entropy": 7.2316563606262205,
      "epoch": 0.041186978959904724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000207,
      "loss": 6.7912,
      "mean_token_accuracy": 0.10149758830666541,
      "num_tokens": 904988.0,
      "step": 415
    },
    {
      "entropy": 7.237285995483399,
      "epoch": 0.04168320762207225,
      "grad_norm": 1.0,
      "learning_rate": 0.0002095,
      "loss": 6.6543,
      "mean_token_accuracy": 0.10472753047943115,
      "num_tokens": 915192.0,
      "step": 420
    },
    {
      "entropy": 7.109501314163208,
      "epoch": 0.04217943628423978,
      "grad_norm": 1.078125,
      "learning_rate": 0.000212,
      "loss": 6.6954,
      "mean_token_accuracy": 0.10364402532577514,
      "num_tokens": 926494.0,
      "step": 425
    },
    {
      "entropy": 7.216678476333618,
      "epoch": 0.0426756649464073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002145,
      "loss": 6.702,
      "mean_token_accuracy": 0.10059248208999634,
      "num_tokens": 938376.0,
      "step": 430
    },
    {
      "entropy": 7.132107162475586,
      "epoch": 0.04317189360857483,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.6097,
      "mean_token_accuracy": 0.1056530199944973,
      "num_tokens": 948640.0,
      "step": 435
    },
    {
      "entropy": 7.096781969070435,
      "epoch": 0.043668122270742356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002195,
      "loss": 6.7184,
      "mean_token_accuracy": 0.09810023754835129,
      "num_tokens": 959759.0,
      "step": 440
    },
    {
      "entropy": 7.138823080062866,
      "epoch": 0.044164350932909886,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000222,
      "loss": 6.6591,
      "mean_token_accuracy": 0.09683494120836258,
      "num_tokens": 970137.0,
      "step": 445
    },
    {
      "entropy": 7.140137815475464,
      "epoch": 0.04466057959507741,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002245,
      "loss": 6.6989,
      "mean_token_accuracy": 0.09436362460255623,
      "num_tokens": 979875.0,
      "step": 450
    },
    {
      "entropy": 7.076799058914185,
      "epoch": 0.04515680825724494,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6654,
      "mean_token_accuracy": 0.10611343160271644,
      "num_tokens": 990324.0,
      "step": 455
    },
    {
      "entropy": 7.221283626556397,
      "epoch": 0.045653036919412464,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.6791,
      "mean_token_accuracy": 0.10299149453639984,
      "num_tokens": 1000259.0,
      "step": 460
    },
    {
      "entropy": 7.089738178253174,
      "epoch": 0.046149265581579994,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.671,
      "mean_token_accuracy": 0.10547975078225136,
      "num_tokens": 1012220.0,
      "step": 465
    },
    {
      "entropy": 7.1531280994415285,
      "epoch": 0.04664549424374752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6543,
      "mean_token_accuracy": 0.10063310712575912,
      "num_tokens": 1022968.0,
      "step": 470
    },
    {
      "entropy": 7.094948148727417,
      "epoch": 0.04714172290591505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000237,
      "loss": 6.6873,
      "mean_token_accuracy": 0.10046143010258675,
      "num_tokens": 1033415.0,
      "step": 475
    },
    {
      "entropy": 7.113246965408325,
      "epoch": 0.04763795156808257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002395,
      "loss": 6.597,
      "mean_token_accuracy": 0.10691144168376923,
      "num_tokens": 1045384.0,
      "step": 480
    },
    {
      "entropy": 7.089356851577759,
      "epoch": 0.0481341802302501,
      "grad_norm": 1.109375,
      "learning_rate": 0.000242,
      "loss": 6.6934,
      "mean_token_accuracy": 0.10633598864078522,
      "num_tokens": 1054982.0,
      "step": 485
    },
    {
      "entropy": 7.126608800888062,
      "epoch": 0.048630408892417626,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002445,
      "loss": 6.6344,
      "mean_token_accuracy": 0.10542937964200974,
      "num_tokens": 1065244.0,
      "step": 490
    },
    {
      "entropy": 7.122588300704956,
      "epoch": 0.04912663755458515,
      "grad_norm": 1.109375,
      "learning_rate": 0.000247,
      "loss": 6.6503,
      "mean_token_accuracy": 0.09597428143024445,
      "num_tokens": 1075236.0,
      "step": 495
    },
    {
      "entropy": 7.064780902862549,
      "epoch": 0.04962286621675268,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002495,
      "loss": 6.6788,
      "mean_token_accuracy": 0.09427490904927253,
      "num_tokens": 1085397.0,
      "step": 500
    },
    {
      "entropy": 7.108021450042725,
      "epoch": 0.0501190948789202,
      "grad_norm": 1.1875,
      "learning_rate": 0.000252,
      "loss": 6.689,
      "mean_token_accuracy": 0.10125237107276916,
      "num_tokens": 1095207.0,
      "step": 505
    },
    {
      "entropy": 7.1391456604003904,
      "epoch": 0.050615323541087734,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002545,
      "loss": 6.6822,
      "mean_token_accuracy": 0.09262849614024163,
      "num_tokens": 1105834.0,
      "step": 510
    },
    {
      "entropy": 7.115177631378174,
      "epoch": 0.05111155220325526,
      "grad_norm": 1.015625,
      "learning_rate": 0.000257,
      "loss": 6.6838,
      "mean_token_accuracy": 0.1050037071108818,
      "num_tokens": 1116105.0,
      "step": 515
    },
    {
      "entropy": 7.111266660690307,
      "epoch": 0.05160778086542279,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002595,
      "loss": 6.6664,
      "mean_token_accuracy": 0.10681836828589439,
      "num_tokens": 1125298.0,
      "step": 520
    },
    {
      "entropy": 7.122176456451416,
      "epoch": 0.05210400952759031,
      "grad_norm": 0.84375,
      "learning_rate": 0.000262,
      "loss": 6.6676,
      "mean_token_accuracy": 0.10127982199192047,
      "num_tokens": 1136797.0,
      "step": 525
    },
    {
      "entropy": 6.9743858814239506,
      "epoch": 0.05260023818975784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6766,
      "mean_token_accuracy": 0.10087173655629159,
      "num_tokens": 1148674.0,
      "step": 530
    },
    {
      "entropy": 7.159593200683593,
      "epoch": 0.053096466851925365,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7479,
      "mean_token_accuracy": 0.09798325300216675,
      "num_tokens": 1160329.0,
      "step": 535
    },
    {
      "entropy": 7.162366485595703,
      "epoch": 0.053592695514092896,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.6908,
      "mean_token_accuracy": 0.09893349334597587,
      "num_tokens": 1170318.0,
      "step": 540
    },
    {
      "entropy": 6.988642835617066,
      "epoch": 0.05408892417626042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5561,
      "mean_token_accuracy": 0.11139835193753242,
      "num_tokens": 1181634.0,
      "step": 545
    },
    {
      "entropy": 7.042670631408692,
      "epoch": 0.05458515283842795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002745,
      "loss": 6.6219,
      "mean_token_accuracy": 0.10727389231324196,
      "num_tokens": 1191829.0,
      "step": 550
    },
    {
      "entropy": 6.997174549102783,
      "epoch": 0.055081381500595473,
      "grad_norm": 1.15625,
      "learning_rate": 0.000277,
      "loss": 6.61,
      "mean_token_accuracy": 0.10348387435078621,
      "num_tokens": 1202265.0,
      "step": 555
    },
    {
      "entropy": 7.030418825149536,
      "epoch": 0.055577610162763004,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002795,
      "loss": 6.6235,
      "mean_token_accuracy": 0.10154768601059913,
      "num_tokens": 1214173.0,
      "step": 560
    },
    {
      "entropy": 7.092197895050049,
      "epoch": 0.05607383882493053,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.5238,
      "mean_token_accuracy": 0.10746525302529335,
      "num_tokens": 1224582.0,
      "step": 565
    },
    {
      "entropy": 7.052742052078247,
      "epoch": 0.05657006748709806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002845,
      "loss": 6.7535,
      "mean_token_accuracy": 0.09928297400474548,
      "num_tokens": 1235865.0,
      "step": 570
    },
    {
      "entropy": 6.987533903121948,
      "epoch": 0.05706629614926558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000287,
      "loss": 6.5779,
      "mean_token_accuracy": 0.10504852086305619,
      "num_tokens": 1247194.0,
      "step": 575
    },
    {
      "entropy": 6.921913146972656,
      "epoch": 0.057562524811433105,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002895,
      "loss": 6.59,
      "mean_token_accuracy": 0.09689045846462249,
      "num_tokens": 1257981.0,
      "step": 580
    },
    {
      "entropy": 7.040601444244385,
      "epoch": 0.058058753473600636,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000292,
      "loss": 6.553,
      "mean_token_accuracy": 0.10087601393461228,
      "num_tokens": 1268443.0,
      "step": 585
    },
    {
      "entropy": 7.027150583267212,
      "epoch": 0.05855498213576816,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002945,
      "loss": 6.666,
      "mean_token_accuracy": 0.10131042078137398,
      "num_tokens": 1279134.0,
      "step": 590
    },
    {
      "entropy": 6.966839933395386,
      "epoch": 0.05905121079793569,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000297,
      "loss": 6.6183,
      "mean_token_accuracy": 0.09604747295379638,
      "num_tokens": 1289965.0,
      "step": 595
    },
    {
      "entropy": 7.076975393295288,
      "epoch": 0.05954743946010321,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002995,
      "loss": 6.5411,
      "mean_token_accuracy": 0.10592939630150795,
      "num_tokens": 1300003.0,
      "step": 600
    },
    {
      "entropy": 6.961626815795898,
      "epoch": 0.060043668122270744,
      "grad_norm": 0.96875,
      "learning_rate": 0.000302,
      "loss": 6.608,
      "mean_token_accuracy": 0.10380170494318008,
      "num_tokens": 1311990.0,
      "step": 605
    },
    {
      "entropy": 6.9057409286499025,
      "epoch": 0.06053989678443827,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003045,
      "loss": 6.6506,
      "mean_token_accuracy": 0.10116632953286171,
      "num_tokens": 1322963.0,
      "step": 610
    },
    {
      "entropy": 7.0877197742462155,
      "epoch": 0.0610361254466058,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000307,
      "loss": 6.6589,
      "mean_token_accuracy": 0.09740115106105804,
      "num_tokens": 1334580.0,
      "step": 615
    },
    {
      "entropy": 7.017366170883179,
      "epoch": 0.06153235410877332,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003095,
      "loss": 6.6027,
      "mean_token_accuracy": 0.10138308554887772,
      "num_tokens": 1346545.0,
      "step": 620
    },
    {
      "entropy": 7.022233295440674,
      "epoch": 0.06202858277094085,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000312,
      "loss": 6.5788,
      "mean_token_accuracy": 0.10256645902991295,
      "num_tokens": 1358375.0,
      "step": 625
    },
    {
      "entropy": 6.930924749374389,
      "epoch": 0.06252481143310838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003145,
      "loss": 6.6474,
      "mean_token_accuracy": 0.09666800163686276,
      "num_tokens": 1368376.0,
      "step": 630
    },
    {
      "entropy": 7.004555892944336,
      "epoch": 0.0630210400952759,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000317,
      "loss": 6.539,
      "mean_token_accuracy": 0.1008249580860138,
      "num_tokens": 1377749.0,
      "step": 635
    },
    {
      "entropy": 6.925582361221314,
      "epoch": 0.06351726875744343,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003195,
      "loss": 6.6182,
      "mean_token_accuracy": 0.1026754193007946,
      "num_tokens": 1389581.0,
      "step": 640
    },
    {
      "entropy": 6.8830712795257565,
      "epoch": 0.06401349741961096,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000322,
      "loss": 6.5373,
      "mean_token_accuracy": 0.10198206380009651,
      "num_tokens": 1399772.0,
      "step": 645
    },
    {
      "entropy": 6.992895269393921,
      "epoch": 0.06450972608177849,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6367,
      "mean_token_accuracy": 0.10756797865033149,
      "num_tokens": 1411158.0,
      "step": 650
    },
    {
      "entropy": 6.994899320602417,
      "epoch": 0.065005954743946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5465,
      "mean_token_accuracy": 0.1033245749771595,
      "num_tokens": 1420651.0,
      "step": 655
    },
    {
      "entropy": 6.84561710357666,
      "epoch": 0.06550218340611354,
      "grad_norm": 1.125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.4921,
      "mean_token_accuracy": 0.10640839785337448,
      "num_tokens": 1431628.0,
      "step": 660
    },
    {
      "entropy": 6.88475866317749,
      "epoch": 0.06599841206828107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5142,
      "mean_token_accuracy": 0.10272238478064537,
      "num_tokens": 1441850.0,
      "step": 665
    },
    {
      "entropy": 7.0489623069763185,
      "epoch": 0.06649464073044858,
      "grad_norm": 1.296875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.5969,
      "mean_token_accuracy": 0.10575282126665116,
      "num_tokens": 1452393.0,
      "step": 670
    },
    {
      "entropy": 6.8803466796875,
      "epoch": 0.06699086939261611,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000337,
      "loss": 6.5199,
      "mean_token_accuracy": 0.10531659424304962,
      "num_tokens": 1463109.0,
      "step": 675
    },
    {
      "entropy": 6.860003042221069,
      "epoch": 0.06748709805478365,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003395,
      "loss": 6.6383,
      "mean_token_accuracy": 0.1039490133523941,
      "num_tokens": 1475082.0,
      "step": 680
    },
    {
      "entropy": 6.982773113250732,
      "epoch": 0.06798332671695118,
      "grad_norm": 1.046875,
      "learning_rate": 0.000342,
      "loss": 6.5005,
      "mean_token_accuracy": 0.10786980241537095,
      "num_tokens": 1485477.0,
      "step": 685
    },
    {
      "entropy": 6.865319013595581,
      "epoch": 0.06847955537911869,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4864,
      "mean_token_accuracy": 0.10533245354890823,
      "num_tokens": 1495422.0,
      "step": 690
    },
    {
      "entropy": 6.881680583953857,
      "epoch": 0.06897578404128622,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000347,
      "loss": 6.5021,
      "mean_token_accuracy": 0.10336431041359902,
      "num_tokens": 1506368.0,
      "step": 695
    },
    {
      "entropy": 6.89243106842041,
      "epoch": 0.06947201270345375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003495,
      "loss": 6.5205,
      "mean_token_accuracy": 0.09701511636376381,
      "num_tokens": 1517437.0,
      "step": 700
    },
    {
      "entropy": 6.802684307098389,
      "epoch": 0.06996824136562128,
      "grad_norm": 1.109375,
      "learning_rate": 0.000352,
      "loss": 6.4691,
      "mean_token_accuracy": 0.10144305750727653,
      "num_tokens": 1527801.0,
      "step": 705
    },
    {
      "entropy": 6.986940240859985,
      "epoch": 0.0704644700277888,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003545,
      "loss": 6.5517,
      "mean_token_accuracy": 0.09867026731371879,
      "num_tokens": 1537621.0,
      "step": 710
    },
    {
      "entropy": 6.879312801361084,
      "epoch": 0.07096069868995633,
      "grad_norm": 1.015625,
      "learning_rate": 0.000357,
      "loss": 6.5558,
      "mean_token_accuracy": 0.10100790783762932,
      "num_tokens": 1549123.0,
      "step": 715
    },
    {
      "entropy": 6.8572304248809814,
      "epoch": 0.07145692735212386,
      "grad_norm": 1.125,
      "learning_rate": 0.0003595,
      "loss": 6.5562,
      "mean_token_accuracy": 0.10545549392700196,
      "num_tokens": 1561167.0,
      "step": 720
    },
    {
      "entropy": 6.959098386764526,
      "epoch": 0.07195315601429139,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000362,
      "loss": 6.4677,
      "mean_token_accuracy": 0.11058438867330551,
      "num_tokens": 1571552.0,
      "step": 725
    },
    {
      "entropy": 6.7413163661956785,
      "epoch": 0.07244938467645891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003645,
      "loss": 6.4498,
      "mean_token_accuracy": 0.10494914427399635,
      "num_tokens": 1582558.0,
      "step": 730
    },
    {
      "entropy": 6.710915040969849,
      "epoch": 0.07294561333862644,
      "grad_norm": 1.015625,
      "learning_rate": 0.000367,
      "loss": 6.4301,
      "mean_token_accuracy": 0.10837240293622016,
      "num_tokens": 1595259.0,
      "step": 735
    },
    {
      "entropy": 6.876967239379883,
      "epoch": 0.07344184200079397,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003695,
      "loss": 6.4578,
      "mean_token_accuracy": 0.108952496945858,
      "num_tokens": 1605585.0,
      "step": 740
    },
    {
      "entropy": 6.710470485687256,
      "epoch": 0.07393807066296149,
      "grad_norm": 0.984375,
      "learning_rate": 0.000372,
      "loss": 6.4133,
      "mean_token_accuracy": 0.10763250440359115,
      "num_tokens": 1616529.0,
      "step": 745
    },
    {
      "entropy": 6.689630460739136,
      "epoch": 0.07443429932512902,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003745,
      "loss": 6.3907,
      "mean_token_accuracy": 0.10895837992429733,
      "num_tokens": 1626223.0,
      "step": 750
    },
    {
      "entropy": 6.887076997756958,
      "epoch": 0.07493052798729655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000377,
      "loss": 6.5011,
      "mean_token_accuracy": 0.10325393304228783,
      "num_tokens": 1638173.0,
      "step": 755
    },
    {
      "entropy": 6.611377286911011,
      "epoch": 0.07542675664946408,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003795,
      "loss": 6.3378,
      "mean_token_accuracy": 0.11054935604333878,
      "num_tokens": 1648919.0,
      "step": 760
    },
    {
      "entropy": 6.933552312850952,
      "epoch": 0.0759229853116316,
      "grad_norm": 0.96875,
      "learning_rate": 0.000382,
      "loss": 6.5917,
      "mean_token_accuracy": 0.09958057925105095,
      "num_tokens": 1660704.0,
      "step": 765
    },
    {
      "entropy": 6.766306972503662,
      "epoch": 0.07641921397379912,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003845,
      "loss": 6.49,
      "mean_token_accuracy": 0.10253961160778999,
      "num_tokens": 1672080.0,
      "step": 770
    },
    {
      "entropy": 6.803144121170044,
      "epoch": 0.07691544263596665,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.473,
      "mean_token_accuracy": 0.10405055657029152,
      "num_tokens": 1681930.0,
      "step": 775
    },
    {
      "entropy": 6.825759029388427,
      "epoch": 0.07741167129813419,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.4713,
      "mean_token_accuracy": 0.10618199408054352,
      "num_tokens": 1693075.0,
      "step": 780
    },
    {
      "entropy": 6.776135778427124,
      "epoch": 0.0779078999603017,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.4653,
      "mean_token_accuracy": 0.10373328626155853,
      "num_tokens": 1704163.0,
      "step": 785
    },
    {
      "entropy": 6.840697479248047,
      "epoch": 0.07840412862246923,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.5917,
      "mean_token_accuracy": 0.09998267292976379,
      "num_tokens": 1714879.0,
      "step": 790
    },
    {
      "entropy": 6.849323797225952,
      "epoch": 0.07890035728463676,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.5946,
      "mean_token_accuracy": 0.1017768420279026,
      "num_tokens": 1725464.0,
      "step": 795
    },
    {
      "entropy": 6.930966424942016,
      "epoch": 0.0793965859468043,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003995,
      "loss": 6.5356,
      "mean_token_accuracy": 0.10469997599720955,
      "num_tokens": 1736947.0,
      "step": 800
    },
    {
      "entropy": 6.695473623275757,
      "epoch": 0.07989281460897181,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000402,
      "loss": 6.3539,
      "mean_token_accuracy": 0.11638421043753625,
      "num_tokens": 1746827.0,
      "step": 805
    },
    {
      "entropy": 6.899624872207641,
      "epoch": 0.08038904327113934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004045,
      "loss": 6.5774,
      "mean_token_accuracy": 0.10568423792719842,
      "num_tokens": 1757079.0,
      "step": 810
    },
    {
      "entropy": 6.826183795928955,
      "epoch": 0.08088527193330687,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5054,
      "mean_token_accuracy": 0.1021289974451065,
      "num_tokens": 1767684.0,
      "step": 815
    },
    {
      "entropy": 6.750361824035645,
      "epoch": 0.0813815005954744,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004095,
      "loss": 6.5018,
      "mean_token_accuracy": 0.10343262404203415,
      "num_tokens": 1778187.0,
      "step": 820
    },
    {
      "entropy": 6.833082675933838,
      "epoch": 0.08187772925764192,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412,
      "loss": 6.4608,
      "mean_token_accuracy": 0.10938398912549019,
      "num_tokens": 1789139.0,
      "step": 825
    },
    {
      "entropy": 6.733291482925415,
      "epoch": 0.08237395791980945,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004145,
      "loss": 6.3351,
      "mean_token_accuracy": 0.1136176273226738,
      "num_tokens": 1799328.0,
      "step": 830
    },
    {
      "entropy": 6.741356658935547,
      "epoch": 0.08287018658197698,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000417,
      "loss": 6.496,
      "mean_token_accuracy": 0.10510264784097671,
      "num_tokens": 1811487.0,
      "step": 835
    },
    {
      "entropy": 6.685542392730713,
      "epoch": 0.0833664152441445,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004195,
      "loss": 6.3968,
      "mean_token_accuracy": 0.11166923865675926,
      "num_tokens": 1821923.0,
      "step": 840
    },
    {
      "entropy": 6.720544052124024,
      "epoch": 0.08386264390631203,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000422,
      "loss": 6.402,
      "mean_token_accuracy": 0.10754152461886406,
      "num_tokens": 1833001.0,
      "step": 845
    },
    {
      "entropy": 6.761799335479736,
      "epoch": 0.08435887256847956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004245,
      "loss": 6.4346,
      "mean_token_accuracy": 0.10533890500664711,
      "num_tokens": 1844473.0,
      "step": 850
    },
    {
      "entropy": 6.626788568496704,
      "epoch": 0.08485510123064709,
      "grad_norm": 0.96875,
      "learning_rate": 0.000427,
      "loss": 6.3387,
      "mean_token_accuracy": 0.11210601925849914,
      "num_tokens": 1855851.0,
      "step": 855
    },
    {
      "entropy": 6.650875186920166,
      "epoch": 0.0853513298928146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004295,
      "loss": 6.3262,
      "mean_token_accuracy": 0.11055968850851058,
      "num_tokens": 1866712.0,
      "step": 860
    },
    {
      "entropy": 6.77458963394165,
      "epoch": 0.08584755855498213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000432,
      "loss": 6.3713,
      "mean_token_accuracy": 0.10980789661407471,
      "num_tokens": 1876443.0,
      "step": 865
    },
    {
      "entropy": 6.640477132797241,
      "epoch": 0.08634378721714966,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004345,
      "loss": 6.4319,
      "mean_token_accuracy": 0.10897266939282417,
      "num_tokens": 1886239.0,
      "step": 870
    },
    {
      "entropy": 6.729007625579834,
      "epoch": 0.0868400158793172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000437,
      "loss": 6.3736,
      "mean_token_accuracy": 0.11199768409132957,
      "num_tokens": 1896457.0,
      "step": 875
    },
    {
      "entropy": 6.60158052444458,
      "epoch": 0.08733624454148471,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004395,
      "loss": 6.3631,
      "mean_token_accuracy": 0.10282145217061042,
      "num_tokens": 1908068.0,
      "step": 880
    },
    {
      "entropy": 6.7822535037994385,
      "epoch": 0.08783247320365224,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000442,
      "loss": 6.5282,
      "mean_token_accuracy": 0.10227628946304321,
      "num_tokens": 1920093.0,
      "step": 885
    },
    {
      "entropy": 6.631855154037476,
      "epoch": 0.08832870186581977,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004445,
      "loss": 6.312,
      "mean_token_accuracy": 0.11389266178011895,
      "num_tokens": 1930401.0,
      "step": 890
    },
    {
      "entropy": 6.710539054870606,
      "epoch": 0.0888249305279873,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000447,
      "loss": 6.5031,
      "mean_token_accuracy": 0.10445247143507004,
      "num_tokens": 1940163.0,
      "step": 895
    },
    {
      "entropy": 6.6758523941040036,
      "epoch": 0.08932115919015482,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3394,
      "mean_token_accuracy": 0.10623634234070778,
      "num_tokens": 1952455.0,
      "step": 900
    },
    {
      "entropy": 6.620533514022827,
      "epoch": 0.08981738785232235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4614,
      "mean_token_accuracy": 0.1033598780632019,
      "num_tokens": 1965079.0,
      "step": 905
    },
    {
      "entropy": 6.606557559967041,
      "epoch": 0.09031361651448988,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.3391,
      "mean_token_accuracy": 0.10922778472304344,
      "num_tokens": 1976620.0,
      "step": 910
    },
    {
      "entropy": 6.650234937667847,
      "epoch": 0.0908098451766574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3271,
      "mean_token_accuracy": 0.10684251859784126,
      "num_tokens": 1987626.0,
      "step": 915
    },
    {
      "entropy": 6.633430910110474,
      "epoch": 0.09130607383882493,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3545,
      "mean_token_accuracy": 0.10809084251523018,
      "num_tokens": 1998250.0,
      "step": 920
    },
    {
      "entropy": 6.696629190444947,
      "epoch": 0.09180230250099246,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000462,
      "loss": 6.3446,
      "mean_token_accuracy": 0.11614455357193947,
      "num_tokens": 2008098.0,
      "step": 925
    },
    {
      "entropy": 6.514123439788818,
      "epoch": 0.09229853116315999,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004645,
      "loss": 6.3045,
      "mean_token_accuracy": 0.10994884520769119,
      "num_tokens": 2020125.0,
      "step": 930
    },
    {
      "entropy": 6.611006355285644,
      "epoch": 0.0927947598253275,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000467,
      "loss": 6.4379,
      "mean_token_accuracy": 0.11345969960093498,
      "num_tokens": 2030791.0,
      "step": 935
    },
    {
      "entropy": 6.681372880935669,
      "epoch": 0.09329098848749504,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004695,
      "loss": 6.4183,
      "mean_token_accuracy": 0.1092501424252987,
      "num_tokens": 2042358.0,
      "step": 940
    },
    {
      "entropy": 6.599484062194824,
      "epoch": 0.09378721714966257,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000472,
      "loss": 6.3679,
      "mean_token_accuracy": 0.11037536859512329,
      "num_tokens": 2054549.0,
      "step": 945
    },
    {
      "entropy": 6.597382736206055,
      "epoch": 0.0942834458118301,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004745,
      "loss": 6.3376,
      "mean_token_accuracy": 0.11201154738664627,
      "num_tokens": 2064903.0,
      "step": 950
    },
    {
      "entropy": 6.615338516235352,
      "epoch": 0.09477967447399761,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000477,
      "loss": 6.3146,
      "mean_token_accuracy": 0.10975310951471329,
      "num_tokens": 2075795.0,
      "step": 955
    },
    {
      "entropy": 6.57103590965271,
      "epoch": 0.09527590313616514,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004795,
      "loss": 6.3637,
      "mean_token_accuracy": 0.10602563321590423,
      "num_tokens": 2087249.0,
      "step": 960
    },
    {
      "entropy": 6.655970811843872,
      "epoch": 0.09577213179833267,
      "grad_norm": 1.0625,
      "learning_rate": 0.000482,
      "loss": 6.3354,
      "mean_token_accuracy": 0.11833583116531372,
      "num_tokens": 2097527.0,
      "step": 965
    },
    {
      "entropy": 6.562977027893067,
      "epoch": 0.0962683604605002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004845,
      "loss": 6.3478,
      "mean_token_accuracy": 0.11001443713903428,
      "num_tokens": 2108174.0,
      "step": 970
    },
    {
      "entropy": 6.554857492446899,
      "epoch": 0.09676458912266772,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000487,
      "loss": 6.3492,
      "mean_token_accuracy": 0.1084967702627182,
      "num_tokens": 2120355.0,
      "step": 975
    },
    {
      "entropy": 6.574288702011108,
      "epoch": 0.09726081778483525,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004895,
      "loss": 6.2304,
      "mean_token_accuracy": 0.11568673327565193,
      "num_tokens": 2130861.0,
      "step": 980
    },
    {
      "entropy": 6.584482145309448,
      "epoch": 0.09775704644700278,
      "grad_norm": 1.0,
      "learning_rate": 0.000492,
      "loss": 6.3911,
      "mean_token_accuracy": 0.1085393950343132,
      "num_tokens": 2142162.0,
      "step": 985
    },
    {
      "entropy": 6.594986820220948,
      "epoch": 0.0982532751091703,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945,
      "loss": 6.318,
      "mean_token_accuracy": 0.11034402921795845,
      "num_tokens": 2153374.0,
      "step": 990
    },
    {
      "entropy": 6.524502325057983,
      "epoch": 0.09874950377133783,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497,
      "loss": 6.2594,
      "mean_token_accuracy": 0.10932201892137527,
      "num_tokens": 2162941.0,
      "step": 995
    },
    {
      "entropy": 6.649286603927612,
      "epoch": 0.09924573243350536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995,
      "loss": 6.3437,
      "mean_token_accuracy": 0.09899675399065018,
      "num_tokens": 2172864.0,
      "step": 1000
    },
    {
      "entropy": 6.512604904174805,
      "epoch": 0.09974196109567289,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999999982145551,
      "loss": 6.3648,
      "mean_token_accuracy": 0.11348994895815849,
      "num_tokens": 2184219.0,
      "step": 1005
    },
    {
      "entropy": 6.696617507934571,
      "epoch": 0.1002381897578404,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999999909611855,
      "loss": 6.3227,
      "mean_token_accuracy": 0.10469643399119377,
      "num_tokens": 2194788.0,
      "step": 1010
    },
    {
      "entropy": 6.4998678207397464,
      "epoch": 0.10073441842000794,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999999781283008,
      "loss": 6.3441,
      "mean_token_accuracy": 0.10121865943074226,
      "num_tokens": 2205493.0,
      "step": 1015
    },
    {
      "entropy": 6.622814416885376,
      "epoch": 0.10123064708217547,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999999597159017,
      "loss": 6.3603,
      "mean_token_accuracy": 0.1107247956097126,
      "num_tokens": 2215354.0,
      "step": 1020
    },
    {
      "entropy": 6.516401863098144,
      "epoch": 0.101726875744343,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999999357239882,
      "loss": 6.2838,
      "mean_token_accuracy": 0.11005397066473961,
      "num_tokens": 2225781.0,
      "step": 1025
    },
    {
      "entropy": 6.545648241043091,
      "epoch": 0.10222310440651051,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999999061525612,
      "loss": 6.3517,
      "mean_token_accuracy": 0.10768992081284523,
      "num_tokens": 2236008.0,
      "step": 1030
    },
    {
      "entropy": 6.544310569763184,
      "epoch": 0.10271933306867805,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999998710016214,
      "loss": 6.279,
      "mean_token_accuracy": 0.1109216794371605,
      "num_tokens": 2246998.0,
      "step": 1035
    },
    {
      "entropy": 6.540542888641357,
      "epoch": 0.10321556173084558,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999998302711696,
      "loss": 6.3958,
      "mean_token_accuracy": 0.10798945054411888,
      "num_tokens": 2257451.0,
      "step": 1040
    },
    {
      "entropy": 6.618514966964722,
      "epoch": 0.1037117903930131,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999997839612069,
      "loss": 6.3545,
      "mean_token_accuracy": 0.1071122370660305,
      "num_tokens": 2268151.0,
      "step": 1045
    },
    {
      "entropy": 6.563951921463013,
      "epoch": 0.10420801905518062,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999997320717342,
      "loss": 6.3776,
      "mean_token_accuracy": 0.10482034385204315,
      "num_tokens": 2278924.0,
      "step": 1050
    },
    {
      "entropy": 6.501388168334961,
      "epoch": 0.10470424771734815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999996746027531,
      "loss": 6.3032,
      "mean_token_accuracy": 0.11530704945325851,
      "num_tokens": 2289229.0,
      "step": 1055
    },
    {
      "entropy": 6.539397239685059,
      "epoch": 0.10520047637951568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499999611554265,
      "loss": 6.3215,
      "mean_token_accuracy": 0.1085787646472454,
      "num_tokens": 2300203.0,
      "step": 1060
    },
    {
      "entropy": 6.557326984405518,
      "epoch": 0.10569670504168321,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999995429262712,
      "loss": 6.3026,
      "mean_token_accuracy": 0.11185992807149887,
      "num_tokens": 2311991.0,
      "step": 1065
    },
    {
      "entropy": 6.499863815307617,
      "epoch": 0.10619293370385073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999994687187737,
      "loss": 6.2687,
      "mean_token_accuracy": 0.10764180719852448,
      "num_tokens": 2322110.0,
      "step": 1070
    },
    {
      "entropy": 6.537433528900147,
      "epoch": 0.10668916236601826,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999993889317742,
      "loss": 6.3196,
      "mean_token_accuracy": 0.1101168490946293,
      "num_tokens": 2332439.0,
      "step": 1075
    },
    {
      "entropy": 6.4248419284820555,
      "epoch": 0.10718539102818579,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999993035652746,
      "loss": 6.1852,
      "mean_token_accuracy": 0.11721493899822236,
      "num_tokens": 2342339.0,
      "step": 1080
    },
    {
      "entropy": 6.510436201095581,
      "epoch": 0.10768161969035331,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999992126192771,
      "loss": 6.324,
      "mean_token_accuracy": 0.11845446228981019,
      "num_tokens": 2353211.0,
      "step": 1085
    },
    {
      "entropy": 6.59377236366272,
      "epoch": 0.10817784835252084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499999116093784,
      "loss": 6.363,
      "mean_token_accuracy": 0.10857008323073387,
      "num_tokens": 2364709.0,
      "step": 1090
    },
    {
      "entropy": 6.536841201782226,
      "epoch": 0.10867407701468837,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999990139887976,
      "loss": 6.2863,
      "mean_token_accuracy": 0.10694537609815598,
      "num_tokens": 2375095.0,
      "step": 1095
    },
    {
      "entropy": 6.508898782730102,
      "epoch": 0.1091703056768559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999989063043206,
      "loss": 6.2757,
      "mean_token_accuracy": 0.11252731531858444,
      "num_tokens": 2384206.0,
      "step": 1100
    },
    {
      "entropy": 6.445655059814453,
      "epoch": 0.10966653433902342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999987930403555,
      "loss": 6.1962,
      "mean_token_accuracy": 0.11554479971528053,
      "num_tokens": 2395139.0,
      "step": 1105
    },
    {
      "entropy": 6.4702643871307375,
      "epoch": 0.11016276300119095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999986741969051,
      "loss": 6.3234,
      "mean_token_accuracy": 0.10869933143258095,
      "num_tokens": 2405652.0,
      "step": 1110
    },
    {
      "entropy": 6.626315641403198,
      "epoch": 0.11065899166335848,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999985497739724,
      "loss": 6.3295,
      "mean_token_accuracy": 0.10884599313139916,
      "num_tokens": 2416386.0,
      "step": 1115
    },
    {
      "entropy": 6.404276943206787,
      "epoch": 0.11115522032552601,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999984197715606,
      "loss": 6.3425,
      "mean_token_accuracy": 0.10958480238914489,
      "num_tokens": 2427807.0,
      "step": 1120
    },
    {
      "entropy": 6.543959474563598,
      "epoch": 0.11165144898769352,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999982841896728,
      "loss": 6.2502,
      "mean_token_accuracy": 0.11363493949174881,
      "num_tokens": 2439481.0,
      "step": 1125
    },
    {
      "entropy": 6.478694009780884,
      "epoch": 0.11214767764986106,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999981430283123,
      "loss": 6.2981,
      "mean_token_accuracy": 0.11177865490317344,
      "num_tokens": 2449857.0,
      "step": 1130
    },
    {
      "entropy": 6.431196975708008,
      "epoch": 0.11264390631202859,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999979962874827,
      "loss": 6.2159,
      "mean_token_accuracy": 0.11506233438849449,
      "num_tokens": 2461569.0,
      "step": 1135
    },
    {
      "entropy": 6.506372594833374,
      "epoch": 0.11314013497419612,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999978439671877,
      "loss": 6.301,
      "mean_token_accuracy": 0.11535706594586373,
      "num_tokens": 2473255.0,
      "step": 1140
    },
    {
      "entropy": 6.511087560653687,
      "epoch": 0.11363636363636363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999976860674309,
      "loss": 6.2554,
      "mean_token_accuracy": 0.10890059396624566,
      "num_tokens": 2484473.0,
      "step": 1145
    },
    {
      "entropy": 6.409585046768188,
      "epoch": 0.11413259229853116,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999975225882164,
      "loss": 6.2632,
      "mean_token_accuracy": 0.11080945432186126,
      "num_tokens": 2495652.0,
      "step": 1150
    },
    {
      "entropy": 6.538078832626343,
      "epoch": 0.1146288209606987,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499997353529548,
      "loss": 6.3189,
      "mean_token_accuracy": 0.10957048162817955,
      "num_tokens": 2507137.0,
      "step": 1155
    },
    {
      "entropy": 6.4585597038269045,
      "epoch": 0.11512504962286621,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999971788914303,
      "loss": 6.2365,
      "mean_token_accuracy": 0.11182398498058319,
      "num_tokens": 2517896.0,
      "step": 1160
    },
    {
      "entropy": 6.409658336639405,
      "epoch": 0.11562127828503374,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999969986738672,
      "loss": 6.2389,
      "mean_token_accuracy": 0.11219030469655991,
      "num_tokens": 2529976.0,
      "step": 1165
    },
    {
      "entropy": 6.589641475677491,
      "epoch": 0.11611750694720127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999968128768634,
      "loss": 6.2993,
      "mean_token_accuracy": 0.10817597731947899,
      "num_tokens": 2540872.0,
      "step": 1170
    },
    {
      "entropy": 6.4878946304321286,
      "epoch": 0.1166137356093688,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999966215004236,
      "loss": 6.3182,
      "mean_token_accuracy": 0.10902737304568291,
      "num_tokens": 2551178.0,
      "step": 1175
    },
    {
      "entropy": 6.421410703659058,
      "epoch": 0.11710996427153632,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999964245445522,
      "loss": 6.2131,
      "mean_token_accuracy": 0.11171767115592957,
      "num_tokens": 2561472.0,
      "step": 1180
    },
    {
      "entropy": 6.479715394973755,
      "epoch": 0.11760619293370385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999962220092545,
      "loss": 6.2229,
      "mean_token_accuracy": 0.10870179533958435,
      "num_tokens": 2572717.0,
      "step": 1185
    },
    {
      "entropy": 6.373128843307495,
      "epoch": 0.11810242159587138,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999960138945351,
      "loss": 6.1331,
      "mean_token_accuracy": 0.1215246357023716,
      "num_tokens": 2582991.0,
      "step": 1190
    },
    {
      "entropy": 6.388510036468506,
      "epoch": 0.11859865025803891,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999958002003996,
      "loss": 6.2167,
      "mean_token_accuracy": 0.11683873906731605,
      "num_tokens": 2594872.0,
      "step": 1195
    },
    {
      "entropy": 6.415997266769409,
      "epoch": 0.11909487892020643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999955809268529,
      "loss": 6.2267,
      "mean_token_accuracy": 0.11254609376192093,
      "num_tokens": 2605991.0,
      "step": 1200
    },
    {
      "entropy": 6.393628072738648,
      "epoch": 0.11959110758237396,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999953560739008,
      "loss": 6.247,
      "mean_token_accuracy": 0.11305257230997086,
      "num_tokens": 2617613.0,
      "step": 1205
    },
    {
      "entropy": 6.516189765930176,
      "epoch": 0.12008733624454149,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999951256415485,
      "loss": 6.3355,
      "mean_token_accuracy": 0.1064811147749424,
      "num_tokens": 2627724.0,
      "step": 1210
    },
    {
      "entropy": 6.452616119384766,
      "epoch": 0.12058356490670902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499994889629802,
      "loss": 6.2306,
      "mean_token_accuracy": 0.1097794696688652,
      "num_tokens": 2638944.0,
      "step": 1215
    },
    {
      "entropy": 6.364139032363892,
      "epoch": 0.12107979356887653,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499994648038667,
      "loss": 6.0846,
      "mean_token_accuracy": 0.12294464185833931,
      "num_tokens": 2649132.0,
      "step": 1220
    },
    {
      "entropy": 6.4025427341461185,
      "epoch": 0.12157602223104406,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999944008681496,
      "loss": 6.2173,
      "mean_token_accuracy": 0.10822719931602479,
      "num_tokens": 2660482.0,
      "step": 1225
    },
    {
      "entropy": 6.506667852401733,
      "epoch": 0.1220722508932116,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999941481182559,
      "loss": 6.2818,
      "mean_token_accuracy": 0.10780982449650764,
      "num_tokens": 2672953.0,
      "step": 1230
    },
    {
      "entropy": 6.3546521186828615,
      "epoch": 0.12256847955537911,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999938897889921,
      "loss": 6.1948,
      "mean_token_accuracy": 0.11872131600975991,
      "num_tokens": 2684958.0,
      "step": 1235
    },
    {
      "entropy": 6.4603499412536625,
      "epoch": 0.12306470821754664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999936258803647,
      "loss": 6.2033,
      "mean_token_accuracy": 0.1136890321969986,
      "num_tokens": 2695375.0,
      "step": 1240
    },
    {
      "entropy": 6.467173624038696,
      "epoch": 0.12356093687971417,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999933563923802,
      "loss": 6.373,
      "mean_token_accuracy": 0.10880970358848571,
      "num_tokens": 2707762.0,
      "step": 1245
    },
    {
      "entropy": 6.37615966796875,
      "epoch": 0.1240571655418817,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999930813250451,
      "loss": 6.1962,
      "mean_token_accuracy": 0.11519552767276764,
      "num_tokens": 2718237.0,
      "step": 1250
    },
    {
      "entropy": 6.407051467895508,
      "epoch": 0.12455339420404922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999928006783667,
      "loss": 6.1689,
      "mean_token_accuracy": 0.1186670869588852,
      "num_tokens": 2728746.0,
      "step": 1255
    },
    {
      "entropy": 6.4216368198394775,
      "epoch": 0.12504962286621676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999925144523515,
      "loss": 6.1714,
      "mean_token_accuracy": 0.11274236291646958,
      "num_tokens": 2739030.0,
      "step": 1260
    },
    {
      "entropy": 6.250958776473999,
      "epoch": 0.12554585152838427,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999922226470066,
      "loss": 6.1324,
      "mean_token_accuracy": 0.12048356831073762,
      "num_tokens": 2750233.0,
      "step": 1265
    },
    {
      "entropy": 6.451997470855713,
      "epoch": 0.1260420801905518,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999919252623397,
      "loss": 6.2138,
      "mean_token_accuracy": 0.10769852250814438,
      "num_tokens": 2761298.0,
      "step": 1270
    },
    {
      "entropy": 6.400608777999878,
      "epoch": 0.12653830885271933,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999916222983577,
      "loss": 6.2612,
      "mean_token_accuracy": 0.11032301038503647,
      "num_tokens": 2771493.0,
      "step": 1275
    },
    {
      "entropy": 6.354720497131348,
      "epoch": 0.12703453751488686,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999913137550682,
      "loss": 6.1673,
      "mean_token_accuracy": 0.11023365557193757,
      "num_tokens": 2782969.0,
      "step": 1280
    },
    {
      "entropy": 6.468644714355468,
      "epoch": 0.1275307661770544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999909996324791,
      "loss": 6.1993,
      "mean_token_accuracy": 0.11823943629860878,
      "num_tokens": 2793355.0,
      "step": 1285
    },
    {
      "entropy": 6.278990459442139,
      "epoch": 0.12802699483922192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499990679930598,
      "loss": 6.1114,
      "mean_token_accuracy": 0.11726746633648873,
      "num_tokens": 2804473.0,
      "step": 1290
    },
    {
      "entropy": 6.389729928970337,
      "epoch": 0.12852322350138945,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999903546494327,
      "loss": 6.1831,
      "mean_token_accuracy": 0.11785808727145194,
      "num_tokens": 2814899.0,
      "step": 1295
    },
    {
      "entropy": 6.435607671737671,
      "epoch": 0.12901945216355698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999900237889916,
      "loss": 6.2205,
      "mean_token_accuracy": 0.11157539635896682,
      "num_tokens": 2825592.0,
      "step": 1300
    },
    {
      "entropy": 6.323909330368042,
      "epoch": 0.12951568082572448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999896873492827,
      "loss": 6.2083,
      "mean_token_accuracy": 0.11462137624621391,
      "num_tokens": 2837444.0,
      "step": 1305
    },
    {
      "entropy": 6.432302522659302,
      "epoch": 0.130011909487892,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999893453303143,
      "loss": 6.2024,
      "mean_token_accuracy": 0.11139100939035415,
      "num_tokens": 2849256.0,
      "step": 1310
    },
    {
      "entropy": 6.387406969070435,
      "epoch": 0.13050813815005954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999889977320949,
      "loss": 6.2585,
      "mean_token_accuracy": 0.11350850909948348,
      "num_tokens": 2860606.0,
      "step": 1315
    },
    {
      "entropy": 6.396147966384888,
      "epoch": 0.13100436681222707,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999886445546332,
      "loss": 6.1914,
      "mean_token_accuracy": 0.12097390070557594,
      "num_tokens": 2871825.0,
      "step": 1320
    },
    {
      "entropy": 6.425122690200806,
      "epoch": 0.1315005954743946,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499988285797938,
      "loss": 6.2041,
      "mean_token_accuracy": 0.12094095051288604,
      "num_tokens": 2883408.0,
      "step": 1325
    },
    {
      "entropy": 6.322391223907471,
      "epoch": 0.13199682413656214,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499987921462018,
      "loss": 6.1787,
      "mean_token_accuracy": 0.12214997559785842,
      "num_tokens": 2895786.0,
      "step": 1330
    },
    {
      "entropy": 6.381843423843383,
      "epoch": 0.13249305279872967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999875515468825,
      "loss": 6.178,
      "mean_token_accuracy": 0.11670075803995132,
      "num_tokens": 2905273.0,
      "step": 1335
    },
    {
      "entropy": 6.471662425994873,
      "epoch": 0.13298928146089717,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999871760525403,
      "loss": 6.2786,
      "mean_token_accuracy": 0.10916746258735657,
      "num_tokens": 2917546.0,
      "step": 1340
    },
    {
      "entropy": 6.495834398269653,
      "epoch": 0.1334855101230647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999867949790011,
      "loss": 6.2294,
      "mean_token_accuracy": 0.11407745182514191,
      "num_tokens": 2928715.0,
      "step": 1345
    },
    {
      "entropy": 6.333612298965454,
      "epoch": 0.13398173878523223,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999864083262743,
      "loss": 6.17,
      "mean_token_accuracy": 0.11962553858757019,
      "num_tokens": 2940260.0,
      "step": 1350
    },
    {
      "entropy": 6.396699094772339,
      "epoch": 0.13447796744739976,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999860160943691,
      "loss": 6.1696,
      "mean_token_accuracy": 0.1181778758764267,
      "num_tokens": 2951543.0,
      "step": 1355
    },
    {
      "entropy": 6.358440589904785,
      "epoch": 0.1349741961095673,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999856182832958,
      "loss": 6.2075,
      "mean_token_accuracy": 0.11538006886839866,
      "num_tokens": 2962504.0,
      "step": 1360
    },
    {
      "entropy": 6.349350261688232,
      "epoch": 0.13547042477173482,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999852148930637,
      "loss": 6.1291,
      "mean_token_accuracy": 0.12085686475038529,
      "num_tokens": 2973784.0,
      "step": 1365
    },
    {
      "entropy": 6.387005567550659,
      "epoch": 0.13596665343390235,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999848059236832,
      "loss": 6.2569,
      "mean_token_accuracy": 0.10802673473954201,
      "num_tokens": 2985281.0,
      "step": 1370
    },
    {
      "entropy": 6.308599185943604,
      "epoch": 0.13646288209606988,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999843913751643,
      "loss": 6.1018,
      "mean_token_accuracy": 0.12057043686509132,
      "num_tokens": 2996090.0,
      "step": 1375
    },
    {
      "entropy": 6.370522308349609,
      "epoch": 0.13695911075823738,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999839712475174,
      "loss": 6.1337,
      "mean_token_accuracy": 0.11793153062462806,
      "num_tokens": 3006235.0,
      "step": 1380
    },
    {
      "entropy": 6.298479413986206,
      "epoch": 0.13745533942040492,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999835455407526,
      "loss": 6.1056,
      "mean_token_accuracy": 0.1146063707768917,
      "num_tokens": 3016481.0,
      "step": 1385
    },
    {
      "entropy": 6.411545610427856,
      "epoch": 0.13795156808257245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999831142548808,
      "loss": 6.162,
      "mean_token_accuracy": 0.12275284230709076,
      "num_tokens": 3026382.0,
      "step": 1390
    },
    {
      "entropy": 6.287210941314697,
      "epoch": 0.13844779674473998,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999826773899124,
      "loss": 6.1672,
      "mean_token_accuracy": 0.10925834402441978,
      "num_tokens": 3036637.0,
      "step": 1395
    },
    {
      "entropy": 6.356094312667847,
      "epoch": 0.1389440254069075,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999822349458586,
      "loss": 6.2199,
      "mean_token_accuracy": 0.11081714704632759,
      "num_tokens": 3047115.0,
      "step": 1400
    },
    {
      "entropy": 6.465851879119873,
      "epoch": 0.13944025406907504,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999817869227301,
      "loss": 6.2426,
      "mean_token_accuracy": 0.11556067392230034,
      "num_tokens": 3057499.0,
      "step": 1405
    },
    {
      "entropy": 6.309928035736084,
      "epoch": 0.13993648273124257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999813333205381,
      "loss": 6.1574,
      "mean_token_accuracy": 0.11418915838003159,
      "num_tokens": 3068503.0,
      "step": 1410
    },
    {
      "entropy": 6.382554817199707,
      "epoch": 0.14043271139341007,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999808741392937,
      "loss": 6.1352,
      "mean_token_accuracy": 0.1202838383615017,
      "num_tokens": 3079220.0,
      "step": 1415
    },
    {
      "entropy": 6.248684358596802,
      "epoch": 0.1409289400555776,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999804093790086,
      "loss": 6.129,
      "mean_token_accuracy": 0.12194435819983482,
      "num_tokens": 3091621.0,
      "step": 1420
    },
    {
      "entropy": 6.3224715232849125,
      "epoch": 0.14142516871774513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499979939039694,
      "loss": 6.1108,
      "mean_token_accuracy": 0.12159208208322525,
      "num_tokens": 3101378.0,
      "step": 1425
    },
    {
      "entropy": 6.2528337955474855,
      "epoch": 0.14192139737991266,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999794631213618,
      "loss": 6.1223,
      "mean_token_accuracy": 0.11627830192446709,
      "num_tokens": 3112996.0,
      "step": 1430
    },
    {
      "entropy": 6.345195579528808,
      "epoch": 0.1424176260420802,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999789816240236,
      "loss": 6.1233,
      "mean_token_accuracy": 0.11082257777452469,
      "num_tokens": 3124262.0,
      "step": 1435
    },
    {
      "entropy": 6.282847118377686,
      "epoch": 0.14291385470424772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999784945476915,
      "loss": 6.1378,
      "mean_token_accuracy": 0.11149395331740379,
      "num_tokens": 3135785.0,
      "step": 1440
    },
    {
      "entropy": 6.278212022781372,
      "epoch": 0.14341008336641525,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999780018923775,
      "loss": 6.0202,
      "mean_token_accuracy": 0.12852933034300804,
      "num_tokens": 3146841.0,
      "step": 1445
    },
    {
      "entropy": 6.276055955886841,
      "epoch": 0.14390631202858278,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999775036580939,
      "loss": 6.1744,
      "mean_token_accuracy": 0.11450461819767951,
      "num_tokens": 3158931.0,
      "step": 1450
    },
    {
      "entropy": 6.341690492630005,
      "epoch": 0.1444025406907503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999769998448529,
      "loss": 6.1153,
      "mean_token_accuracy": 0.11504201218485832,
      "num_tokens": 3169320.0,
      "step": 1455
    },
    {
      "entropy": 6.3100001335144045,
      "epoch": 0.14489876935291782,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999764904526672,
      "loss": 6.1483,
      "mean_token_accuracy": 0.11918737143278121,
      "num_tokens": 3179163.0,
      "step": 1460
    },
    {
      "entropy": 6.281153011322021,
      "epoch": 0.14539499801508535,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999759754815492,
      "loss": 6.083,
      "mean_token_accuracy": 0.11143830195069312,
      "num_tokens": 3187987.0,
      "step": 1465
    },
    {
      "entropy": 6.255558013916016,
      "epoch": 0.14589122667725288,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999754549315118,
      "loss": 6.0775,
      "mean_token_accuracy": 0.12257000356912613,
      "num_tokens": 3198052.0,
      "step": 1470
    },
    {
      "entropy": 6.222886371612549,
      "epoch": 0.1463874553394204,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499974928802568,
      "loss": 6.0643,
      "mean_token_accuracy": 0.11924131661653518,
      "num_tokens": 3209019.0,
      "step": 1475
    },
    {
      "entropy": 6.209675025939942,
      "epoch": 0.14688368400158794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999743970947306,
      "loss": 6.0273,
      "mean_token_accuracy": 0.12434762641787529,
      "num_tokens": 3219130.0,
      "step": 1480
    },
    {
      "entropy": 6.21720700263977,
      "epoch": 0.14737991266375547,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499973859808013,
      "loss": 6.0421,
      "mean_token_accuracy": 0.12010939195752143,
      "num_tokens": 3230882.0,
      "step": 1485
    },
    {
      "entropy": 6.222798299789429,
      "epoch": 0.14787614132592297,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999733169424284,
      "loss": 6.1729,
      "mean_token_accuracy": 0.11634461581707001,
      "num_tokens": 3241752.0,
      "step": 1490
    },
    {
      "entropy": 6.488711023330689,
      "epoch": 0.1483723699880905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999727684979904,
      "loss": 6.1832,
      "mean_token_accuracy": 0.11442616507411003,
      "num_tokens": 3252291.0,
      "step": 1495
    },
    {
      "entropy": 6.317147922515869,
      "epoch": 0.14886859865025803,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999722144747125,
      "loss": 6.0964,
      "mean_token_accuracy": 0.11958790943026543,
      "num_tokens": 3263525.0,
      "step": 1500
    },
    {
      "entropy": 6.26204400062561,
      "epoch": 0.14936482731242556,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999716548726085,
      "loss": 6.0651,
      "mean_token_accuracy": 0.12191420197486877,
      "num_tokens": 3274164.0,
      "step": 1505
    },
    {
      "entropy": 6.1559864521026615,
      "epoch": 0.1498610559745931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999710896916921,
      "loss": 6.0561,
      "mean_token_accuracy": 0.12110427170991897,
      "num_tokens": 3286244.0,
      "step": 1510
    },
    {
      "entropy": 6.305137777328492,
      "epoch": 0.15035728463676062,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999705189319774,
      "loss": 6.1474,
      "mean_token_accuracy": 0.11220626085996628,
      "num_tokens": 3296569.0,
      "step": 1515
    },
    {
      "entropy": 6.250655698776245,
      "epoch": 0.15085351329892815,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999699425934788,
      "loss": 6.1675,
      "mean_token_accuracy": 0.12036664485931396,
      "num_tokens": 3307259.0,
      "step": 1520
    },
    {
      "entropy": 6.2826783657073975,
      "epoch": 0.15134974196109569,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999693606762103,
      "loss": 5.9762,
      "mean_token_accuracy": 0.12630216851830484,
      "num_tokens": 3317473.0,
      "step": 1525
    },
    {
      "entropy": 6.221875429153442,
      "epoch": 0.1518459706232632,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999687731801865,
      "loss": 6.0801,
      "mean_token_accuracy": 0.11735879704356193,
      "num_tokens": 3328783.0,
      "step": 1530
    },
    {
      "entropy": 6.221706104278565,
      "epoch": 0.15234219928543072,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999681801054217,
      "loss": 6.1161,
      "mean_token_accuracy": 0.11810076832771302,
      "num_tokens": 3338308.0,
      "step": 1535
    },
    {
      "entropy": 6.273039722442627,
      "epoch": 0.15283842794759825,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499967581451931,
      "loss": 6.0367,
      "mean_token_accuracy": 0.12085681930184364,
      "num_tokens": 3348802.0,
      "step": 1540
    },
    {
      "entropy": 6.250692558288574,
      "epoch": 0.15333465660976578,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499966977219729,
      "loss": 6.0505,
      "mean_token_accuracy": 0.122286207228899,
      "num_tokens": 3358781.0,
      "step": 1545
    },
    {
      "entropy": 6.182136583328247,
      "epoch": 0.1538308852719333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999663674088306,
      "loss": 6.004,
      "mean_token_accuracy": 0.12099720686674117,
      "num_tokens": 3369360.0,
      "step": 1550
    },
    {
      "entropy": 6.30761604309082,
      "epoch": 0.15432711393410084,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999657520192511,
      "loss": 6.1196,
      "mean_token_accuracy": 0.11720000579953194,
      "num_tokens": 3380111.0,
      "step": 1555
    },
    {
      "entropy": 6.182616472244263,
      "epoch": 0.15482334259626837,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999651310510058,
      "loss": 6.0476,
      "mean_token_accuracy": 0.12010963633656502,
      "num_tokens": 3391932.0,
      "step": 1560
    },
    {
      "entropy": 6.286138010025025,
      "epoch": 0.15531957125843587,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999645045041099,
      "loss": 6.0978,
      "mean_token_accuracy": 0.11951644867658615,
      "num_tokens": 3402192.0,
      "step": 1565
    },
    {
      "entropy": 6.25222749710083,
      "epoch": 0.1558157999206034,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999638723785791,
      "loss": 5.9897,
      "mean_token_accuracy": 0.12218942567706108,
      "num_tokens": 3412382.0,
      "step": 1570
    },
    {
      "entropy": 6.230171489715576,
      "epoch": 0.15631202858277093,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499963234674429,
      "loss": 6.0516,
      "mean_token_accuracy": 0.11803107783198356,
      "num_tokens": 3422471.0,
      "step": 1575
    },
    {
      "entropy": 6.2233234405517575,
      "epoch": 0.15680825724493846,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999625913916754,
      "loss": 6.0301,
      "mean_token_accuracy": 0.12271839007735252,
      "num_tokens": 3432348.0,
      "step": 1580
    },
    {
      "entropy": 6.185507822036743,
      "epoch": 0.157304485907106,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999619425303343,
      "loss": 6.0307,
      "mean_token_accuracy": 0.11823347955942154,
      "num_tokens": 3442549.0,
      "step": 1585
    },
    {
      "entropy": 6.245360565185547,
      "epoch": 0.15780071456927353,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999612880904219,
      "loss": 6.1235,
      "mean_token_accuracy": 0.11440040320158004,
      "num_tokens": 3452317.0,
      "step": 1590
    },
    {
      "entropy": 6.224632358551025,
      "epoch": 0.15829694323144106,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999606280719542,
      "loss": 6.0532,
      "mean_token_accuracy": 0.11879203096032143,
      "num_tokens": 3463432.0,
      "step": 1595
    },
    {
      "entropy": 6.308253622055053,
      "epoch": 0.1587931718936086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999599624749475,
      "loss": 6.1023,
      "mean_token_accuracy": 0.11679881960153579,
      "num_tokens": 3473800.0,
      "step": 1600
    },
    {
      "entropy": 6.227562952041626,
      "epoch": 0.1592894005557761,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999592912994187,
      "loss": 6.0823,
      "mean_token_accuracy": 0.11684635356068611,
      "num_tokens": 3484803.0,
      "step": 1605
    },
    {
      "entropy": 6.235965061187744,
      "epoch": 0.15978562921794362,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999586145453842,
      "loss": 6.0284,
      "mean_token_accuracy": 0.12258547395467759,
      "num_tokens": 3494687.0,
      "step": 1610
    },
    {
      "entropy": 6.244058561325073,
      "epoch": 0.16028185788011115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999579322128608,
      "loss": 6.0741,
      "mean_token_accuracy": 0.11799983456730842,
      "num_tokens": 3505586.0,
      "step": 1615
    },
    {
      "entropy": 6.10240535736084,
      "epoch": 0.16077808654227868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999572443018654,
      "loss": 5.9235,
      "mean_token_accuracy": 0.13439927846193314,
      "num_tokens": 3516702.0,
      "step": 1620
    },
    {
      "entropy": 6.232553434371948,
      "epoch": 0.1612743152044462,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999565508124151,
      "loss": 6.0416,
      "mean_token_accuracy": 0.12300089374184608,
      "num_tokens": 3527240.0,
      "step": 1625
    },
    {
      "entropy": 6.2715003490448,
      "epoch": 0.16177054386661374,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499955851744527,
      "loss": 6.1539,
      "mean_token_accuracy": 0.11600246727466583,
      "num_tokens": 3539368.0,
      "step": 1630
    },
    {
      "entropy": 6.200672674179077,
      "epoch": 0.16226677252878127,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999551470982185,
      "loss": 6.0307,
      "mean_token_accuracy": 0.12346499860286712,
      "num_tokens": 3551128.0,
      "step": 1635
    },
    {
      "entropy": 6.121882009506225,
      "epoch": 0.1627630011909488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999544368735073,
      "loss": 6.0322,
      "mean_token_accuracy": 0.12636793553829193,
      "num_tokens": 3562385.0,
      "step": 1640
    },
    {
      "entropy": 6.254643726348877,
      "epoch": 0.1632592298531163,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999537210704106,
      "loss": 6.0262,
      "mean_token_accuracy": 0.1224179670214653,
      "num_tokens": 3572737.0,
      "step": 1645
    },
    {
      "entropy": 6.137145900726319,
      "epoch": 0.16375545851528384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999529996889463,
      "loss": 6.0462,
      "mean_token_accuracy": 0.11289570853114128,
      "num_tokens": 3583872.0,
      "step": 1650
    },
    {
      "entropy": 6.286539888381958,
      "epoch": 0.16425168717745137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999522727291325,
      "loss": 6.0236,
      "mean_token_accuracy": 0.11809901297092437,
      "num_tokens": 3595012.0,
      "step": 1655
    },
    {
      "entropy": 6.250521469116211,
      "epoch": 0.1647479158396189,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499951540190987,
      "loss": 6.14,
      "mean_token_accuracy": 0.11284013092517853,
      "num_tokens": 3605733.0,
      "step": 1660
    },
    {
      "entropy": 6.2044517517089846,
      "epoch": 0.16524414450178643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999508020745281,
      "loss": 6.0625,
      "mean_token_accuracy": 0.12692697569727898,
      "num_tokens": 3615997.0,
      "step": 1665
    },
    {
      "entropy": 6.166298866271973,
      "epoch": 0.16574037316395396,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999500583797739,
      "loss": 6.0486,
      "mean_token_accuracy": 0.12888892963528634,
      "num_tokens": 3628742.0,
      "step": 1670
    },
    {
      "entropy": 6.150118827819824,
      "epoch": 0.1662366018261215,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999493091067431,
      "loss": 6.0301,
      "mean_token_accuracy": 0.1232452742755413,
      "num_tokens": 3639702.0,
      "step": 1675
    },
    {
      "entropy": 6.249062776565552,
      "epoch": 0.166732830488289,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999485542554541,
      "loss": 6.0093,
      "mean_token_accuracy": 0.12544036880135537,
      "num_tokens": 3650414.0,
      "step": 1680
    },
    {
      "entropy": 6.117237901687622,
      "epoch": 0.16722905915045652,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999477938259256,
      "loss": 5.9927,
      "mean_token_accuracy": 0.11784988194704056,
      "num_tokens": 3660739.0,
      "step": 1685
    },
    {
      "entropy": 6.2560646533966064,
      "epoch": 0.16772528781262405,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999470278181766,
      "loss": 6.075,
      "mean_token_accuracy": 0.12103780433535576,
      "num_tokens": 3673109.0,
      "step": 1690
    },
    {
      "entropy": 6.2072841167449955,
      "epoch": 0.16822151647479158,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999462562322261,
      "loss": 6.0392,
      "mean_token_accuracy": 0.123382568359375,
      "num_tokens": 3683168.0,
      "step": 1695
    },
    {
      "entropy": 6.17884316444397,
      "epoch": 0.1687177451369591,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499945479068093,
      "loss": 6.0407,
      "mean_token_accuracy": 0.12351762875914574,
      "num_tokens": 3694309.0,
      "step": 1700
    },
    {
      "entropy": 6.194273328781128,
      "epoch": 0.16921397379912664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999446963257969,
      "loss": 6.1322,
      "mean_token_accuracy": 0.11986947655677796,
      "num_tokens": 3706165.0,
      "step": 1705
    },
    {
      "entropy": 6.218581247329712,
      "epoch": 0.16971020246129417,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999439080053571,
      "loss": 6.1305,
      "mean_token_accuracy": 0.12079597637057304,
      "num_tokens": 3717818.0,
      "step": 1710
    },
    {
      "entropy": 6.240843820571899,
      "epoch": 0.1702064311234617,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999431141067928,
      "loss": 5.9971,
      "mean_token_accuracy": 0.1241668090224266,
      "num_tokens": 3727837.0,
      "step": 1715
    },
    {
      "entropy": 6.140651416778565,
      "epoch": 0.1707026597856292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999423146301242,
      "loss": 5.9883,
      "mean_token_accuracy": 0.11822554692626,
      "num_tokens": 3738827.0,
      "step": 1720
    },
    {
      "entropy": 6.252899122238159,
      "epoch": 0.17119888844779674,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999415095753709,
      "loss": 6.0,
      "mean_token_accuracy": 0.12453790381550789,
      "num_tokens": 3749859.0,
      "step": 1725
    },
    {
      "entropy": 6.116316032409668,
      "epoch": 0.17169511710996427,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999406989425527,
      "loss": 5.999,
      "mean_token_accuracy": 0.1228027805685997,
      "num_tokens": 3760999.0,
      "step": 1730
    },
    {
      "entropy": 6.207907724380493,
      "epoch": 0.1721913457721318,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999398827316901,
      "loss": 6.025,
      "mean_token_accuracy": 0.12453913390636444,
      "num_tokens": 3771541.0,
      "step": 1735
    },
    {
      "entropy": 6.126678371429444,
      "epoch": 0.17268757443429933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499939060942803,
      "loss": 5.983,
      "mean_token_accuracy": 0.1210265539586544,
      "num_tokens": 3781503.0,
      "step": 1740
    },
    {
      "entropy": 6.190077447891236,
      "epoch": 0.17318380309646686,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999382335759119,
      "loss": 6.0486,
      "mean_token_accuracy": 0.11707601249217987,
      "num_tokens": 3791278.0,
      "step": 1745
    },
    {
      "entropy": 6.263814401626587,
      "epoch": 0.1736800317586344,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999374006310372,
      "loss": 6.0787,
      "mean_token_accuracy": 0.12131959795951844,
      "num_tokens": 3802995.0,
      "step": 1750
    },
    {
      "entropy": 6.144543790817261,
      "epoch": 0.1741762604208019,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999365621081999,
      "loss": 6.0164,
      "mean_token_accuracy": 0.12310748770833016,
      "num_tokens": 3813912.0,
      "step": 1755
    },
    {
      "entropy": 6.259069395065308,
      "epoch": 0.17467248908296942,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999357180074204,
      "loss": 5.9649,
      "mean_token_accuracy": 0.1212263897061348,
      "num_tokens": 3823513.0,
      "step": 1760
    },
    {
      "entropy": 6.128121471405029,
      "epoch": 0.17516871774513695,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999348683287199,
      "loss": 6.0029,
      "mean_token_accuracy": 0.1258741870522499,
      "num_tokens": 3835602.0,
      "step": 1765
    },
    {
      "entropy": 6.148974323272705,
      "epoch": 0.17566494640730448,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999340130721192,
      "loss": 6.0626,
      "mean_token_accuracy": 0.11835796162486076,
      "num_tokens": 3847031.0,
      "step": 1770
    },
    {
      "entropy": 6.139218711853028,
      "epoch": 0.17616117506947201,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999331522376396,
      "loss": 5.9138,
      "mean_token_accuracy": 0.1234263926744461,
      "num_tokens": 3857714.0,
      "step": 1775
    },
    {
      "entropy": 6.157423210144043,
      "epoch": 0.17665740373163955,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999322858253027,
      "loss": 6.0043,
      "mean_token_accuracy": 0.1264146938920021,
      "num_tokens": 3868818.0,
      "step": 1780
    },
    {
      "entropy": 6.190987920761108,
      "epoch": 0.17715363239380708,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999314138351297,
      "loss": 6.0497,
      "mean_token_accuracy": 0.12253930419683456,
      "num_tokens": 3878802.0,
      "step": 1785
    },
    {
      "entropy": 6.232629203796387,
      "epoch": 0.1776498610559746,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999305362671424,
      "loss": 6.0532,
      "mean_token_accuracy": 0.12156687378883362,
      "num_tokens": 3890045.0,
      "step": 1790
    },
    {
      "entropy": 6.123168706893921,
      "epoch": 0.1781460897181421,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999296531213624,
      "loss": 5.9338,
      "mean_token_accuracy": 0.12736912518739701,
      "num_tokens": 3901429.0,
      "step": 1795
    },
    {
      "entropy": 6.106752252578735,
      "epoch": 0.17864231838030964,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999287643978117,
      "loss": 5.9172,
      "mean_token_accuracy": 0.12626976445317267,
      "num_tokens": 3912201.0,
      "step": 1800
    },
    {
      "entropy": 6.071602916717529,
      "epoch": 0.17913854704247717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999278700965124,
      "loss": 5.9402,
      "mean_token_accuracy": 0.12588722854852677,
      "num_tokens": 3922338.0,
      "step": 1805
    },
    {
      "entropy": 6.153251314163208,
      "epoch": 0.1796347757046447,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999269702174866,
      "loss": 5.9837,
      "mean_token_accuracy": 0.12280823290348053,
      "num_tokens": 3934469.0,
      "step": 1810
    },
    {
      "entropy": 6.121962928771973,
      "epoch": 0.18013100436681223,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999260647607565,
      "loss": 5.9517,
      "mean_token_accuracy": 0.12701089158654214,
      "num_tokens": 3946136.0,
      "step": 1815
    },
    {
      "entropy": 6.157146072387695,
      "epoch": 0.18062723302897976,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999251537263448,
      "loss": 6.0053,
      "mean_token_accuracy": 0.12121656090021134,
      "num_tokens": 3956976.0,
      "step": 1820
    },
    {
      "entropy": 6.189162302017212,
      "epoch": 0.1811234616911473,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999242371142739,
      "loss": 6.0184,
      "mean_token_accuracy": 0.12531510293483733,
      "num_tokens": 3967929.0,
      "step": 1825
    },
    {
      "entropy": 6.028120803833008,
      "epoch": 0.1816196903533148,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999233149245665,
      "loss": 5.9661,
      "mean_token_accuracy": 0.13040608540177345,
      "num_tokens": 3979585.0,
      "step": 1830
    },
    {
      "entropy": 6.194211387634278,
      "epoch": 0.18211591901548232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999223871572456,
      "loss": 5.9483,
      "mean_token_accuracy": 0.12807398438453674,
      "num_tokens": 3990097.0,
      "step": 1835
    },
    {
      "entropy": 6.195305252075196,
      "epoch": 0.18261214767764986,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999214538123342,
      "loss": 6.0159,
      "mean_token_accuracy": 0.11925126612186432,
      "num_tokens": 3999864.0,
      "step": 1840
    },
    {
      "entropy": 6.140563869476319,
      "epoch": 0.18310837633981739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999205148898554,
      "loss": 6.0352,
      "mean_token_accuracy": 0.12029140070080757,
      "num_tokens": 4012986.0,
      "step": 1845
    },
    {
      "entropy": 6.133894157409668,
      "epoch": 0.18360460500198492,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999195703898325,
      "loss": 6.014,
      "mean_token_accuracy": 0.12564092352986336,
      "num_tokens": 4024231.0,
      "step": 1850
    },
    {
      "entropy": 6.100856161117553,
      "epoch": 0.18410083366415245,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999186203122889,
      "loss": 6.014,
      "mean_token_accuracy": 0.1193986639380455,
      "num_tokens": 4035709.0,
      "step": 1855
    },
    {
      "entropy": 6.097147274017334,
      "epoch": 0.18459706232631998,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999176646572482,
      "loss": 5.8987,
      "mean_token_accuracy": 0.12348575592041015,
      "num_tokens": 4046504.0,
      "step": 1860
    },
    {
      "entropy": 6.13851490020752,
      "epoch": 0.1850932909884875,
      "grad_norm": 1.0,
      "learning_rate": 0.000499916703424734,
      "loss": 5.9443,
      "mean_token_accuracy": 0.11826654374599457,
      "num_tokens": 4056864.0,
      "step": 1865
    },
    {
      "entropy": 6.073440790176392,
      "epoch": 0.185589519650655,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999157366147702,
      "loss": 5.962,
      "mean_token_accuracy": 0.12339324504137039,
      "num_tokens": 4067682.0,
      "step": 1870
    },
    {
      "entropy": 6.170620965957641,
      "epoch": 0.18608574831282254,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999147642273809,
      "loss": 5.9651,
      "mean_token_accuracy": 0.12521442025899887,
      "num_tokens": 4078623.0,
      "step": 1875
    },
    {
      "entropy": 6.154106092453003,
      "epoch": 0.18658197697499007,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049991378626259,
      "loss": 6.0492,
      "mean_token_accuracy": 0.12163211032748222,
      "num_tokens": 4090479.0,
      "step": 1880
    },
    {
      "entropy": 6.190439748764038,
      "epoch": 0.1870782056371576,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999128027204219,
      "loss": 5.9575,
      "mean_token_accuracy": 0.1205236978828907,
      "num_tokens": 4100379.0,
      "step": 1885
    },
    {
      "entropy": 6.100738382339477,
      "epoch": 0.18757443429932513,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499911813600901,
      "loss": 6.102,
      "mean_token_accuracy": 0.11705014109611511,
      "num_tokens": 4111738.0,
      "step": 1890
    },
    {
      "entropy": 6.248137140274048,
      "epoch": 0.18807066296149266,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999108189040516,
      "loss": 6.0623,
      "mean_token_accuracy": 0.11439691036939621,
      "num_tokens": 4123602.0,
      "step": 1895
    },
    {
      "entropy": 6.1028767108917235,
      "epoch": 0.1885668916236602,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999098186298986,
      "loss": 6.007,
      "mean_token_accuracy": 0.12231771126389504,
      "num_tokens": 4133558.0,
      "step": 1900
    },
    {
      "entropy": 6.093136978149414,
      "epoch": 0.1890631202858277,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999088127784667,
      "loss": 5.9219,
      "mean_token_accuracy": 0.12927619069814683,
      "num_tokens": 4143901.0,
      "step": 1905
    },
    {
      "entropy": 6.120997333526612,
      "epoch": 0.18955934894799523,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499907801349781,
      "loss": 5.9625,
      "mean_token_accuracy": 0.11881752163171769,
      "num_tokens": 4154180.0,
      "step": 1910
    },
    {
      "entropy": 6.1055737972259525,
      "epoch": 0.19005557761016276,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999067843438664,
      "loss": 5.9002,
      "mean_token_accuracy": 0.12527662962675096,
      "num_tokens": 4163540.0,
      "step": 1915
    },
    {
      "entropy": 6.11438364982605,
      "epoch": 0.1905518062723303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999057617607482,
      "loss": 5.9836,
      "mean_token_accuracy": 0.12190114930272103,
      "num_tokens": 4173723.0,
      "step": 1920
    },
    {
      "entropy": 6.151767587661743,
      "epoch": 0.19104803493449782,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999047336004517,
      "loss": 5.9828,
      "mean_token_accuracy": 0.12671934813261032,
      "num_tokens": 4184771.0,
      "step": 1925
    },
    {
      "entropy": 6.133954286575317,
      "epoch": 0.19154426359666535,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999036998630024,
      "loss": 5.9683,
      "mean_token_accuracy": 0.12368112206459045,
      "num_tokens": 4195791.0,
      "step": 1930
    },
    {
      "entropy": 6.083200073242187,
      "epoch": 0.19204049225883288,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499902660548426,
      "loss": 6.0029,
      "mean_token_accuracy": 0.12152576521039009,
      "num_tokens": 4207393.0,
      "step": 1935
    },
    {
      "entropy": 6.179332113265991,
      "epoch": 0.1925367209210004,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999016156567482,
      "loss": 5.9361,
      "mean_token_accuracy": 0.1270752251148224,
      "num_tokens": 4218278.0,
      "step": 1940
    },
    {
      "entropy": 6.1392213821411135,
      "epoch": 0.1930329495831679,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499900565187995,
      "loss": 6.0965,
      "mean_token_accuracy": 0.11639922857284546,
      "num_tokens": 4229448.0,
      "step": 1945
    },
    {
      "entropy": 6.2056193351745605,
      "epoch": 0.19352917824533544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998995091421924,
      "loss": 6.0121,
      "mean_token_accuracy": 0.12356942817568779,
      "num_tokens": 4240750.0,
      "step": 1950
    },
    {
      "entropy": 6.125873804092407,
      "epoch": 0.19402540690750297,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998984475193665,
      "loss": 5.9919,
      "mean_token_accuracy": 0.1261401057243347,
      "num_tokens": 4252086.0,
      "step": 1955
    },
    {
      "entropy": 6.123996639251709,
      "epoch": 0.1945216355696705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998973803195439,
      "loss": 5.9428,
      "mean_token_accuracy": 0.1254188634455204,
      "num_tokens": 4262756.0,
      "step": 1960
    },
    {
      "entropy": 6.069361400604248,
      "epoch": 0.19501786423183803,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998963075427508,
      "loss": 5.999,
      "mean_token_accuracy": 0.12879851460456848,
      "num_tokens": 4272965.0,
      "step": 1965
    },
    {
      "entropy": 6.210020208358765,
      "epoch": 0.19551409289400556,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998952291890138,
      "loss": 6.052,
      "mean_token_accuracy": 0.12350022047758102,
      "num_tokens": 4284465.0,
      "step": 1970
    },
    {
      "entropy": 6.133313083648682,
      "epoch": 0.1960103215561731,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998941452583599,
      "loss": 5.9768,
      "mean_token_accuracy": 0.12168458700180054,
      "num_tokens": 4295266.0,
      "step": 1975
    },
    {
      "entropy": 6.1962378978729244,
      "epoch": 0.1965065502183406,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998930557508156,
      "loss": 6.017,
      "mean_token_accuracy": 0.11654685959219932,
      "num_tokens": 4306477.0,
      "step": 1980
    },
    {
      "entropy": 6.12655668258667,
      "epoch": 0.19700277888050813,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004998919606664082,
      "loss": 6.0534,
      "mean_token_accuracy": 0.11759311258792877,
      "num_tokens": 4319150.0,
      "step": 1985
    },
    {
      "entropy": 6.129172420501709,
      "epoch": 0.19749900754267566,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998908600051648,
      "loss": 6.0234,
      "mean_token_accuracy": 0.12395478188991546,
      "num_tokens": 4330679.0,
      "step": 1990
    },
    {
      "entropy": 6.184890651702881,
      "epoch": 0.1979952362048432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998897537671126,
      "loss": 5.9498,
      "mean_token_accuracy": 0.1217629611492157,
      "num_tokens": 4340597.0,
      "step": 1995
    },
    {
      "entropy": 6.026102209091187,
      "epoch": 0.19849146486701072,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998886419522792,
      "loss": 5.951,
      "mean_token_accuracy": 0.12472691535949706,
      "num_tokens": 4351814.0,
      "step": 2000
    },
    {
      "entropy": 6.186213159561158,
      "epoch": 0.19898769352917825,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004998875245606919,
      "loss": 6.0405,
      "mean_token_accuracy": 0.1263307310640812,
      "num_tokens": 4363930.0,
      "step": 2005
    },
    {
      "entropy": 6.186520051956177,
      "epoch": 0.19948392219134578,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998864015923788,
      "loss": 5.9422,
      "mean_token_accuracy": 0.12555222287774087,
      "num_tokens": 4376127.0,
      "step": 2010
    },
    {
      "entropy": 6.03753776550293,
      "epoch": 0.1999801508535133,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998852730473674,
      "loss": 5.9369,
      "mean_token_accuracy": 0.12351071462035179,
      "num_tokens": 4386095.0,
      "step": 2015
    },
    {
      "entropy": 6.076215410232544,
      "epoch": 0.2004763795156808,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998841389256859,
      "loss": 5.8703,
      "mean_token_accuracy": 0.12462282553315163,
      "num_tokens": 4397441.0,
      "step": 2020
    },
    {
      "entropy": 6.071394205093384,
      "epoch": 0.20097260817784834,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998829992273622,
      "loss": 5.9734,
      "mean_token_accuracy": 0.12160781174898147,
      "num_tokens": 4408295.0,
      "step": 2025
    },
    {
      "entropy": 6.043331718444824,
      "epoch": 0.20146883684001587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998818539524248,
      "loss": 5.9636,
      "mean_token_accuracy": 0.12636535987257957,
      "num_tokens": 4419307.0,
      "step": 2030
    },
    {
      "entropy": 6.137074136734009,
      "epoch": 0.2019650655021834,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499880703100902,
      "loss": 5.9539,
      "mean_token_accuracy": 0.12338273897767067,
      "num_tokens": 4431908.0,
      "step": 2035
    },
    {
      "entropy": 6.1705132007598875,
      "epoch": 0.20246129416435094,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998795466728223,
      "loss": 5.9858,
      "mean_token_accuracy": 0.12354729771614074,
      "num_tokens": 4443417.0,
      "step": 2040
    },
    {
      "entropy": 6.078726291656494,
      "epoch": 0.20295752282651847,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998783846682144,
      "loss": 6.0307,
      "mean_token_accuracy": 0.1215967983007431,
      "num_tokens": 4453224.0,
      "step": 2045
    },
    {
      "entropy": 6.076732397079468,
      "epoch": 0.203453751488686,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998772170871072,
      "loss": 5.8468,
      "mean_token_accuracy": 0.12905316129326821,
      "num_tokens": 4463231.0,
      "step": 2050
    },
    {
      "entropy": 6.058457279205323,
      "epoch": 0.20394998015085353,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998760439295295,
      "loss": 5.9344,
      "mean_token_accuracy": 0.12181595787405967,
      "num_tokens": 4474397.0,
      "step": 2055
    },
    {
      "entropy": 6.075131225585937,
      "epoch": 0.20444620881302103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998748651955105,
      "loss": 5.9376,
      "mean_token_accuracy": 0.12992871999740602,
      "num_tokens": 4485623.0,
      "step": 2060
    },
    {
      "entropy": 6.1033796787261965,
      "epoch": 0.20494243747518856,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998736808850795,
      "loss": 5.9249,
      "mean_token_accuracy": 0.12822652086615563,
      "num_tokens": 4495247.0,
      "step": 2065
    },
    {
      "entropy": 6.0438454151153564,
      "epoch": 0.2054386661373561,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998724909982656,
      "loss": 5.9906,
      "mean_token_accuracy": 0.12130568027496338,
      "num_tokens": 4506490.0,
      "step": 2070
    },
    {
      "entropy": 6.184813690185547,
      "epoch": 0.20593489479952362,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998712955350986,
      "loss": 6.0001,
      "mean_token_accuracy": 0.12409835308790207,
      "num_tokens": 4517511.0,
      "step": 2075
    },
    {
      "entropy": 6.033582639694214,
      "epoch": 0.20643112346169115,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998700944956081,
      "loss": 5.9458,
      "mean_token_accuracy": 0.13070451766252517,
      "num_tokens": 4528655.0,
      "step": 2080
    },
    {
      "entropy": 6.07148003578186,
      "epoch": 0.20692735212385868,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998688878798237,
      "loss": 5.9863,
      "mean_token_accuracy": 0.12557874843478203,
      "num_tokens": 4539369.0,
      "step": 2085
    },
    {
      "entropy": 6.109499120712281,
      "epoch": 0.2074235807860262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998676756877753,
      "loss": 5.8801,
      "mean_token_accuracy": 0.12313915863633156,
      "num_tokens": 4548747.0,
      "step": 2090
    },
    {
      "entropy": 5.956798696517945,
      "epoch": 0.20791980944819372,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998664579194934,
      "loss": 5.8747,
      "mean_token_accuracy": 0.12651906609535218,
      "num_tokens": 4561283.0,
      "step": 2095
    },
    {
      "entropy": 6.101844596862793,
      "epoch": 0.20841603811036125,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998652345750077,
      "loss": 6.0087,
      "mean_token_accuracy": 0.12195251435041428,
      "num_tokens": 4571929.0,
      "step": 2100
    },
    {
      "entropy": 6.0962621688842775,
      "epoch": 0.20891226677252878,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998640056543487,
      "loss": 5.9936,
      "mean_token_accuracy": 0.12055393010377884,
      "num_tokens": 4583105.0,
      "step": 2105
    },
    {
      "entropy": 6.128101778030396,
      "epoch": 0.2094084954346963,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998627711575469,
      "loss": 6.0601,
      "mean_token_accuracy": 0.11935122683644295,
      "num_tokens": 4595360.0,
      "step": 2110
    },
    {
      "entropy": 6.121275186538696,
      "epoch": 0.20990472409686384,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499861531084633,
      "loss": 6.0214,
      "mean_token_accuracy": 0.12107700780034066,
      "num_tokens": 4607621.0,
      "step": 2115
    },
    {
      "entropy": 6.0780713081359865,
      "epoch": 0.21040095275903137,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998602854356376,
      "loss": 5.8639,
      "mean_token_accuracy": 0.1276746354997158,
      "num_tokens": 4618286.0,
      "step": 2120
    },
    {
      "entropy": 6.060158061981201,
      "epoch": 0.2108971814211989,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998590342105917,
      "loss": 5.9454,
      "mean_token_accuracy": 0.1269027128815651,
      "num_tokens": 4628345.0,
      "step": 2125
    },
    {
      "entropy": 6.151575517654419,
      "epoch": 0.21139341008336643,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998577774095262,
      "loss": 5.9933,
      "mean_token_accuracy": 0.12428949922323226,
      "num_tokens": 4639074.0,
      "step": 2130
    },
    {
      "entropy": 6.060426807403564,
      "epoch": 0.21188963874553393,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998565150324724,
      "loss": 6.0473,
      "mean_token_accuracy": 0.12472090199589729,
      "num_tokens": 4652698.0,
      "step": 2135
    },
    {
      "entropy": 6.079493188858033,
      "epoch": 0.21238586740770146,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998552470794616,
      "loss": 5.8508,
      "mean_token_accuracy": 0.13163320422172547,
      "num_tokens": 4663924.0,
      "step": 2140
    },
    {
      "entropy": 6.1048369884490965,
      "epoch": 0.212882096069869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998539735505251,
      "loss": 5.965,
      "mean_token_accuracy": 0.1236407183110714,
      "num_tokens": 4673821.0,
      "step": 2145
    },
    {
      "entropy": 6.084499216079712,
      "epoch": 0.21337832473203652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998526944456946,
      "loss": 5.9017,
      "mean_token_accuracy": 0.12818676382303237,
      "num_tokens": 4686646.0,
      "step": 2150
    },
    {
      "entropy": 6.096099758148194,
      "epoch": 0.21387455339420405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998514097650018,
      "loss": 5.9023,
      "mean_token_accuracy": 0.1319211333990097,
      "num_tokens": 4698118.0,
      "step": 2155
    },
    {
      "entropy": 6.026635837554932,
      "epoch": 0.21437078205637158,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998501195084786,
      "loss": 5.9088,
      "mean_token_accuracy": 0.12052568271756173,
      "num_tokens": 4708869.0,
      "step": 2160
    },
    {
      "entropy": 6.034014320373535,
      "epoch": 0.21486701071853911,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998488236761569,
      "loss": 5.9338,
      "mean_token_accuracy": 0.1213805504143238,
      "num_tokens": 4720172.0,
      "step": 2165
    },
    {
      "entropy": 6.08177638053894,
      "epoch": 0.21536323938070662,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998475222680688,
      "loss": 5.9905,
      "mean_token_accuracy": 0.12283061891794204,
      "num_tokens": 4731886.0,
      "step": 2170
    },
    {
      "entropy": 6.061129188537597,
      "epoch": 0.21585946804287415,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998462152842468,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1260004922747612,
      "num_tokens": 4742381.0,
      "step": 2175
    },
    {
      "entropy": 6.03049898147583,
      "epoch": 0.21635569670504168,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499844902724723,
      "loss": 5.8151,
      "mean_token_accuracy": 0.13139328137040138,
      "num_tokens": 4752851.0,
      "step": 2180
    },
    {
      "entropy": 6.024123573303223,
      "epoch": 0.2168519253672092,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998435845895302,
      "loss": 5.9025,
      "mean_token_accuracy": 0.12522170394659043,
      "num_tokens": 4764271.0,
      "step": 2185
    },
    {
      "entropy": 6.008631229400635,
      "epoch": 0.21734815402937674,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499842260878701,
      "loss": 5.9643,
      "mean_token_accuracy": 0.12548010498285295,
      "num_tokens": 4776216.0,
      "step": 2190
    },
    {
      "entropy": 6.02488341331482,
      "epoch": 0.21784438269154427,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998409315922683,
      "loss": 5.9141,
      "mean_token_accuracy": 0.1233865849673748,
      "num_tokens": 4786666.0,
      "step": 2195
    },
    {
      "entropy": 6.102518749237061,
      "epoch": 0.2183406113537118,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998395967302649,
      "loss": 5.9867,
      "mean_token_accuracy": 0.11829503029584884,
      "num_tokens": 4798062.0,
      "step": 2200
    },
    {
      "entropy": 6.052379560470581,
      "epoch": 0.21883684001587933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499838256292724,
      "loss": 5.9647,
      "mean_token_accuracy": 0.12422867640852928,
      "num_tokens": 4808837.0,
      "step": 2205
    },
    {
      "entropy": 6.059252071380615,
      "epoch": 0.21933306867804683,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998369102796787,
      "loss": 5.8906,
      "mean_token_accuracy": 0.13139605671167373,
      "num_tokens": 4819894.0,
      "step": 2210
    },
    {
      "entropy": 5.953090047836303,
      "epoch": 0.21982929734021436,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998355586911627,
      "loss": 5.8606,
      "mean_token_accuracy": 0.12828084975481033,
      "num_tokens": 4830976.0,
      "step": 2215
    },
    {
      "entropy": 6.128897476196289,
      "epoch": 0.2203255260023819,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998342015272093,
      "loss": 6.0499,
      "mean_token_accuracy": 0.12443847954273224,
      "num_tokens": 4844292.0,
      "step": 2220
    },
    {
      "entropy": 6.087233448028565,
      "epoch": 0.22082175466454942,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998328387878521,
      "loss": 5.8356,
      "mean_token_accuracy": 0.13550032079219818,
      "num_tokens": 4855284.0,
      "step": 2225
    },
    {
      "entropy": 6.04253740310669,
      "epoch": 0.22131798332671695,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998314704731249,
      "loss": 5.9686,
      "mean_token_accuracy": 0.12750899493694307,
      "num_tokens": 4867801.0,
      "step": 2230
    },
    {
      "entropy": 6.014859342575074,
      "epoch": 0.22181421198888449,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998300965830618,
      "loss": 5.7639,
      "mean_token_accuracy": 0.12936904281377792,
      "num_tokens": 4877392.0,
      "step": 2235
    },
    {
      "entropy": 5.9705253601074215,
      "epoch": 0.22231044065105202,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998287171176968,
      "loss": 5.8695,
      "mean_token_accuracy": 0.12599683701992034,
      "num_tokens": 4888772.0,
      "step": 2240
    },
    {
      "entropy": 6.065585517883301,
      "epoch": 0.22280666931321952,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998273320770641,
      "loss": 5.9199,
      "mean_token_accuracy": 0.12483752593398094,
      "num_tokens": 4899739.0,
      "step": 2245
    },
    {
      "entropy": 6.090867757797241,
      "epoch": 0.22330289797538705,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499825941461198,
      "loss": 5.8943,
      "mean_token_accuracy": 0.1287125341594219,
      "num_tokens": 4910455.0,
      "step": 2250
    },
    {
      "entropy": 6.070242071151734,
      "epoch": 0.22379912663755458,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998245452701329,
      "loss": 5.8417,
      "mean_token_accuracy": 0.12609231770038604,
      "num_tokens": 4921156.0,
      "step": 2255
    },
    {
      "entropy": 6.000742483139038,
      "epoch": 0.2242953552997221,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998231435039036,
      "loss": 5.8579,
      "mean_token_accuracy": 0.1287795290350914,
      "num_tokens": 4931771.0,
      "step": 2260
    },
    {
      "entropy": 5.991134738922119,
      "epoch": 0.22479158396188964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998217361625449,
      "loss": 5.8663,
      "mean_token_accuracy": 0.12458878979086876,
      "num_tokens": 4942790.0,
      "step": 2265
    },
    {
      "entropy": 5.995892524719238,
      "epoch": 0.22528781262405717,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998203232460915,
      "loss": 5.874,
      "mean_token_accuracy": 0.12552486434578897,
      "num_tokens": 4952752.0,
      "step": 2270
    },
    {
      "entropy": 6.009762763977051,
      "epoch": 0.2257840412862247,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998189047545785,
      "loss": 5.9124,
      "mean_token_accuracy": 0.12559892162680625,
      "num_tokens": 4963659.0,
      "step": 2275
    },
    {
      "entropy": 6.0851356983184814,
      "epoch": 0.22628026994839223,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998174806880411,
      "loss": 5.8793,
      "mean_token_accuracy": 0.12859429866075517,
      "num_tokens": 4976121.0,
      "step": 2280
    },
    {
      "entropy": 6.003554630279541,
      "epoch": 0.22677649861055973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998160510465146,
      "loss": 5.9175,
      "mean_token_accuracy": 0.12735827937722205,
      "num_tokens": 4986515.0,
      "step": 2285
    },
    {
      "entropy": 5.9042627811431885,
      "epoch": 0.22727272727272727,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998146158300345,
      "loss": 5.7962,
      "mean_token_accuracy": 0.13313199654221536,
      "num_tokens": 4998231.0,
      "step": 2290
    },
    {
      "entropy": 6.041743040084839,
      "epoch": 0.2277689559348948,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998131750386364,
      "loss": 5.8249,
      "mean_token_accuracy": 0.1294488824903965,
      "num_tokens": 5010309.0,
      "step": 2295
    },
    {
      "entropy": 5.973780727386474,
      "epoch": 0.22826518459706233,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998117286723559,
      "loss": 5.9081,
      "mean_token_accuracy": 0.1309161476790905,
      "num_tokens": 5020864.0,
      "step": 2300
    },
    {
      "entropy": 6.018421363830567,
      "epoch": 0.22876141325922986,
      "grad_norm": 1.0,
      "learning_rate": 0.000499810276731229,
      "loss": 5.9005,
      "mean_token_accuracy": 0.13277044668793678,
      "num_tokens": 5032224.0,
      "step": 2305
    },
    {
      "entropy": 6.084892845153808,
      "epoch": 0.2292576419213974,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998088192152918,
      "loss": 5.8614,
      "mean_token_accuracy": 0.1324816547334194,
      "num_tokens": 5044318.0,
      "step": 2310
    },
    {
      "entropy": 5.994323968887329,
      "epoch": 0.22975387058356492,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998073561245801,
      "loss": 5.9598,
      "mean_token_accuracy": 0.12364556863904,
      "num_tokens": 5055270.0,
      "step": 2315
    },
    {
      "entropy": 6.087515926361084,
      "epoch": 0.23025009924573242,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998058874591305,
      "loss": 5.8998,
      "mean_token_accuracy": 0.13236292973160743,
      "num_tokens": 5064771.0,
      "step": 2320
    },
    {
      "entropy": 6.012254524230957,
      "epoch": 0.23074632790789995,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998044132189793,
      "loss": 5.9151,
      "mean_token_accuracy": 0.12145323902368546,
      "num_tokens": 5075509.0,
      "step": 2325
    },
    {
      "entropy": 5.989501953125,
      "epoch": 0.23124255657006748,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499802933404163,
      "loss": 5.8608,
      "mean_token_accuracy": 0.12346896156668663,
      "num_tokens": 5086270.0,
      "step": 2330
    },
    {
      "entropy": 5.993625497817993,
      "epoch": 0.231738785232235,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998014480147183,
      "loss": 5.8026,
      "mean_token_accuracy": 0.134233608096838,
      "num_tokens": 5096695.0,
      "step": 2335
    },
    {
      "entropy": 6.032075071334839,
      "epoch": 0.23223501389440254,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997999570506823,
      "loss": 5.8928,
      "mean_token_accuracy": 0.1275573804974556,
      "num_tokens": 5107775.0,
      "step": 2340
    },
    {
      "entropy": 5.994977617263794,
      "epoch": 0.23273124255657007,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997984605120917,
      "loss": 5.8537,
      "mean_token_accuracy": 0.1317949816584587,
      "num_tokens": 5118215.0,
      "step": 2345
    },
    {
      "entropy": 6.086479043960571,
      "epoch": 0.2332274712187376,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997969583989835,
      "loss": 5.8716,
      "mean_token_accuracy": 0.1287857286632061,
      "num_tokens": 5129509.0,
      "step": 2350
    },
    {
      "entropy": 5.943372297286987,
      "epoch": 0.23372369988090513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997954507113952,
      "loss": 5.8281,
      "mean_token_accuracy": 0.12884464114904404,
      "num_tokens": 5140689.0,
      "step": 2355
    },
    {
      "entropy": 5.984193468093872,
      "epoch": 0.23421992854307264,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997939374493642,
      "loss": 5.818,
      "mean_token_accuracy": 0.1295442223548889,
      "num_tokens": 5151060.0,
      "step": 2360
    },
    {
      "entropy": 5.96844630241394,
      "epoch": 0.23471615720524017,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499792418612928,
      "loss": 5.7961,
      "mean_token_accuracy": 0.12893855795264245,
      "num_tokens": 5161481.0,
      "step": 2365
    },
    {
      "entropy": 5.9971160888671875,
      "epoch": 0.2352123858674077,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997908942021241,
      "loss": 5.9171,
      "mean_token_accuracy": 0.12333358600735664,
      "num_tokens": 5171961.0,
      "step": 2370
    },
    {
      "entropy": 6.144476413726807,
      "epoch": 0.23570861452957523,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997893642169904,
      "loss": 5.9614,
      "mean_token_accuracy": 0.12242959663271905,
      "num_tokens": 5184158.0,
      "step": 2375
    },
    {
      "entropy": 5.974719429016114,
      "epoch": 0.23620484319174276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997878286575648,
      "loss": 5.9808,
      "mean_token_accuracy": 0.12806885838508605,
      "num_tokens": 5195942.0,
      "step": 2380
    },
    {
      "entropy": 6.029640007019043,
      "epoch": 0.2367010718539103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997862875238855,
      "loss": 5.8598,
      "mean_token_accuracy": 0.12884648367762566,
      "num_tokens": 5208063.0,
      "step": 2385
    },
    {
      "entropy": 6.0588702201843265,
      "epoch": 0.23719730051607782,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997847408159907,
      "loss": 5.9084,
      "mean_token_accuracy": 0.13252404928207398,
      "num_tokens": 5219424.0,
      "step": 2390
    },
    {
      "entropy": 6.0522301197052,
      "epoch": 0.23769352917824532,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997831885339186,
      "loss": 5.9037,
      "mean_token_accuracy": 0.12284734547138214,
      "num_tokens": 5230351.0,
      "step": 2395
    },
    {
      "entropy": 5.945491743087769,
      "epoch": 0.23818975784041285,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997816306777077,
      "loss": 5.79,
      "mean_token_accuracy": 0.12974425926804542,
      "num_tokens": 5242906.0,
      "step": 2400
    },
    {
      "entropy": 5.940090894699097,
      "epoch": 0.23868598650258038,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997800672473969,
      "loss": 5.8165,
      "mean_token_accuracy": 0.13514286726713182,
      "num_tokens": 5253689.0,
      "step": 2405
    },
    {
      "entropy": 6.0798712253570555,
      "epoch": 0.2391822151647479,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997784982430247,
      "loss": 5.9217,
      "mean_token_accuracy": 0.13028180822730065,
      "num_tokens": 5265896.0,
      "step": 2410
    },
    {
      "entropy": 5.999238395690918,
      "epoch": 0.23967844382691544,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997769236646301,
      "loss": 5.832,
      "mean_token_accuracy": 0.12702908590435982,
      "num_tokens": 5276800.0,
      "step": 2415
    },
    {
      "entropy": 5.974053955078125,
      "epoch": 0.24017467248908297,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997753435122522,
      "loss": 5.8315,
      "mean_token_accuracy": 0.13149850741028785,
      "num_tokens": 5289113.0,
      "step": 2420
    },
    {
      "entropy": 5.999874496459961,
      "epoch": 0.2406709011512505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049977375778593,
      "loss": 5.7911,
      "mean_token_accuracy": 0.12800634130835534,
      "num_tokens": 5298825.0,
      "step": 2425
    },
    {
      "entropy": 5.885820531845093,
      "epoch": 0.24116712981341804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997721664857032,
      "loss": 5.8154,
      "mean_token_accuracy": 0.1327855445444584,
      "num_tokens": 5310376.0,
      "step": 2430
    },
    {
      "entropy": 5.9202381610870365,
      "epoch": 0.24166335847558554,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997705696116108,
      "loss": 5.8319,
      "mean_token_accuracy": 0.12816173657774926,
      "num_tokens": 5320483.0,
      "step": 2435
    },
    {
      "entropy": 6.028395891189575,
      "epoch": 0.24215958713775307,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997689671636926,
      "loss": 5.7414,
      "mean_token_accuracy": 0.13838978856801987,
      "num_tokens": 5330995.0,
      "step": 2440
    },
    {
      "entropy": 5.933304786682129,
      "epoch": 0.2426558157999206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997673591419884,
      "loss": 5.8568,
      "mean_token_accuracy": 0.1328567832708359,
      "num_tokens": 5341208.0,
      "step": 2445
    },
    {
      "entropy": 5.9697181224823,
      "epoch": 0.24315204446208813,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997657455465381,
      "loss": 5.9195,
      "mean_token_accuracy": 0.13485939353704451,
      "num_tokens": 5351769.0,
      "step": 2450
    },
    {
      "entropy": 5.988297748565674,
      "epoch": 0.24364827312425566,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997641263773815,
      "loss": 5.7017,
      "mean_token_accuracy": 0.13556905314326287,
      "num_tokens": 5361033.0,
      "step": 2455
    },
    {
      "entropy": 5.82569694519043,
      "epoch": 0.2441445017864232,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499762501634559,
      "loss": 5.7772,
      "mean_token_accuracy": 0.1308267556130886,
      "num_tokens": 5372381.0,
      "step": 2460
    },
    {
      "entropy": 6.007634067535401,
      "epoch": 0.24464073044859072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997608713181108,
      "loss": 5.838,
      "mean_token_accuracy": 0.13373954892158507,
      "num_tokens": 5383101.0,
      "step": 2465
    },
    {
      "entropy": 5.999208927154541,
      "epoch": 0.24513695911075822,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997592354280772,
      "loss": 5.866,
      "mean_token_accuracy": 0.12492905780673028,
      "num_tokens": 5393661.0,
      "step": 2470
    },
    {
      "entropy": 6.018182182312012,
      "epoch": 0.24563318777292575,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499757593964499,
      "loss": 5.8775,
      "mean_token_accuracy": 0.1348043665289879,
      "num_tokens": 5405377.0,
      "step": 2475
    },
    {
      "entropy": 5.926452541351319,
      "epoch": 0.24612941643509328,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997559469274166,
      "loss": 5.7918,
      "mean_token_accuracy": 0.12829321920871734,
      "num_tokens": 5417172.0,
      "step": 2480
    },
    {
      "entropy": 5.959072780609131,
      "epoch": 0.24662564509726081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997542943168712,
      "loss": 5.9262,
      "mean_token_accuracy": 0.12317180931568146,
      "num_tokens": 5428053.0,
      "step": 2485
    },
    {
      "entropy": 6.076896238327026,
      "epoch": 0.24712187375942835,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997526361329035,
      "loss": 5.8564,
      "mean_token_accuracy": 0.12735359892249107,
      "num_tokens": 5438282.0,
      "step": 2490
    },
    {
      "entropy": 5.928952550888061,
      "epoch": 0.24761810242159588,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997509723755547,
      "loss": 5.8189,
      "mean_token_accuracy": 0.13142750561237335,
      "num_tokens": 5448940.0,
      "step": 2495
    },
    {
      "entropy": 5.812957143783569,
      "epoch": 0.2481143310837634,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997493030448662,
      "loss": 5.6781,
      "mean_token_accuracy": 0.13580155298113822,
      "num_tokens": 5458471.0,
      "step": 2500
    },
    {
      "entropy": 5.949634838104248,
      "epoch": 0.24861055974593094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997476281408793,
      "loss": 5.7861,
      "mean_token_accuracy": 0.13658471554517745,
      "num_tokens": 5467979.0,
      "step": 2505
    },
    {
      "entropy": 6.00718846321106,
      "epoch": 0.24910678840809844,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997459476636354,
      "loss": 5.9151,
      "mean_token_accuracy": 0.11804315224289894,
      "num_tokens": 5479852.0,
      "step": 2510
    },
    {
      "entropy": 5.938953304290772,
      "epoch": 0.24960301707026597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997442616131763,
      "loss": 5.7642,
      "mean_token_accuracy": 0.12964859008789062,
      "num_tokens": 5491827.0,
      "step": 2515
    },
    {
      "entropy": 5.895040082931518,
      "epoch": 0.25009924573243353,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997425699895438,
      "loss": 5.7411,
      "mean_token_accuracy": 0.13389707282185553,
      "num_tokens": 5502206.0,
      "step": 2520
    },
    {
      "entropy": 5.923152685165405,
      "epoch": 0.25059547439460106,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997408727927798,
      "loss": 5.779,
      "mean_token_accuracy": 0.12976060807704926,
      "num_tokens": 5514352.0,
      "step": 2525
    },
    {
      "entropy": 5.96109037399292,
      "epoch": 0.25109170305676853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997391700229264,
      "loss": 5.7926,
      "mean_token_accuracy": 0.1397273398935795,
      "num_tokens": 5524413.0,
      "step": 2530
    },
    {
      "entropy": 5.91346435546875,
      "epoch": 0.25158793171893606,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499737461680026,
      "loss": 5.7945,
      "mean_token_accuracy": 0.12549405321478843,
      "num_tokens": 5533692.0,
      "step": 2535
    },
    {
      "entropy": 6.0045164108276365,
      "epoch": 0.2520841603811036,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997357477641207,
      "loss": 5.9264,
      "mean_token_accuracy": 0.13241172581911087,
      "num_tokens": 5545955.0,
      "step": 2540
    },
    {
      "entropy": 5.92925500869751,
      "epoch": 0.2525803890432711,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997340282752532,
      "loss": 5.8668,
      "mean_token_accuracy": 0.13234830275177956,
      "num_tokens": 5557667.0,
      "step": 2545
    },
    {
      "entropy": 5.862270545959473,
      "epoch": 0.25307661770543866,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499732303213466,
      "loss": 5.7624,
      "mean_token_accuracy": 0.12981996908783913,
      "num_tokens": 5567635.0,
      "step": 2550
    },
    {
      "entropy": 5.9047070980072025,
      "epoch": 0.2535728463676062,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997305725788019,
      "loss": 5.7698,
      "mean_token_accuracy": 0.13442888036370276,
      "num_tokens": 5578584.0,
      "step": 2555
    },
    {
      "entropy": 6.005180597305298,
      "epoch": 0.2540690750297737,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499728836371304,
      "loss": 5.8818,
      "mean_token_accuracy": 0.12881023287773133,
      "num_tokens": 5589679.0,
      "step": 2560
    },
    {
      "entropy": 5.980772924423218,
      "epoch": 0.25456530369194125,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997270945910151,
      "loss": 5.8516,
      "mean_token_accuracy": 0.12968761026859282,
      "num_tokens": 5601550.0,
      "step": 2565
    },
    {
      "entropy": 5.974395275115967,
      "epoch": 0.2550615323541088,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997253472379785,
      "loss": 5.8688,
      "mean_token_accuracy": 0.1274749144911766,
      "num_tokens": 5611777.0,
      "step": 2570
    },
    {
      "entropy": 5.999769496917724,
      "epoch": 0.2555577610162763,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997235943122375,
      "loss": 5.8864,
      "mean_token_accuracy": 0.13235417976975442,
      "num_tokens": 5623792.0,
      "step": 2575
    },
    {
      "entropy": 5.866275358200073,
      "epoch": 0.25605398967844384,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997218358138356,
      "loss": 5.6778,
      "mean_token_accuracy": 0.13726607859134674,
      "num_tokens": 5634072.0,
      "step": 2580
    },
    {
      "entropy": 5.999044752120971,
      "epoch": 0.25655021834061137,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997200717428165,
      "loss": 5.9544,
      "mean_token_accuracy": 0.12634836062788962,
      "num_tokens": 5645046.0,
      "step": 2585
    },
    {
      "entropy": 5.904431629180908,
      "epoch": 0.2570464470027789,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997183020992237,
      "loss": 5.809,
      "mean_token_accuracy": 0.1285477615892887,
      "num_tokens": 5655156.0,
      "step": 2590
    },
    {
      "entropy": 5.938637113571167,
      "epoch": 0.25754267566494643,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997165268831014,
      "loss": 5.7818,
      "mean_token_accuracy": 0.13181826770305632,
      "num_tokens": 5665621.0,
      "step": 2595
    },
    {
      "entropy": 5.93091344833374,
      "epoch": 0.25803890432711396,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997147460944933,
      "loss": 5.8289,
      "mean_token_accuracy": 0.12735895141959191,
      "num_tokens": 5677072.0,
      "step": 2600
    },
    {
      "entropy": 6.021169185638428,
      "epoch": 0.25853513298928144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997129597334438,
      "loss": 5.7873,
      "mean_token_accuracy": 0.1364290252327919,
      "num_tokens": 5687926.0,
      "step": 2605
    },
    {
      "entropy": 5.8465227603912355,
      "epoch": 0.25903136165144897,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997111677999971,
      "loss": 5.7934,
      "mean_token_accuracy": 0.13265275880694388,
      "num_tokens": 5698923.0,
      "step": 2610
    },
    {
      "entropy": 5.994477939605713,
      "epoch": 0.2595275903136165,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997093702941976,
      "loss": 5.9077,
      "mean_token_accuracy": 0.12608013227581977,
      "num_tokens": 5709879.0,
      "step": 2615
    },
    {
      "entropy": 6.012147760391235,
      "epoch": 0.260023818975784,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049970756721609,
      "loss": 5.849,
      "mean_token_accuracy": 0.13115165755152702,
      "num_tokens": 5720698.0,
      "step": 2620
    },
    {
      "entropy": 5.843365859985352,
      "epoch": 0.26052004763795156,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499705758565719,
      "loss": 5.8038,
      "mean_token_accuracy": 0.13133684992790223,
      "num_tokens": 5732890.0,
      "step": 2625
    },
    {
      "entropy": 5.973391675949097,
      "epoch": 0.2610162763001191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997039443431292,
      "loss": 5.7878,
      "mean_token_accuracy": 0.12773183211684228,
      "num_tokens": 5743741.0,
      "step": 2630
    },
    {
      "entropy": 5.905499076843261,
      "epoch": 0.2615125049622866,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997021245483659,
      "loss": 5.7137,
      "mean_token_accuracy": 0.13542912974953653,
      "num_tokens": 5754730.0,
      "step": 2635
    },
    {
      "entropy": 5.871477842330933,
      "epoch": 0.26200873362445415,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499700299181474,
      "loss": 5.7528,
      "mean_token_accuracy": 0.13477372154593467,
      "num_tokens": 5764705.0,
      "step": 2640
    },
    {
      "entropy": 5.9722490310668945,
      "epoch": 0.2625049622866217,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996984682424989,
      "loss": 5.865,
      "mean_token_accuracy": 0.12664012461900712,
      "num_tokens": 5776803.0,
      "step": 2645
    },
    {
      "entropy": 6.041519641876221,
      "epoch": 0.2630011909487892,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499696631731486,
      "loss": 5.8975,
      "mean_token_accuracy": 0.1250507615506649,
      "num_tokens": 5788203.0,
      "step": 2650
    },
    {
      "entropy": 6.021883153915406,
      "epoch": 0.26349741961095674,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996947896484809,
      "loss": 5.8195,
      "mean_token_accuracy": 0.12791492864489556,
      "num_tokens": 5800476.0,
      "step": 2655
    },
    {
      "entropy": 5.887351179122925,
      "epoch": 0.26399364827312427,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499692941993529,
      "loss": 5.8694,
      "mean_token_accuracy": 0.12800540328025817,
      "num_tokens": 5811652.0,
      "step": 2660
    },
    {
      "entropy": 5.990099906921387,
      "epoch": 0.2644898769352918,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996910887666765,
      "loss": 5.7816,
      "mean_token_accuracy": 0.1410517655313015,
      "num_tokens": 5823036.0,
      "step": 2665
    },
    {
      "entropy": 5.909138345718384,
      "epoch": 0.26498610559745933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499689229967969,
      "loss": 5.7264,
      "mean_token_accuracy": 0.13503767773509026,
      "num_tokens": 5833220.0,
      "step": 2670
    },
    {
      "entropy": 5.86794605255127,
      "epoch": 0.26548233425962686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996873655974529,
      "loss": 5.8144,
      "mean_token_accuracy": 0.12698522582650185,
      "num_tokens": 5843777.0,
      "step": 2675
    },
    {
      "entropy": 5.916513824462891,
      "epoch": 0.26597856292179434,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996854956551743,
      "loss": 5.8224,
      "mean_token_accuracy": 0.1307629682123661,
      "num_tokens": 5854942.0,
      "step": 2680
    },
    {
      "entropy": 5.988381481170654,
      "epoch": 0.26647479158396187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996836201411794,
      "loss": 5.8396,
      "mean_token_accuracy": 0.1324163593351841,
      "num_tokens": 5866536.0,
      "step": 2685
    },
    {
      "entropy": 5.9175981998443605,
      "epoch": 0.2669710202461294,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499681739055515,
      "loss": 5.7219,
      "mean_token_accuracy": 0.1395683117210865,
      "num_tokens": 5876653.0,
      "step": 2690
    },
    {
      "entropy": 5.919457578659058,
      "epoch": 0.26746724890829693,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996798523982276,
      "loss": 5.7461,
      "mean_token_accuracy": 0.12860351651906968,
      "num_tokens": 5887092.0,
      "step": 2695
    },
    {
      "entropy": 5.938131856918335,
      "epoch": 0.26796347757046446,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499677960169364,
      "loss": 5.8721,
      "mean_token_accuracy": 0.1272119738161564,
      "num_tokens": 5898224.0,
      "step": 2700
    },
    {
      "entropy": 5.945243692398071,
      "epoch": 0.268459706232632,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996760623689711,
      "loss": 5.7136,
      "mean_token_accuracy": 0.13979979678988458,
      "num_tokens": 5909235.0,
      "step": 2705
    },
    {
      "entropy": 5.943400955200195,
      "epoch": 0.2689559348947995,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499674158997096,
      "loss": 5.8699,
      "mean_token_accuracy": 0.13342854082584382,
      "num_tokens": 5921299.0,
      "step": 2710
    },
    {
      "entropy": 5.9730690002441404,
      "epoch": 0.26945216355696705,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996722500537859,
      "loss": 5.8334,
      "mean_token_accuracy": 0.12727866619825362,
      "num_tokens": 5933249.0,
      "step": 2715
    },
    {
      "entropy": 5.9161705493927,
      "epoch": 0.2699483922191346,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996703355390881,
      "loss": 5.8614,
      "mean_token_accuracy": 0.12605185210704803,
      "num_tokens": 5945942.0,
      "step": 2720
    },
    {
      "entropy": 5.93488187789917,
      "epoch": 0.2704446208813021,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049966841545305,
      "loss": 5.7666,
      "mean_token_accuracy": 0.1335825651884079,
      "num_tokens": 5956315.0,
      "step": 2725
    },
    {
      "entropy": 5.966945314407349,
      "epoch": 0.27094084954346964,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996664897957195,
      "loss": 5.854,
      "mean_token_accuracy": 0.12438153848052025,
      "num_tokens": 5966566.0,
      "step": 2730
    },
    {
      "entropy": 5.866949892044067,
      "epoch": 0.2714370782056372,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996645585671441,
      "loss": 5.7093,
      "mean_token_accuracy": 0.13864025473594666,
      "num_tokens": 5977137.0,
      "step": 2735
    },
    {
      "entropy": 5.936009550094605,
      "epoch": 0.2719333068678047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996626217673717,
      "loss": 5.8313,
      "mean_token_accuracy": 0.1317468449473381,
      "num_tokens": 5987653.0,
      "step": 2740
    },
    {
      "entropy": 5.963485193252564,
      "epoch": 0.27242953552997223,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996606793964504,
      "loss": 5.887,
      "mean_token_accuracy": 0.1295153670012951,
      "num_tokens": 5998530.0,
      "step": 2745
    },
    {
      "entropy": 5.893201017379761,
      "epoch": 0.27292576419213976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996587314544284,
      "loss": 5.7099,
      "mean_token_accuracy": 0.13821978718042374,
      "num_tokens": 6009550.0,
      "step": 2750
    },
    {
      "entropy": 5.940571975708008,
      "epoch": 0.27342199285430724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996567779413539,
      "loss": 5.9101,
      "mean_token_accuracy": 0.12932207062840462,
      "num_tokens": 6021522.0,
      "step": 2755
    },
    {
      "entropy": 5.959524393081665,
      "epoch": 0.27391822151647477,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996548188572755,
      "loss": 5.7807,
      "mean_token_accuracy": 0.13457212075591088,
      "num_tokens": 6033993.0,
      "step": 2760
    },
    {
      "entropy": 5.988877677917481,
      "epoch": 0.2744144501786423,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996528542022416,
      "loss": 5.9458,
      "mean_token_accuracy": 0.1262620911002159,
      "num_tokens": 6044804.0,
      "step": 2765
    },
    {
      "entropy": 5.921116924285888,
      "epoch": 0.27491067884080983,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499650883976301,
      "loss": 5.7866,
      "mean_token_accuracy": 0.1267233669757843,
      "num_tokens": 6054878.0,
      "step": 2770
    },
    {
      "entropy": 5.956991338729859,
      "epoch": 0.27540690750297736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996489081795026,
      "loss": 5.7891,
      "mean_token_accuracy": 0.13266447335481643,
      "num_tokens": 6065492.0,
      "step": 2775
    },
    {
      "entropy": 5.925124549865723,
      "epoch": 0.2759031361651449,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996469268118953,
      "loss": 5.811,
      "mean_token_accuracy": 0.12755917012691498,
      "num_tokens": 6077360.0,
      "step": 2780
    },
    {
      "entropy": 5.918424940109253,
      "epoch": 0.2763993648273124,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996449398735283,
      "loss": 5.7938,
      "mean_token_accuracy": 0.12594451755285263,
      "num_tokens": 6087530.0,
      "step": 2785
    },
    {
      "entropy": 5.934395456314087,
      "epoch": 0.27689559348947995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996429473644509,
      "loss": 5.7659,
      "mean_token_accuracy": 0.13185836747288704,
      "num_tokens": 6098064.0,
      "step": 2790
    },
    {
      "entropy": 5.828168869018555,
      "epoch": 0.2773918221516475,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996409492847124,
      "loss": 5.701,
      "mean_token_accuracy": 0.13495547845959663,
      "num_tokens": 6108446.0,
      "step": 2795
    },
    {
      "entropy": 5.9052245140075685,
      "epoch": 0.277888050813815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996389456343624,
      "loss": 5.7486,
      "mean_token_accuracy": 0.13034897968173026,
      "num_tokens": 6118923.0,
      "step": 2800
    },
    {
      "entropy": 5.881168937683105,
      "epoch": 0.27838427947598254,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996369364134507,
      "loss": 5.6589,
      "mean_token_accuracy": 0.1399786502122879,
      "num_tokens": 6130583.0,
      "step": 2805
    },
    {
      "entropy": 5.936345386505127,
      "epoch": 0.2788805081381501,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499634921622027,
      "loss": 5.8739,
      "mean_token_accuracy": 0.12864407300949096,
      "num_tokens": 6142351.0,
      "step": 2810
    },
    {
      "entropy": 5.869996118545532,
      "epoch": 0.2793767368003176,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004996329012601413,
      "loss": 5.69,
      "mean_token_accuracy": 0.135867390781641,
      "num_tokens": 6152964.0,
      "step": 2815
    },
    {
      "entropy": 5.823973894119263,
      "epoch": 0.27987296546248513,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996308753278437,
      "loss": 5.7623,
      "mean_token_accuracy": 0.1387517362833023,
      "num_tokens": 6163522.0,
      "step": 2820
    },
    {
      "entropy": 5.948193883895874,
      "epoch": 0.28036919412465267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996288438251843,
      "loss": 5.7879,
      "mean_token_accuracy": 0.12974680438637734,
      "num_tokens": 6174060.0,
      "step": 2825
    },
    {
      "entropy": 5.917904472351074,
      "epoch": 0.28086542278682014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996268067522137,
      "loss": 5.701,
      "mean_token_accuracy": 0.13888110667467118,
      "num_tokens": 6183850.0,
      "step": 2830
    },
    {
      "entropy": 5.905739784240723,
      "epoch": 0.28136165144898767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996247641089823,
      "loss": 5.7671,
      "mean_token_accuracy": 0.12889229655265808,
      "num_tokens": 6195073.0,
      "step": 2835
    },
    {
      "entropy": 5.953272962570191,
      "epoch": 0.2818578801111552,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996227158955409,
      "loss": 5.8007,
      "mean_token_accuracy": 0.12981446608901023,
      "num_tokens": 6206658.0,
      "step": 2840
    },
    {
      "entropy": 5.779361295700073,
      "epoch": 0.28235410877332273,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049962066211194,
      "loss": 5.6868,
      "mean_token_accuracy": 0.13519837632775306,
      "num_tokens": 6217183.0,
      "step": 2845
    },
    {
      "entropy": 5.8369691371917725,
      "epoch": 0.28285033743549026,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996186027582308,
      "loss": 5.7279,
      "mean_token_accuracy": 0.13328900784254075,
      "num_tokens": 6227575.0,
      "step": 2850
    },
    {
      "entropy": 5.891994619369507,
      "epoch": 0.2833465660976578,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996165378344642,
      "loss": 5.7363,
      "mean_token_accuracy": 0.13181785643100738,
      "num_tokens": 6239698.0,
      "step": 2855
    },
    {
      "entropy": 5.858404922485351,
      "epoch": 0.2838427947598253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996144673406914,
      "loss": 5.7423,
      "mean_token_accuracy": 0.1368052139878273,
      "num_tokens": 6249730.0,
      "step": 2860
    },
    {
      "entropy": 5.834589242935181,
      "epoch": 0.28433902342199285,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996123912769639,
      "loss": 5.7313,
      "mean_token_accuracy": 0.13907284140586854,
      "num_tokens": 6259493.0,
      "step": 2865
    },
    {
      "entropy": 5.90723762512207,
      "epoch": 0.2848352520841604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996103096433332,
      "loss": 5.74,
      "mean_token_accuracy": 0.13290882706642151,
      "num_tokens": 6269568.0,
      "step": 2870
    },
    {
      "entropy": 5.922413015365601,
      "epoch": 0.2853314807463279,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996082224398506,
      "loss": 5.7645,
      "mean_token_accuracy": 0.1388328976929188,
      "num_tokens": 6280028.0,
      "step": 2875
    },
    {
      "entropy": 5.826960611343384,
      "epoch": 0.28582770940849545,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996061296665682,
      "loss": 5.7286,
      "mean_token_accuracy": 0.1314896821975708,
      "num_tokens": 6290321.0,
      "step": 2880
    },
    {
      "entropy": 5.862280654907226,
      "epoch": 0.286323938070663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996040313235377,
      "loss": 5.728,
      "mean_token_accuracy": 0.1321492835879326,
      "num_tokens": 6300729.0,
      "step": 2885
    },
    {
      "entropy": 5.939273405075073,
      "epoch": 0.2868201667328305,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996019274108112,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1290019102394581,
      "num_tokens": 6311051.0,
      "step": 2890
    },
    {
      "entropy": 5.802321290969848,
      "epoch": 0.28731639539499804,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499599817928441,
      "loss": 5.6792,
      "mean_token_accuracy": 0.1415045864880085,
      "num_tokens": 6322201.0,
      "step": 2895
    },
    {
      "entropy": 5.875109529495239,
      "epoch": 0.28781262405716557,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995977028764792,
      "loss": 5.7885,
      "mean_token_accuracy": 0.133917373418808,
      "num_tokens": 6332372.0,
      "step": 2900
    },
    {
      "entropy": 5.954572486877441,
      "epoch": 0.28830885271933304,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995955822549783,
      "loss": 5.899,
      "mean_token_accuracy": 0.13203557208180428,
      "num_tokens": 6344070.0,
      "step": 2905
    },
    {
      "entropy": 5.788153886795044,
      "epoch": 0.2888050813815006,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995934560639909,
      "loss": 5.7194,
      "mean_token_accuracy": 0.13280046507716178,
      "num_tokens": 6355118.0,
      "step": 2910
    },
    {
      "entropy": 5.910306978225708,
      "epoch": 0.2893013100436681,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995913243035699,
      "loss": 5.7816,
      "mean_token_accuracy": 0.13115150183439256,
      "num_tokens": 6366722.0,
      "step": 2915
    },
    {
      "entropy": 5.846759939193726,
      "epoch": 0.28979753870583563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995891869737678,
      "loss": 5.7226,
      "mean_token_accuracy": 0.13123665228486062,
      "num_tokens": 6376190.0,
      "step": 2920
    },
    {
      "entropy": 5.852357482910156,
      "epoch": 0.29029376736800316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995870440746379,
      "loss": 5.8527,
      "mean_token_accuracy": 0.1258479841053486,
      "num_tokens": 6385883.0,
      "step": 2925
    },
    {
      "entropy": 5.875295686721802,
      "epoch": 0.2907899960301707,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995848956062333,
      "loss": 5.7043,
      "mean_token_accuracy": 0.1373466357588768,
      "num_tokens": 6397762.0,
      "step": 2930
    },
    {
      "entropy": 5.876016473770141,
      "epoch": 0.2912862246923382,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995827415686071,
      "loss": 5.8161,
      "mean_token_accuracy": 0.12998180985450744,
      "num_tokens": 6408972.0,
      "step": 2935
    },
    {
      "entropy": 5.8990397453308105,
      "epoch": 0.29178245335450576,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499580581961813,
      "loss": 5.6794,
      "mean_token_accuracy": 0.13430978432297708,
      "num_tokens": 6418998.0,
      "step": 2940
    },
    {
      "entropy": 5.945958471298217,
      "epoch": 0.2922786820166733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995784167859043,
      "loss": 5.7651,
      "mean_token_accuracy": 0.13749361261725426,
      "num_tokens": 6430046.0,
      "step": 2945
    },
    {
      "entropy": 5.919869470596313,
      "epoch": 0.2927749106788408,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995762460409349,
      "loss": 5.7428,
      "mean_token_accuracy": 0.13194330036640167,
      "num_tokens": 6441351.0,
      "step": 2950
    },
    {
      "entropy": 5.815007972717285,
      "epoch": 0.29327113934100835,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995740697269584,
      "loss": 5.7621,
      "mean_token_accuracy": 0.1333144061267376,
      "num_tokens": 6453259.0,
      "step": 2955
    },
    {
      "entropy": 5.883697414398194,
      "epoch": 0.2937673680031759,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995718878440289,
      "loss": 5.7516,
      "mean_token_accuracy": 0.1323343835771084,
      "num_tokens": 6464819.0,
      "step": 2960
    },
    {
      "entropy": 5.958118867874146,
      "epoch": 0.2942635966653434,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995697003922006,
      "loss": 5.8363,
      "mean_token_accuracy": 0.13035116493701934,
      "num_tokens": 6477315.0,
      "step": 2965
    },
    {
      "entropy": 5.872132253646851,
      "epoch": 0.29475982532751094,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995675073715274,
      "loss": 5.7451,
      "mean_token_accuracy": 0.13140062168240546,
      "num_tokens": 6488461.0,
      "step": 2970
    },
    {
      "entropy": 5.804051685333252,
      "epoch": 0.29525605398967847,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995653087820641,
      "loss": 5.7166,
      "mean_token_accuracy": 0.13911211416125296,
      "num_tokens": 6499486.0,
      "step": 2975
    },
    {
      "entropy": 5.947262525558472,
      "epoch": 0.29575228265184594,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499563104623865,
      "loss": 5.8091,
      "mean_token_accuracy": 0.13345877602696418,
      "num_tokens": 6511337.0,
      "step": 2980
    },
    {
      "entropy": 5.84906907081604,
      "epoch": 0.2962485113140135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995608948969849,
      "loss": 5.6918,
      "mean_token_accuracy": 0.13708140105009078,
      "num_tokens": 6521066.0,
      "step": 2985
    },
    {
      "entropy": 5.782366609573364,
      "epoch": 0.296744739976181,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995586796014784,
      "loss": 5.7117,
      "mean_token_accuracy": 0.13593292757868766,
      "num_tokens": 6531267.0,
      "step": 2990
    },
    {
      "entropy": 5.831585788726807,
      "epoch": 0.29724096863834853,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995564587374005,
      "loss": 5.7673,
      "mean_token_accuracy": 0.12948553413152694,
      "num_tokens": 6541891.0,
      "step": 2995
    },
    {
      "entropy": 5.903474807739258,
      "epoch": 0.29773719730051607,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995542323048063,
      "loss": 5.7197,
      "mean_token_accuracy": 0.13689327314496041,
      "num_tokens": 6551974.0,
      "step": 3000
    },
    {
      "epoch": 0.29773719730051607,
      "eval_entropy": 5.687596190596,
      "eval_loss": 5.815427303314209,
      "eval_mean_token_accuracy": 0.13528745296530922,
      "eval_num_tokens": 6551974.0,
      "eval_runtime": 20.4138,
      "eval_samples_per_second": 1649.324,
      "eval_steps_per_second": 206.184,
      "step": 3000
    },
    {
      "entropy": 5.882981634140014,
      "epoch": 0.2982334259626836,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499552000303751,
      "loss": 5.723,
      "mean_token_accuracy": 0.1339751213788986,
      "num_tokens": 6563296.0,
      "step": 3005
    },
    {
      "entropy": 5.90754656791687,
      "epoch": 0.2987296546248511,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049954976273429,
      "loss": 5.7498,
      "mean_token_accuracy": 0.13449311405420303,
      "num_tokens": 6573699.0,
      "step": 3010
    },
    {
      "entropy": 5.8666716575622555,
      "epoch": 0.29922588328701866,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995475195964787,
      "loss": 5.7606,
      "mean_token_accuracy": 0.13438182771205903,
      "num_tokens": 6585111.0,
      "step": 3015
    },
    {
      "entropy": 5.81152982711792,
      "epoch": 0.2997221119491862,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995452708903727,
      "loss": 5.7462,
      "mean_token_accuracy": 0.13328394740819932,
      "num_tokens": 6596102.0,
      "step": 3020
    },
    {
      "entropy": 5.835455322265625,
      "epoch": 0.3002183406113537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499543016616028,
      "loss": 5.7082,
      "mean_token_accuracy": 0.136111568659544,
      "num_tokens": 6605989.0,
      "step": 3025
    },
    {
      "entropy": 5.906055116653443,
      "epoch": 0.30071456927352125,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995407567735002,
      "loss": 5.7201,
      "mean_token_accuracy": 0.1376255288720131,
      "num_tokens": 6615771.0,
      "step": 3030
    },
    {
      "entropy": 5.843295335769653,
      "epoch": 0.3012107979356888,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995384913628456,
      "loss": 5.6698,
      "mean_token_accuracy": 0.14295522272586822,
      "num_tokens": 6626707.0,
      "step": 3035
    },
    {
      "entropy": 5.8286217212677,
      "epoch": 0.3017070265978563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995362203841202,
      "loss": 5.7206,
      "mean_token_accuracy": 0.1341872878372669,
      "num_tokens": 6637539.0,
      "step": 3040
    },
    {
      "entropy": 5.806274366378784,
      "epoch": 0.30220325526002384,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995339438373804,
      "loss": 5.7765,
      "mean_token_accuracy": 0.13231148198246956,
      "num_tokens": 6648372.0,
      "step": 3045
    },
    {
      "entropy": 5.85441632270813,
      "epoch": 0.30269948392219137,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995316617226825,
      "loss": 5.7798,
      "mean_token_accuracy": 0.12775983437895774,
      "num_tokens": 6658732.0,
      "step": 3050
    },
    {
      "entropy": 5.814883470535278,
      "epoch": 0.30319571258435885,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995293740400833,
      "loss": 5.6723,
      "mean_token_accuracy": 0.1364850677549839,
      "num_tokens": 6670108.0,
      "step": 3055
    },
    {
      "entropy": 5.913384056091308,
      "epoch": 0.3036919412465264,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995270807896395,
      "loss": 5.7863,
      "mean_token_accuracy": 0.1281758651137352,
      "num_tokens": 6681337.0,
      "step": 3060
    },
    {
      "entropy": 5.834350061416626,
      "epoch": 0.3041881699086939,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995247819714079,
      "loss": 5.6932,
      "mean_token_accuracy": 0.13551875576376915,
      "num_tokens": 6691221.0,
      "step": 3065
    },
    {
      "entropy": 5.903665971755982,
      "epoch": 0.30468439857086144,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995224775854454,
      "loss": 5.8462,
      "mean_token_accuracy": 0.13598475977778435,
      "num_tokens": 6703363.0,
      "step": 3070
    },
    {
      "entropy": 5.861040687561035,
      "epoch": 0.30518062723302897,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004995201676318093,
      "loss": 5.7654,
      "mean_token_accuracy": 0.13301002755761146,
      "num_tokens": 6715526.0,
      "step": 3075
    },
    {
      "entropy": 5.880378770828247,
      "epoch": 0.3056768558951965,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995178521105568,
      "loss": 5.7365,
      "mean_token_accuracy": 0.13436401709914209,
      "num_tokens": 6725607.0,
      "step": 3080
    },
    {
      "entropy": 5.82697434425354,
      "epoch": 0.30617308455736403,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995155310217454,
      "loss": 5.6883,
      "mean_token_accuracy": 0.13515269681811332,
      "num_tokens": 6735762.0,
      "step": 3085
    },
    {
      "entropy": 5.901894903182983,
      "epoch": 0.30666931321953156,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995132043654327,
      "loss": 5.7937,
      "mean_token_accuracy": 0.1385086104273796,
      "num_tokens": 6746334.0,
      "step": 3090
    },
    {
      "entropy": 5.829314661026001,
      "epoch": 0.3071655418816991,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995108721416763,
      "loss": 5.7596,
      "mean_token_accuracy": 0.1301199622452259,
      "num_tokens": 6756082.0,
      "step": 3095
    },
    {
      "entropy": 5.889105224609375,
      "epoch": 0.3076617705438666,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995085343505339,
      "loss": 5.7931,
      "mean_token_accuracy": 0.13027504384517669,
      "num_tokens": 6765723.0,
      "step": 3100
    },
    {
      "entropy": 5.871817350387573,
      "epoch": 0.30815799920603415,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995061909920638,
      "loss": 5.7832,
      "mean_token_accuracy": 0.13101766183972358,
      "num_tokens": 6778567.0,
      "step": 3105
    },
    {
      "entropy": 5.938895654678345,
      "epoch": 0.3086542278682017,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995038420663238,
      "loss": 5.8083,
      "mean_token_accuracy": 0.13469240814447403,
      "num_tokens": 6789721.0,
      "step": 3110
    },
    {
      "entropy": 5.8680150508880615,
      "epoch": 0.3091504565303692,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995014875733724,
      "loss": 5.7106,
      "mean_token_accuracy": 0.13767961412668228,
      "num_tokens": 6800606.0,
      "step": 3115
    },
    {
      "entropy": 5.868447160720825,
      "epoch": 0.30964668519253674,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994991275132678,
      "loss": 5.7612,
      "mean_token_accuracy": 0.13593339994549752,
      "num_tokens": 6813068.0,
      "step": 3120
    },
    {
      "entropy": 5.829099702835083,
      "epoch": 0.31014291385470427,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994967618860685,
      "loss": 5.71,
      "mean_token_accuracy": 0.13283495754003524,
      "num_tokens": 6824502.0,
      "step": 3125
    },
    {
      "entropy": 5.7927391052246096,
      "epoch": 0.31063914251687175,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994943906918333,
      "loss": 5.735,
      "mean_token_accuracy": 0.13616894111037253,
      "num_tokens": 6835434.0,
      "step": 3130
    },
    {
      "entropy": 5.864362525939941,
      "epoch": 0.3111353711790393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994920139306211,
      "loss": 5.7729,
      "mean_token_accuracy": 0.13050154894590377,
      "num_tokens": 6846754.0,
      "step": 3135
    },
    {
      "entropy": 5.929287338256836,
      "epoch": 0.3116315998412068,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994896316024906,
      "loss": 5.7156,
      "mean_token_accuracy": 0.13192819207906722,
      "num_tokens": 6858050.0,
      "step": 3140
    },
    {
      "entropy": 5.853343677520752,
      "epoch": 0.31212782850337434,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499487243707501,
      "loss": 5.7851,
      "mean_token_accuracy": 0.13169284760951996,
      "num_tokens": 6869940.0,
      "step": 3145
    },
    {
      "entropy": 5.852452278137207,
      "epoch": 0.31262405716554187,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994848502457114,
      "loss": 5.7262,
      "mean_token_accuracy": 0.1333612710237503,
      "num_tokens": 6882161.0,
      "step": 3150
    },
    {
      "entropy": 5.9310907363891605,
      "epoch": 0.3131202858277094,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994824512171812,
      "loss": 5.7443,
      "mean_token_accuracy": 0.1340456210076809,
      "num_tokens": 6893236.0,
      "step": 3155
    },
    {
      "entropy": 5.868790340423584,
      "epoch": 0.31361651448987693,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994800466219701,
      "loss": 5.7497,
      "mean_token_accuracy": 0.1378454878926277,
      "num_tokens": 6904934.0,
      "step": 3160
    },
    {
      "entropy": 5.864860486984253,
      "epoch": 0.31411274315204446,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994776364601374,
      "loss": 5.7541,
      "mean_token_accuracy": 0.13124591931700708,
      "num_tokens": 6917537.0,
      "step": 3165
    },
    {
      "entropy": 5.872111368179321,
      "epoch": 0.314608971814212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994752207317431,
      "loss": 5.7994,
      "mean_token_accuracy": 0.13613324910402297,
      "num_tokens": 6928354.0,
      "step": 3170
    },
    {
      "entropy": 5.814365100860596,
      "epoch": 0.3151052004763795,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499472799436847,
      "loss": 5.684,
      "mean_token_accuracy": 0.13755037263035774,
      "num_tokens": 6939553.0,
      "step": 3175
    },
    {
      "entropy": 5.825185632705688,
      "epoch": 0.31560142913854705,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994703725755092,
      "loss": 5.6875,
      "mean_token_accuracy": 0.13778624460101127,
      "num_tokens": 6950233.0,
      "step": 3180
    },
    {
      "entropy": 5.839802265167236,
      "epoch": 0.3160976578007146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994679401477899,
      "loss": 5.7014,
      "mean_token_accuracy": 0.14181062579154968,
      "num_tokens": 6960716.0,
      "step": 3185
    },
    {
      "entropy": 5.808838844299316,
      "epoch": 0.3165938864628821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994655021537492,
      "loss": 5.667,
      "mean_token_accuracy": 0.1420467786490917,
      "num_tokens": 6970391.0,
      "step": 3190
    },
    {
      "entropy": 5.872873067855835,
      "epoch": 0.31709011512504964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994630585934478,
      "loss": 5.6756,
      "mean_token_accuracy": 0.13863585740327836,
      "num_tokens": 6981067.0,
      "step": 3195
    },
    {
      "entropy": 5.823529815673828,
      "epoch": 0.3175863437872172,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994606094669463,
      "loss": 5.7471,
      "mean_token_accuracy": 0.1366729259490967,
      "num_tokens": 6992000.0,
      "step": 3200
    },
    {
      "entropy": 5.871815347671509,
      "epoch": 0.3180825724493847,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994581547743053,
      "loss": 5.6927,
      "mean_token_accuracy": 0.13655175417661666,
      "num_tokens": 7001791.0,
      "step": 3205
    },
    {
      "entropy": 5.864743328094482,
      "epoch": 0.3185788011115522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994556945155857,
      "loss": 5.7496,
      "mean_token_accuracy": 0.13631674349308015,
      "num_tokens": 7011635.0,
      "step": 3210
    },
    {
      "entropy": 5.829900646209717,
      "epoch": 0.3190750297737197,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004994532286908486,
      "loss": 5.7548,
      "mean_token_accuracy": 0.1374804101884365,
      "num_tokens": 7023011.0,
      "step": 3215
    },
    {
      "entropy": 5.872418022155761,
      "epoch": 0.31957125843588724,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499450757300155,
      "loss": 5.6644,
      "mean_token_accuracy": 0.13625743612647057,
      "num_tokens": 7033756.0,
      "step": 3220
    },
    {
      "entropy": 5.811505079269409,
      "epoch": 0.32006748709805477,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994482803435664,
      "loss": 5.763,
      "mean_token_accuracy": 0.12896935492753983,
      "num_tokens": 7045038.0,
      "step": 3225
    },
    {
      "entropy": 5.8763984680175785,
      "epoch": 0.3205637157602223,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499445797821144,
      "loss": 5.7462,
      "mean_token_accuracy": 0.13087846636772155,
      "num_tokens": 7057181.0,
      "step": 3230
    },
    {
      "entropy": 5.877984523773193,
      "epoch": 0.32105994442238983,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994433097329494,
      "loss": 5.8529,
      "mean_token_accuracy": 0.1271327830851078,
      "num_tokens": 7069441.0,
      "step": 3235
    },
    {
      "entropy": 5.862765216827393,
      "epoch": 0.32155617308455736,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994408160790444,
      "loss": 5.7301,
      "mean_token_accuracy": 0.13759879618883133,
      "num_tokens": 7080017.0,
      "step": 3240
    },
    {
      "entropy": 5.809614372253418,
      "epoch": 0.3220524017467249,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994383168594907,
      "loss": 5.7454,
      "mean_token_accuracy": 0.1369093529880047,
      "num_tokens": 7091034.0,
      "step": 3245
    },
    {
      "entropy": 5.857351207733155,
      "epoch": 0.3225486304088924,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994358120743506,
      "loss": 5.6766,
      "mean_token_accuracy": 0.1324448935687542,
      "num_tokens": 7101227.0,
      "step": 3250
    },
    {
      "entropy": 5.806218767166138,
      "epoch": 0.32304485907105995,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994333017236858,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1320954255759716,
      "num_tokens": 7112056.0,
      "step": 3255
    },
    {
      "entropy": 5.820009040832519,
      "epoch": 0.3235410877332275,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994307858075587,
      "loss": 5.674,
      "mean_token_accuracy": 0.13769950717687607,
      "num_tokens": 7121895.0,
      "step": 3260
    },
    {
      "entropy": 5.820459938049316,
      "epoch": 0.324037316395395,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994282643260319,
      "loss": 5.5998,
      "mean_token_accuracy": 0.14075906649231912,
      "num_tokens": 7132726.0,
      "step": 3265
    },
    {
      "entropy": 5.792088985443115,
      "epoch": 0.32453354505756254,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994257372791676,
      "loss": 5.674,
      "mean_token_accuracy": 0.13695003762841224,
      "num_tokens": 7143394.0,
      "step": 3270
    },
    {
      "entropy": 5.800633430480957,
      "epoch": 0.3250297737197301,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994232046670286,
      "loss": 5.676,
      "mean_token_accuracy": 0.13709723353385925,
      "num_tokens": 7154416.0,
      "step": 3275
    },
    {
      "entropy": 5.834659719467163,
      "epoch": 0.3255260023818976,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994206664896778,
      "loss": 5.761,
      "mean_token_accuracy": 0.13632258474826814,
      "num_tokens": 7164876.0,
      "step": 3280
    },
    {
      "entropy": 5.830479335784912,
      "epoch": 0.3260222310440651,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994181227471779,
      "loss": 5.6436,
      "mean_token_accuracy": 0.14082633778452874,
      "num_tokens": 7176053.0,
      "step": 3285
    },
    {
      "entropy": 5.790161228179931,
      "epoch": 0.3265184597062326,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994155734395923,
      "loss": 5.7964,
      "mean_token_accuracy": 0.13124819248914718,
      "num_tokens": 7188180.0,
      "step": 3290
    },
    {
      "entropy": 5.907830667495728,
      "epoch": 0.32701468836840014,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499413018566984,
      "loss": 5.8258,
      "mean_token_accuracy": 0.12964817509055138,
      "num_tokens": 7198012.0,
      "step": 3295
    },
    {
      "entropy": 5.774489688873291,
      "epoch": 0.32751091703056767,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994104581294163,
      "loss": 5.6503,
      "mean_token_accuracy": 0.14036565124988556,
      "num_tokens": 7209465.0,
      "step": 3300
    },
    {
      "entropy": 5.828147268295288,
      "epoch": 0.3280071456927352,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994078921269529,
      "loss": 5.7186,
      "mean_token_accuracy": 0.13260019421577454,
      "num_tokens": 7220546.0,
      "step": 3305
    },
    {
      "entropy": 5.8164819240570065,
      "epoch": 0.32850337435490273,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994053205596573,
      "loss": 5.7431,
      "mean_token_accuracy": 0.13402992039918898,
      "num_tokens": 7232870.0,
      "step": 3310
    },
    {
      "entropy": 5.745219469070435,
      "epoch": 0.32899960301707026,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994027434275934,
      "loss": 5.6136,
      "mean_token_accuracy": 0.14180929958820343,
      "num_tokens": 7242052.0,
      "step": 3315
    },
    {
      "entropy": 5.778253364562988,
      "epoch": 0.3294958316792378,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994001607308249,
      "loss": 5.6931,
      "mean_token_accuracy": 0.1311803065240383,
      "num_tokens": 7253060.0,
      "step": 3320
    },
    {
      "entropy": 5.939854621887207,
      "epoch": 0.3299920603414053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499397572469416,
      "loss": 5.8246,
      "mean_token_accuracy": 0.12777294889092444,
      "num_tokens": 7264034.0,
      "step": 3325
    },
    {
      "entropy": 5.9013756275177,
      "epoch": 0.33048828900357285,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993949786434308,
      "loss": 5.7524,
      "mean_token_accuracy": 0.13291485533118247,
      "num_tokens": 7275865.0,
      "step": 3330
    },
    {
      "entropy": 5.788418531417847,
      "epoch": 0.3309845176657404,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993923792529337,
      "loss": 5.612,
      "mean_token_accuracy": 0.13734735026955605,
      "num_tokens": 7286652.0,
      "step": 3335
    },
    {
      "entropy": 5.7966104507446286,
      "epoch": 0.3314807463279079,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004993897742979892,
      "loss": 5.7748,
      "mean_token_accuracy": 0.1369300901889801,
      "num_tokens": 7298032.0,
      "step": 3340
    },
    {
      "entropy": 5.865872478485107,
      "epoch": 0.33197697499007545,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993871637786616,
      "loss": 5.7163,
      "mean_token_accuracy": 0.1389184132218361,
      "num_tokens": 7310883.0,
      "step": 3345
    },
    {
      "entropy": 5.824404382705689,
      "epoch": 0.332473203652243,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499384547695016,
      "loss": 5.7027,
      "mean_token_accuracy": 0.13594060242176056,
      "num_tokens": 7320934.0,
      "step": 3350
    },
    {
      "entropy": 5.824160146713257,
      "epoch": 0.3329694323144105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499381926047117,
      "loss": 5.7346,
      "mean_token_accuracy": 0.13489899337291716,
      "num_tokens": 7331137.0,
      "step": 3355
    },
    {
      "entropy": 5.900968980789185,
      "epoch": 0.333465660976578,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00049937929883503,
      "loss": 5.7548,
      "mean_token_accuracy": 0.1311749204993248,
      "num_tokens": 7341697.0,
      "step": 3360
    },
    {
      "entropy": 5.811262893676758,
      "epoch": 0.3339618896387455,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993766660588196,
      "loss": 5.6705,
      "mean_token_accuracy": 0.13751054629683496,
      "num_tokens": 7352952.0,
      "step": 3365
    },
    {
      "entropy": 5.816409969329834,
      "epoch": 0.33445811830091304,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993740277185515,
      "loss": 5.6562,
      "mean_token_accuracy": 0.13477067425847053,
      "num_tokens": 7362904.0,
      "step": 3370
    },
    {
      "entropy": 5.86172308921814,
      "epoch": 0.3349543469630806,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499371383814291,
      "loss": 5.7228,
      "mean_token_accuracy": 0.13629542663693428,
      "num_tokens": 7372457.0,
      "step": 3375
    },
    {
      "entropy": 5.805365371704101,
      "epoch": 0.3354505756252481,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993687343461036,
      "loss": 5.7635,
      "mean_token_accuracy": 0.13371680974960326,
      "num_tokens": 7383732.0,
      "step": 3380
    },
    {
      "entropy": 5.803901481628418,
      "epoch": 0.33594680428741563,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993660793140551,
      "loss": 5.6232,
      "mean_token_accuracy": 0.14322842955589293,
      "num_tokens": 7395985.0,
      "step": 3385
    },
    {
      "entropy": 5.777278280258178,
      "epoch": 0.33644303294958317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993634187182113,
      "loss": 5.6112,
      "mean_token_accuracy": 0.13524336218833924,
      "num_tokens": 7405508.0,
      "step": 3390
    },
    {
      "entropy": 5.827437019348144,
      "epoch": 0.3369392616117507,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993607525586381,
      "loss": 5.7334,
      "mean_token_accuracy": 0.13744548633694648,
      "num_tokens": 7416369.0,
      "step": 3395
    },
    {
      "entropy": 5.858999013900757,
      "epoch": 0.3374354902739182,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993580808354018,
      "loss": 5.7504,
      "mean_token_accuracy": 0.13679931461811065,
      "num_tokens": 7427357.0,
      "step": 3400
    },
    {
      "entropy": 5.929110050201416,
      "epoch": 0.33793171893608576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993554035485685,
      "loss": 5.6525,
      "mean_token_accuracy": 0.1368493355810642,
      "num_tokens": 7438019.0,
      "step": 3405
    },
    {
      "entropy": 5.839983987808227,
      "epoch": 0.3384279475982533,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993527206982047,
      "loss": 5.7426,
      "mean_token_accuracy": 0.13790037259459495,
      "num_tokens": 7449063.0,
      "step": 3410
    },
    {
      "entropy": 5.795181846618652,
      "epoch": 0.3389241762604208,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993500322843767,
      "loss": 5.6322,
      "mean_token_accuracy": 0.13845305293798446,
      "num_tokens": 7459731.0,
      "step": 3415
    },
    {
      "entropy": 5.876691818237305,
      "epoch": 0.33942040492258835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993473383071514,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1342042103409767,
      "num_tokens": 7470763.0,
      "step": 3420
    },
    {
      "entropy": 5.721733140945434,
      "epoch": 0.3399166335847559,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993446387665956,
      "loss": 5.6467,
      "mean_token_accuracy": 0.14225304871797562,
      "num_tokens": 7482281.0,
      "step": 3425
    },
    {
      "entropy": 5.8150310039520265,
      "epoch": 0.3404128622469234,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993419336627761,
      "loss": 5.7149,
      "mean_token_accuracy": 0.12838553860783578,
      "num_tokens": 7494622.0,
      "step": 3430
    },
    {
      "entropy": 5.812551784515381,
      "epoch": 0.3409090909090909,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993392229957602,
      "loss": 5.6458,
      "mean_token_accuracy": 0.14302448630332948,
      "num_tokens": 7505554.0,
      "step": 3435
    },
    {
      "entropy": 5.889950942993164,
      "epoch": 0.3414053195712584,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993365067656148,
      "loss": 5.7392,
      "mean_token_accuracy": 0.13819737657904624,
      "num_tokens": 7517308.0,
      "step": 3440
    },
    {
      "entropy": 5.832235288619995,
      "epoch": 0.34190154823342594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993337849724076,
      "loss": 5.7959,
      "mean_token_accuracy": 0.13020050451159476,
      "num_tokens": 7527937.0,
      "step": 3445
    },
    {
      "entropy": 5.814507007598877,
      "epoch": 0.3423977768955935,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993310576162059,
      "loss": 5.6939,
      "mean_token_accuracy": 0.1342161625623703,
      "num_tokens": 7538987.0,
      "step": 3450
    },
    {
      "entropy": 5.765037298202515,
      "epoch": 0.342894005557761,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993283246970772,
      "loss": 5.6618,
      "mean_token_accuracy": 0.13538794219493866,
      "num_tokens": 7550526.0,
      "step": 3455
    },
    {
      "entropy": 5.777789354324341,
      "epoch": 0.34339023421992854,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993255862150896,
      "loss": 5.6918,
      "mean_token_accuracy": 0.14151629582047462,
      "num_tokens": 7561705.0,
      "step": 3460
    },
    {
      "entropy": 5.869984912872314,
      "epoch": 0.34388646288209607,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993228421703108,
      "loss": 5.7532,
      "mean_token_accuracy": 0.12912511378526687,
      "num_tokens": 7573789.0,
      "step": 3465
    },
    {
      "entropy": 5.775468492507935,
      "epoch": 0.3443826915442636,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993200925628088,
      "loss": 5.5987,
      "mean_token_accuracy": 0.13950269743800164,
      "num_tokens": 7585516.0,
      "step": 3470
    },
    {
      "entropy": 5.856809949874878,
      "epoch": 0.3448789202064311,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499317337392652,
      "loss": 5.7337,
      "mean_token_accuracy": 0.1322330914437771,
      "num_tokens": 7596324.0,
      "step": 3475
    },
    {
      "entropy": 5.9036643505096436,
      "epoch": 0.34537514886859866,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993145766599085,
      "loss": 5.7481,
      "mean_token_accuracy": 0.13191625624895095,
      "num_tokens": 7608103.0,
      "step": 3480
    },
    {
      "entropy": 5.76446042060852,
      "epoch": 0.3458713775307662,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993118103646469,
      "loss": 5.6446,
      "mean_token_accuracy": 0.13612786456942558,
      "num_tokens": 7619442.0,
      "step": 3485
    },
    {
      "entropy": 5.824655771255493,
      "epoch": 0.3463676061929337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993090385069357,
      "loss": 5.7224,
      "mean_token_accuracy": 0.14050939530134202,
      "num_tokens": 7630066.0,
      "step": 3490
    },
    {
      "entropy": 5.747276163101196,
      "epoch": 0.34686383485510125,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993062610868437,
      "loss": 5.6047,
      "mean_token_accuracy": 0.13780295699834824,
      "num_tokens": 7641209.0,
      "step": 3495
    },
    {
      "entropy": 5.670969772338867,
      "epoch": 0.3473600635172688,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993034781044398,
      "loss": 5.6636,
      "mean_token_accuracy": 0.1388409838080406,
      "num_tokens": 7652957.0,
      "step": 3500
    },
    {
      "entropy": 5.755992269515991,
      "epoch": 0.3478562921794363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499300689559793,
      "loss": 5.6574,
      "mean_token_accuracy": 0.13754271045327188,
      "num_tokens": 7664538.0,
      "step": 3505
    },
    {
      "entropy": 5.8704993724823,
      "epoch": 0.3483525208416038,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992978954529724,
      "loss": 5.6779,
      "mean_token_accuracy": 0.13819587901234626,
      "num_tokens": 7674596.0,
      "step": 3510
    },
    {
      "entropy": 5.7532305240631105,
      "epoch": 0.3488487495037713,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992950957840473,
      "loss": 5.6497,
      "mean_token_accuracy": 0.13185194060206412,
      "num_tokens": 7685496.0,
      "step": 3515
    },
    {
      "entropy": 5.846642875671387,
      "epoch": 0.34934497816593885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499292290553087,
      "loss": 5.6766,
      "mean_token_accuracy": 0.13124992325901985,
      "num_tokens": 7696202.0,
      "step": 3520
    },
    {
      "entropy": 5.722525644302368,
      "epoch": 0.3498412068281064,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992894797601614,
      "loss": 5.5982,
      "mean_token_accuracy": 0.14212893396615983,
      "num_tokens": 7706570.0,
      "step": 3525
    },
    {
      "entropy": 5.713914155960083,
      "epoch": 0.3503374354902739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049928666340534,
      "loss": 5.5978,
      "mean_token_accuracy": 0.14290860816836357,
      "num_tokens": 7716903.0,
      "step": 3530
    },
    {
      "entropy": 5.817956018447876,
      "epoch": 0.35083366415244144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992838414886925,
      "loss": 5.6679,
      "mean_token_accuracy": 0.13681603819131852,
      "num_tokens": 7727059.0,
      "step": 3535
    },
    {
      "entropy": 5.844083213806153,
      "epoch": 0.35132989281460897,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992810140102891,
      "loss": 5.817,
      "mean_token_accuracy": 0.132791168987751,
      "num_tokens": 7739089.0,
      "step": 3540
    },
    {
      "entropy": 5.902175951004028,
      "epoch": 0.3518261214767765,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004992781809701999,
      "loss": 5.7011,
      "mean_token_accuracy": 0.13764540925621987,
      "num_tokens": 7750953.0,
      "step": 3545
    },
    {
      "entropy": 5.7494512557983395,
      "epoch": 0.35232235013894403,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992753423684951,
      "loss": 5.6575,
      "mean_token_accuracy": 0.1336328312754631,
      "num_tokens": 7761535.0,
      "step": 3550
    },
    {
      "entropy": 5.885289049148559,
      "epoch": 0.35281857880111156,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992724982052449,
      "loss": 5.8164,
      "mean_token_accuracy": 0.13148643001914023,
      "num_tokens": 7773497.0,
      "step": 3555
    },
    {
      "entropy": 5.791857385635376,
      "epoch": 0.3533148074632791,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00049926964848052,
      "loss": 5.6892,
      "mean_token_accuracy": 0.136527568846941,
      "num_tokens": 7785764.0,
      "step": 3560
    },
    {
      "entropy": 5.803077554702758,
      "epoch": 0.3538110361254466,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992667931943913,
      "loss": 5.6081,
      "mean_token_accuracy": 0.1423060715198517,
      "num_tokens": 7797231.0,
      "step": 3565
    },
    {
      "entropy": 5.818060493469238,
      "epoch": 0.35430726478761415,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992639323469293,
      "loss": 5.6565,
      "mean_token_accuracy": 0.1348166771233082,
      "num_tokens": 7809218.0,
      "step": 3570
    },
    {
      "entropy": 5.820521354675293,
      "epoch": 0.3548034934497817,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499261065938205,
      "loss": 5.729,
      "mean_token_accuracy": 0.13844639733433722,
      "num_tokens": 7820418.0,
      "step": 3575
    },
    {
      "entropy": 5.759602737426758,
      "epoch": 0.3552997221119492,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992581939682895,
      "loss": 5.5965,
      "mean_token_accuracy": 0.13811820447444917,
      "num_tokens": 7831034.0,
      "step": 3580
    },
    {
      "entropy": 5.7398828029632565,
      "epoch": 0.3557959507741167,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499255316437254,
      "loss": 5.65,
      "mean_token_accuracy": 0.13426289111375808,
      "num_tokens": 7843703.0,
      "step": 3585
    },
    {
      "entropy": 5.789868974685669,
      "epoch": 0.3562921794362842,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004992524333451699,
      "loss": 5.6411,
      "mean_token_accuracy": 0.14170614182949065,
      "num_tokens": 7855806.0,
      "step": 3590
    },
    {
      "entropy": 5.8667271614074705,
      "epoch": 0.35678840809845175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992495446921087,
      "loss": 5.771,
      "mean_token_accuracy": 0.13592937514185904,
      "num_tokens": 7868638.0,
      "step": 3595
    },
    {
      "entropy": 5.832484102249145,
      "epoch": 0.3572846367606193,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992466504781419,
      "loss": 5.61,
      "mean_token_accuracy": 0.13937883526086808,
      "num_tokens": 7880269.0,
      "step": 3600
    },
    {
      "entropy": 5.655418539047242,
      "epoch": 0.3577808654227868,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992437507033415,
      "loss": 5.5792,
      "mean_token_accuracy": 0.14191157892346382,
      "num_tokens": 7889870.0,
      "step": 3605
    },
    {
      "entropy": 5.8278790473937985,
      "epoch": 0.35827709408495434,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992408453677792,
      "loss": 5.7588,
      "mean_token_accuracy": 0.1345938950777054,
      "num_tokens": 7902025.0,
      "step": 3610
    },
    {
      "entropy": 5.720179271697998,
      "epoch": 0.35877332274712187,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992379344715273,
      "loss": 5.5518,
      "mean_token_accuracy": 0.1420609712600708,
      "num_tokens": 7912994.0,
      "step": 3615
    },
    {
      "entropy": 5.690258884429932,
      "epoch": 0.3592695514092894,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992350180146577,
      "loss": 5.6346,
      "mean_token_accuracy": 0.14315218478441238,
      "num_tokens": 7924184.0,
      "step": 3620
    },
    {
      "entropy": 5.766508722305298,
      "epoch": 0.35976578007145693,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992320959972428,
      "loss": 5.6139,
      "mean_token_accuracy": 0.1391090139746666,
      "num_tokens": 7935394.0,
      "step": 3625
    },
    {
      "entropy": 5.826987028121948,
      "epoch": 0.36026200873362446,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992291684193551,
      "loss": 5.6947,
      "mean_token_accuracy": 0.14261544197797776,
      "num_tokens": 7946699.0,
      "step": 3630
    },
    {
      "entropy": 5.727631664276123,
      "epoch": 0.360758237395792,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992262352810672,
      "loss": 5.6244,
      "mean_token_accuracy": 0.14122700542211533,
      "num_tokens": 7957362.0,
      "step": 3635
    },
    {
      "entropy": 5.7845306396484375,
      "epoch": 0.3612544660579595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992232965824518,
      "loss": 5.6247,
      "mean_token_accuracy": 0.13957104384899138,
      "num_tokens": 7967316.0,
      "step": 3640
    },
    {
      "entropy": 5.745174980163574,
      "epoch": 0.36175069472012705,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499220352323582,
      "loss": 5.5914,
      "mean_token_accuracy": 0.13948371410369872,
      "num_tokens": 7978163.0,
      "step": 3645
    },
    {
      "entropy": 5.7267382621765135,
      "epoch": 0.3622469233822946,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992174025045304,
      "loss": 5.608,
      "mean_token_accuracy": 0.1417738080024719,
      "num_tokens": 7988441.0,
      "step": 3650
    },
    {
      "entropy": 5.768844985961914,
      "epoch": 0.3627431520444621,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992144471253704,
      "loss": 5.6009,
      "mean_token_accuracy": 0.1405134156346321,
      "num_tokens": 7999779.0,
      "step": 3655
    },
    {
      "entropy": 5.757862758636475,
      "epoch": 0.3632393807066296,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992114861861753,
      "loss": 5.7477,
      "mean_token_accuracy": 0.14123099967837333,
      "num_tokens": 8010921.0,
      "step": 3660
    },
    {
      "entropy": 5.829793691635132,
      "epoch": 0.3637356093687971,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992085196870184,
      "loss": 5.7302,
      "mean_token_accuracy": 0.13367349281907082,
      "num_tokens": 8020918.0,
      "step": 3665
    },
    {
      "entropy": 5.803525733947754,
      "epoch": 0.36423183803096465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992055476279734,
      "loss": 5.7629,
      "mean_token_accuracy": 0.13460451290011405,
      "num_tokens": 8032238.0,
      "step": 3670
    },
    {
      "entropy": 5.767388010025025,
      "epoch": 0.3647280666931322,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992025700091139,
      "loss": 5.5947,
      "mean_token_accuracy": 0.1451639488339424,
      "num_tokens": 8043288.0,
      "step": 3675
    },
    {
      "entropy": 5.678320407867432,
      "epoch": 0.3652242953552997,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991995868305138,
      "loss": 5.575,
      "mean_token_accuracy": 0.14363528192043304,
      "num_tokens": 8053356.0,
      "step": 3680
    },
    {
      "entropy": 5.754720401763916,
      "epoch": 0.36572052401746724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991965980922471,
      "loss": 5.5695,
      "mean_token_accuracy": 0.143849016726017,
      "num_tokens": 8062641.0,
      "step": 3685
    },
    {
      "entropy": 5.825892162322998,
      "epoch": 0.36621675267963477,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991936037943878,
      "loss": 5.7519,
      "mean_token_accuracy": 0.13373227939009666,
      "num_tokens": 8074232.0,
      "step": 3690
    },
    {
      "entropy": 5.803862142562866,
      "epoch": 0.3667129813418023,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991906039370104,
      "loss": 5.5952,
      "mean_token_accuracy": 0.1464700773358345,
      "num_tokens": 8085276.0,
      "step": 3695
    },
    {
      "entropy": 5.7944032669067385,
      "epoch": 0.36720921000396983,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991875985201889,
      "loss": 5.6434,
      "mean_token_accuracy": 0.14082100316882135,
      "num_tokens": 8095769.0,
      "step": 3700
    },
    {
      "entropy": 5.80278844833374,
      "epoch": 0.36770543866613736,
      "grad_norm": 0.875,
      "learning_rate": 0.000499184587543998,
      "loss": 5.5932,
      "mean_token_accuracy": 0.1406601808965206,
      "num_tokens": 8106932.0,
      "step": 3705
    },
    {
      "entropy": 5.849697589874268,
      "epoch": 0.3682016673283049,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991815710085126,
      "loss": 5.6995,
      "mean_token_accuracy": 0.14065353721380233,
      "num_tokens": 8118307.0,
      "step": 3710
    },
    {
      "entropy": 5.700325536727905,
      "epoch": 0.3686978959904724,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991785489138072,
      "loss": 5.62,
      "mean_token_accuracy": 0.1361625850200653,
      "num_tokens": 8129940.0,
      "step": 3715
    },
    {
      "entropy": 5.791812515258789,
      "epoch": 0.36919412465263995,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991755212599569,
      "loss": 5.6925,
      "mean_token_accuracy": 0.13987574577331544,
      "num_tokens": 8141362.0,
      "step": 3720
    },
    {
      "entropy": 5.749632120132446,
      "epoch": 0.3696903533148075,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991724880470366,
      "loss": 5.6204,
      "mean_token_accuracy": 0.13807517513632775,
      "num_tokens": 8151795.0,
      "step": 3725
    },
    {
      "entropy": 5.730589914321899,
      "epoch": 0.370186581976975,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991694492751217,
      "loss": 5.6279,
      "mean_token_accuracy": 0.13879199996590613,
      "num_tokens": 8162490.0,
      "step": 3730
    },
    {
      "entropy": 5.788167142868042,
      "epoch": 0.3706828106391425,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991664049442876,
      "loss": 5.609,
      "mean_token_accuracy": 0.14127933382987976,
      "num_tokens": 8173546.0,
      "step": 3735
    },
    {
      "entropy": 5.755032300949097,
      "epoch": 0.37117903930131,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991633550546097,
      "loss": 5.6779,
      "mean_token_accuracy": 0.1373685859143734,
      "num_tokens": 8184079.0,
      "step": 3740
    },
    {
      "entropy": 5.807260894775391,
      "epoch": 0.37167526796347755,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991602996061635,
      "loss": 5.7216,
      "mean_token_accuracy": 0.13596394509077073,
      "num_tokens": 8195603.0,
      "step": 3745
    },
    {
      "entropy": 5.8333714485168455,
      "epoch": 0.3721714966256451,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991572385990249,
      "loss": 5.6465,
      "mean_token_accuracy": 0.13658567145466805,
      "num_tokens": 8206605.0,
      "step": 3750
    },
    {
      "entropy": 5.756965494155883,
      "epoch": 0.3726677252878126,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991541720332698,
      "loss": 5.6923,
      "mean_token_accuracy": 0.13872257843613625,
      "num_tokens": 8216950.0,
      "step": 3755
    },
    {
      "entropy": 5.783972215652466,
      "epoch": 0.37316395394998014,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991510999089743,
      "loss": 5.7008,
      "mean_token_accuracy": 0.13334987908601761,
      "num_tokens": 8227272.0,
      "step": 3760
    },
    {
      "entropy": 5.7975975513458256,
      "epoch": 0.3736601826121477,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991480222262145,
      "loss": 5.6797,
      "mean_token_accuracy": 0.1449749782681465,
      "num_tokens": 8238062.0,
      "step": 3765
    },
    {
      "entropy": 5.760014581680298,
      "epoch": 0.3741564112743152,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991449389850668,
      "loss": 5.6489,
      "mean_token_accuracy": 0.1373499646782875,
      "num_tokens": 8248851.0,
      "step": 3770
    },
    {
      "entropy": 5.825553894042969,
      "epoch": 0.37465263993648273,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991418501856076,
      "loss": 5.7235,
      "mean_token_accuracy": 0.13612475544214248,
      "num_tokens": 8258535.0,
      "step": 3775
    },
    {
      "entropy": 5.769029140472412,
      "epoch": 0.37514886859865026,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991387558279134,
      "loss": 5.6499,
      "mean_token_accuracy": 0.1392560936510563,
      "num_tokens": 8271666.0,
      "step": 3780
    },
    {
      "entropy": 5.819280052185059,
      "epoch": 0.3756450972608178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991356559120612,
      "loss": 5.7209,
      "mean_token_accuracy": 0.14241980984807015,
      "num_tokens": 8281869.0,
      "step": 3785
    },
    {
      "entropy": 5.8115629196167,
      "epoch": 0.3761413259229853,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991325504381275,
      "loss": 5.6971,
      "mean_token_accuracy": 0.13476150706410409,
      "num_tokens": 8292699.0,
      "step": 3790
    },
    {
      "entropy": 5.8070549964904785,
      "epoch": 0.37663755458515286,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991294394061896,
      "loss": 5.6069,
      "mean_token_accuracy": 0.14085992574691772,
      "num_tokens": 8302832.0,
      "step": 3795
    },
    {
      "entropy": 5.854214143753052,
      "epoch": 0.3771337832473204,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991263228163246,
      "loss": 5.7081,
      "mean_token_accuracy": 0.13581942692399024,
      "num_tokens": 8314721.0,
      "step": 3800
    },
    {
      "entropy": 5.720626640319824,
      "epoch": 0.3776300119094879,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991232006686096,
      "loss": 5.5751,
      "mean_token_accuracy": 0.14003885090351104,
      "num_tokens": 8325140.0,
      "step": 3805
    },
    {
      "entropy": 5.800462341308593,
      "epoch": 0.3781262405716554,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991200729631223,
      "loss": 5.6418,
      "mean_token_accuracy": 0.13866187483072281,
      "num_tokens": 8336509.0,
      "step": 3810
    },
    {
      "entropy": 5.7718969821929935,
      "epoch": 0.3786224692338229,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991169396999401,
      "loss": 5.6419,
      "mean_token_accuracy": 0.14797513633966447,
      "num_tokens": 8346999.0,
      "step": 3815
    },
    {
      "entropy": 5.7060408115386965,
      "epoch": 0.37911869789599045,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991138008791408,
      "loss": 5.577,
      "mean_token_accuracy": 0.14889908730983734,
      "num_tokens": 8358697.0,
      "step": 3820
    },
    {
      "entropy": 5.87032299041748,
      "epoch": 0.379614926558158,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991106565008021,
      "loss": 5.698,
      "mean_token_accuracy": 0.14287034720182418,
      "num_tokens": 8370365.0,
      "step": 3825
    },
    {
      "entropy": 5.7740264415740965,
      "epoch": 0.3801111552203255,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991075065650019,
      "loss": 5.6022,
      "mean_token_accuracy": 0.13797869756817818,
      "num_tokens": 8381137.0,
      "step": 3830
    },
    {
      "entropy": 5.766485166549683,
      "epoch": 0.38060738388249304,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991043510718186,
      "loss": 5.7333,
      "mean_token_accuracy": 0.14498398676514626,
      "num_tokens": 8393299.0,
      "step": 3835
    },
    {
      "entropy": 5.800220966339111,
      "epoch": 0.3811036125446606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991011900213303,
      "loss": 5.6736,
      "mean_token_accuracy": 0.13635399639606477,
      "num_tokens": 8404564.0,
      "step": 3840
    },
    {
      "entropy": 5.8245477199554445,
      "epoch": 0.3815998412068281,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990980234136152,
      "loss": 5.6168,
      "mean_token_accuracy": 0.1448028489947319,
      "num_tokens": 8413850.0,
      "step": 3845
    },
    {
      "entropy": 5.660411930084228,
      "epoch": 0.38209606986899564,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990948512487522,
      "loss": 5.6055,
      "mean_token_accuracy": 0.14423333257436752,
      "num_tokens": 8425226.0,
      "step": 3850
    },
    {
      "entropy": 5.832178592681885,
      "epoch": 0.38259229853116317,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004990916735268198,
      "loss": 5.7127,
      "mean_token_accuracy": 0.13525868132710456,
      "num_tokens": 8435714.0,
      "step": 3855
    },
    {
      "entropy": 5.867836236953735,
      "epoch": 0.3830885271933307,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990884902478966,
      "loss": 5.6165,
      "mean_token_accuracy": 0.14263269007205964,
      "num_tokens": 8446691.0,
      "step": 3860
    },
    {
      "entropy": 5.7082971096038815,
      "epoch": 0.3835847558554982,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990853014120619,
      "loss": 5.6964,
      "mean_token_accuracy": 0.13977687433362007,
      "num_tokens": 8457376.0,
      "step": 3865
    },
    {
      "entropy": 5.740559101104736,
      "epoch": 0.38408098451766576,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990821070193944,
      "loss": 5.6439,
      "mean_token_accuracy": 0.13726354837417604,
      "num_tokens": 8469456.0,
      "step": 3870
    },
    {
      "entropy": 5.908693313598633,
      "epoch": 0.3845772131798333,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990789070699736,
      "loss": 5.6818,
      "mean_token_accuracy": 0.13804868757724761,
      "num_tokens": 8481142.0,
      "step": 3875
    },
    {
      "entropy": 5.807751131057739,
      "epoch": 0.3850734418420008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990757015638787,
      "loss": 5.7742,
      "mean_token_accuracy": 0.13008464649319648,
      "num_tokens": 8492900.0,
      "step": 3880
    },
    {
      "entropy": 5.720052480697632,
      "epoch": 0.3855696705041683,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990724905011894,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1429848238825798,
      "num_tokens": 8503323.0,
      "step": 3885
    },
    {
      "entropy": 5.745321655273438,
      "epoch": 0.3860658991663358,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499069273881985,
      "loss": 5.6884,
      "mean_token_accuracy": 0.13964812234044074,
      "num_tokens": 8514713.0,
      "step": 3890
    },
    {
      "entropy": 5.787202024459839,
      "epoch": 0.38656212782850335,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990660517063455,
      "loss": 5.5265,
      "mean_token_accuracy": 0.14724477753043175,
      "num_tokens": 8524907.0,
      "step": 3895
    },
    {
      "entropy": 5.70192666053772,
      "epoch": 0.3870583564906709,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990628239743508,
      "loss": 5.5562,
      "mean_token_accuracy": 0.14487576633691787,
      "num_tokens": 8535769.0,
      "step": 3900
    },
    {
      "entropy": 5.772091341018677,
      "epoch": 0.3875545851528384,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990595906860807,
      "loss": 5.7239,
      "mean_token_accuracy": 0.133495282381773,
      "num_tokens": 8547754.0,
      "step": 3905
    },
    {
      "entropy": 5.754996252059937,
      "epoch": 0.38805081381500595,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990563518416157,
      "loss": 5.6319,
      "mean_token_accuracy": 0.13957725316286088,
      "num_tokens": 8558896.0,
      "step": 3910
    },
    {
      "entropy": 5.704990673065185,
      "epoch": 0.3885470424771735,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990531074410358,
      "loss": 5.5356,
      "mean_token_accuracy": 0.14678843393921853,
      "num_tokens": 8568721.0,
      "step": 3915
    },
    {
      "entropy": 5.747829723358154,
      "epoch": 0.389043271139341,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990498574844218,
      "loss": 5.6026,
      "mean_token_accuracy": 0.14324251040816308,
      "num_tokens": 8579199.0,
      "step": 3920
    },
    {
      "entropy": 5.807630395889282,
      "epoch": 0.38953949980150854,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990466019718542,
      "loss": 5.7108,
      "mean_token_accuracy": 0.13559925258159639,
      "num_tokens": 8590362.0,
      "step": 3925
    },
    {
      "entropy": 5.819594430923462,
      "epoch": 0.39003572846367607,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990433409034134,
      "loss": 5.6471,
      "mean_token_accuracy": 0.1393386095762253,
      "num_tokens": 8600603.0,
      "step": 3930
    },
    {
      "entropy": 5.806178140640259,
      "epoch": 0.3905319571258436,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990400742791806,
      "loss": 5.6356,
      "mean_token_accuracy": 0.13758892714977264,
      "num_tokens": 8611436.0,
      "step": 3935
    },
    {
      "entropy": 5.742629671096802,
      "epoch": 0.39102818578801113,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990368020992368,
      "loss": 5.7252,
      "mean_token_accuracy": 0.13721646517515182,
      "num_tokens": 8624152.0,
      "step": 3940
    },
    {
      "entropy": 5.714175748825073,
      "epoch": 0.39152441445017866,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499033524363663,
      "loss": 5.537,
      "mean_token_accuracy": 0.13984050825238228,
      "num_tokens": 8634762.0,
      "step": 3945
    },
    {
      "entropy": 5.751743459701538,
      "epoch": 0.3920206431123462,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990302410725405,
      "loss": 5.5328,
      "mean_token_accuracy": 0.14400038719177247,
      "num_tokens": 8646204.0,
      "step": 3950
    },
    {
      "entropy": 5.6987059116363525,
      "epoch": 0.3925168717745137,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990269522259509,
      "loss": 5.5748,
      "mean_token_accuracy": 0.14200279861688614,
      "num_tokens": 8656662.0,
      "step": 3955
    },
    {
      "entropy": 5.7163866519927975,
      "epoch": 0.3930131004366812,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990236578239755,
      "loss": 5.5743,
      "mean_token_accuracy": 0.1435669742524624,
      "num_tokens": 8668180.0,
      "step": 3960
    },
    {
      "entropy": 5.773800706863403,
      "epoch": 0.3935093290988487,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990203578666962,
      "loss": 5.6141,
      "mean_token_accuracy": 0.14074693843722344,
      "num_tokens": 8679522.0,
      "step": 3965
    },
    {
      "entropy": 5.742657661437988,
      "epoch": 0.39400555776101626,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990170523541948,
      "loss": 5.7092,
      "mean_token_accuracy": 0.13487801551818848,
      "num_tokens": 8691451.0,
      "step": 3970
    },
    {
      "entropy": 5.752674293518067,
      "epoch": 0.3945017864231838,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990137412865531,
      "loss": 5.5689,
      "mean_token_accuracy": 0.14206789955496787,
      "num_tokens": 8702437.0,
      "step": 3975
    },
    {
      "entropy": 5.741733694076538,
      "epoch": 0.3949980150853513,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990104246638533,
      "loss": 5.5927,
      "mean_token_accuracy": 0.1416883200407028,
      "num_tokens": 8714612.0,
      "step": 3980
    },
    {
      "entropy": 5.680162668228149,
      "epoch": 0.39549424374751885,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990071024861778,
      "loss": 5.5748,
      "mean_token_accuracy": 0.13880547657608985,
      "num_tokens": 8725065.0,
      "step": 3985
    },
    {
      "entropy": 5.717659330368042,
      "epoch": 0.3959904724096864,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990037747536087,
      "loss": 5.5698,
      "mean_token_accuracy": 0.1459265246987343,
      "num_tokens": 8736142.0,
      "step": 3990
    },
    {
      "entropy": 5.804776287078857,
      "epoch": 0.3964867010718539,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990004414662288,
      "loss": 5.6869,
      "mean_token_accuracy": 0.1372356206178665,
      "num_tokens": 8746708.0,
      "step": 3995
    },
    {
      "entropy": 5.815282821655273,
      "epoch": 0.39698292973402144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989971026241205,
      "loss": 5.587,
      "mean_token_accuracy": 0.14142581522464753,
      "num_tokens": 8756756.0,
      "step": 4000
    },
    {
      "entropy": 5.711295700073242,
      "epoch": 0.39747915839618897,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989937582273668,
      "loss": 5.617,
      "mean_token_accuracy": 0.1431009478867054,
      "num_tokens": 8767151.0,
      "step": 4005
    },
    {
      "entropy": 5.77561936378479,
      "epoch": 0.3979753870583565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989904082760506,
      "loss": 5.7365,
      "mean_token_accuracy": 0.13625101894140243,
      "num_tokens": 8778169.0,
      "step": 4010
    },
    {
      "entropy": 5.763059139251709,
      "epoch": 0.39847161572052403,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498987052770255,
      "loss": 5.5918,
      "mean_token_accuracy": 0.1434145525097847,
      "num_tokens": 8788981.0,
      "step": 4015
    },
    {
      "entropy": 5.7446869850158695,
      "epoch": 0.39896784438269156,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498983691710063,
      "loss": 5.6182,
      "mean_token_accuracy": 0.14746309965848922,
      "num_tokens": 8799190.0,
      "step": 4020
    },
    {
      "entropy": 5.7633524417877195,
      "epoch": 0.3994640730448591,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498980325095558,
      "loss": 5.5929,
      "mean_token_accuracy": 0.13670487627387046,
      "num_tokens": 8810030.0,
      "step": 4025
    },
    {
      "entropy": 5.694856023788452,
      "epoch": 0.3999603017070266,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989769529268238,
      "loss": 5.5992,
      "mean_token_accuracy": 0.13823745623230935,
      "num_tokens": 8820715.0,
      "step": 4030
    },
    {
      "entropy": 5.776308155059814,
      "epoch": 0.4004565303691941,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989735752039436,
      "loss": 5.6101,
      "mean_token_accuracy": 0.13810632526874542,
      "num_tokens": 8830484.0,
      "step": 4035
    },
    {
      "entropy": 5.774669456481933,
      "epoch": 0.4009527590313616,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989701919270015,
      "loss": 5.6137,
      "mean_token_accuracy": 0.1400647670030594,
      "num_tokens": 8840732.0,
      "step": 4040
    },
    {
      "entropy": 5.767325305938721,
      "epoch": 0.40144898769352916,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004989668030960812,
      "loss": 5.6311,
      "mean_token_accuracy": 0.13918099254369737,
      "num_tokens": 8853334.0,
      "step": 4045
    },
    {
      "entropy": 5.777545690536499,
      "epoch": 0.4019452163556967,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989634087112667,
      "loss": 5.6421,
      "mean_token_accuracy": 0.14060635939240457,
      "num_tokens": 8864129.0,
      "step": 4050
    },
    {
      "entropy": 5.679741430282593,
      "epoch": 0.4024414450178642,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989600087726425,
      "loss": 5.5977,
      "mean_token_accuracy": 0.13984159380197525,
      "num_tokens": 8875353.0,
      "step": 4055
    },
    {
      "entropy": 5.7884073734283445,
      "epoch": 0.40293767368003175,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989566032802924,
      "loss": 5.6555,
      "mean_token_accuracy": 0.14024916887283326,
      "num_tokens": 8886933.0,
      "step": 4060
    },
    {
      "entropy": 5.774551105499268,
      "epoch": 0.4034339023421993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989531922343013,
      "loss": 5.6171,
      "mean_token_accuracy": 0.13920446932315828,
      "num_tokens": 8898465.0,
      "step": 4065
    },
    {
      "entropy": 5.688659954071045,
      "epoch": 0.4039301310043668,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989497756347534,
      "loss": 5.5442,
      "mean_token_accuracy": 0.1478254973888397,
      "num_tokens": 8909072.0,
      "step": 4070
    },
    {
      "entropy": 5.735241317749024,
      "epoch": 0.40442635966653434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989463534817339,
      "loss": 5.6269,
      "mean_token_accuracy": 0.14025538563728332,
      "num_tokens": 8918841.0,
      "step": 4075
    },
    {
      "entropy": 5.8261096477508545,
      "epoch": 0.40492258832870187,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989429257753273,
      "loss": 5.6731,
      "mean_token_accuracy": 0.13695992827415465,
      "num_tokens": 8929718.0,
      "step": 4080
    },
    {
      "entropy": 5.639446258544922,
      "epoch": 0.4054188169908694,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989394925156187,
      "loss": 5.5564,
      "mean_token_accuracy": 0.14727575927972794,
      "num_tokens": 8938737.0,
      "step": 4085
    },
    {
      "entropy": 5.777335691452026,
      "epoch": 0.40591504565303693,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989360537026932,
      "loss": 5.6863,
      "mean_token_accuracy": 0.14337181970477103,
      "num_tokens": 8949121.0,
      "step": 4090
    },
    {
      "entropy": 5.73737473487854,
      "epoch": 0.40641127431520446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498932609336636,
      "loss": 5.5279,
      "mean_token_accuracy": 0.1430393636226654,
      "num_tokens": 8958585.0,
      "step": 4095
    },
    {
      "entropy": 5.743276834487915,
      "epoch": 0.406907502977372,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989291594175327,
      "loss": 5.5744,
      "mean_token_accuracy": 0.14032098054885864,
      "num_tokens": 8968528.0,
      "step": 4100
    },
    {
      "entropy": 5.726383113861084,
      "epoch": 0.4074037316395395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989257039454689,
      "loss": 5.6589,
      "mean_token_accuracy": 0.1403864450752735,
      "num_tokens": 8980870.0,
      "step": 4105
    },
    {
      "entropy": 5.73947114944458,
      "epoch": 0.40789996030170705,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049892224292053,
      "loss": 5.6037,
      "mean_token_accuracy": 0.1405671626329422,
      "num_tokens": 8992147.0,
      "step": 4110
    },
    {
      "entropy": 5.745876502990723,
      "epoch": 0.40839618896387453,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989187763428021,
      "loss": 5.5912,
      "mean_token_accuracy": 0.14025968909263611,
      "num_tokens": 9003768.0,
      "step": 4115
    },
    {
      "entropy": 5.739760398864746,
      "epoch": 0.40889241762604206,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498915304212371,
      "loss": 5.6987,
      "mean_token_accuracy": 0.13496260195970536,
      "num_tokens": 9017259.0,
      "step": 4120
    },
    {
      "entropy": 5.71677885055542,
      "epoch": 0.4093886462882096,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989118265293228,
      "loss": 5.596,
      "mean_token_accuracy": 0.14214176386594773,
      "num_tokens": 9028696.0,
      "step": 4125
    },
    {
      "entropy": 5.7507734298706055,
      "epoch": 0.4098848749503771,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989083432937439,
      "loss": 5.6122,
      "mean_token_accuracy": 0.13986908346414567,
      "num_tokens": 9040231.0,
      "step": 4130
    },
    {
      "entropy": 5.655962562561035,
      "epoch": 0.41038110361254465,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989048545057204,
      "loss": 5.4839,
      "mean_token_accuracy": 0.14580330699682237,
      "num_tokens": 9050254.0,
      "step": 4135
    },
    {
      "entropy": 5.686241626739502,
      "epoch": 0.4108773322747122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989013601653391,
      "loss": 5.6307,
      "mean_token_accuracy": 0.13935362100601195,
      "num_tokens": 9061046.0,
      "step": 4140
    },
    {
      "entropy": 5.806539058685303,
      "epoch": 0.4113735609368797,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988978602726866,
      "loss": 5.6312,
      "mean_token_accuracy": 0.13362794667482375,
      "num_tokens": 9071899.0,
      "step": 4145
    },
    {
      "entropy": 5.746391820907593,
      "epoch": 0.41186978959904724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988943548278494,
      "loss": 5.6255,
      "mean_token_accuracy": 0.1440863624215126,
      "num_tokens": 9081781.0,
      "step": 4150
    },
    {
      "entropy": 5.713353157043457,
      "epoch": 0.4123660182612148,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988908438309149,
      "loss": 5.6143,
      "mean_token_accuracy": 0.14353400766849517,
      "num_tokens": 9092284.0,
      "step": 4155
    },
    {
      "entropy": 5.755454444885254,
      "epoch": 0.4128622469233823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988873272819697,
      "loss": 5.6028,
      "mean_token_accuracy": 0.1370057925581932,
      "num_tokens": 9102031.0,
      "step": 4160
    },
    {
      "entropy": 5.76092791557312,
      "epoch": 0.41335847558554983,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988838051811014,
      "loss": 5.592,
      "mean_token_accuracy": 0.14244649186730385,
      "num_tokens": 9113312.0,
      "step": 4165
    },
    {
      "entropy": 5.691739225387574,
      "epoch": 0.41385470424771736,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498880277528397,
      "loss": 5.6391,
      "mean_token_accuracy": 0.1454815074801445,
      "num_tokens": 9124111.0,
      "step": 4170
    },
    {
      "entropy": 5.749805736541748,
      "epoch": 0.4143509329098849,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988767443239443,
      "loss": 5.5658,
      "mean_token_accuracy": 0.14652878195047378,
      "num_tokens": 9134463.0,
      "step": 4175
    },
    {
      "entropy": 5.838459539413452,
      "epoch": 0.4148471615720524,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004988732055678307,
      "loss": 5.7591,
      "mean_token_accuracy": 0.13751131072640418,
      "num_tokens": 9146064.0,
      "step": 4180
    },
    {
      "entropy": 5.73297085762024,
      "epoch": 0.41534339023421996,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498869661260144,
      "loss": 5.5391,
      "mean_token_accuracy": 0.1499718338251114,
      "num_tokens": 9156272.0,
      "step": 4185
    },
    {
      "entropy": 5.681269264221191,
      "epoch": 0.41583961889638743,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498866111400972,
      "loss": 5.5989,
      "mean_token_accuracy": 0.14205178022384643,
      "num_tokens": 9166477.0,
      "step": 4190
    },
    {
      "entropy": 5.810617733001709,
      "epoch": 0.41633584755855496,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988625559904029,
      "loss": 5.6155,
      "mean_token_accuracy": 0.13759971186518669,
      "num_tokens": 9178577.0,
      "step": 4195
    },
    {
      "entropy": 5.688449668884277,
      "epoch": 0.4168320762207225,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988589950285248,
      "loss": 5.5597,
      "mean_token_accuracy": 0.14190229624509812,
      "num_tokens": 9188020.0,
      "step": 4200
    },
    {
      "entropy": 5.7322094440460205,
      "epoch": 0.41732830488289,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988554285154259,
      "loss": 5.6584,
      "mean_token_accuracy": 0.1414238192141056,
      "num_tokens": 9198571.0,
      "step": 4205
    },
    {
      "entropy": 5.767828273773193,
      "epoch": 0.41782453354505755,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004988518564511948,
      "loss": 5.6234,
      "mean_token_accuracy": 0.1430214002728462,
      "num_tokens": 9210515.0,
      "step": 4210
    },
    {
      "entropy": 5.652394771575928,
      "epoch": 0.4183207622072251,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988482788359202,
      "loss": 5.4977,
      "mean_token_accuracy": 0.15405959039926528,
      "num_tokens": 9220756.0,
      "step": 4215
    },
    {
      "entropy": 5.7456409454345705,
      "epoch": 0.4188169908693926,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988446956696904,
      "loss": 5.5718,
      "mean_token_accuracy": 0.14581375122070311,
      "num_tokens": 9231713.0,
      "step": 4220
    },
    {
      "entropy": 5.779129076004028,
      "epoch": 0.41931321953156014,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988411069525947,
      "loss": 5.5938,
      "mean_token_accuracy": 0.13980774953961372,
      "num_tokens": 9242430.0,
      "step": 4225
    },
    {
      "entropy": 5.767920112609863,
      "epoch": 0.4198094481937277,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988375126847216,
      "loss": 5.7145,
      "mean_token_accuracy": 0.13447293937206267,
      "num_tokens": 9254202.0,
      "step": 4230
    },
    {
      "entropy": 5.702891445159912,
      "epoch": 0.4203056768558952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988339128661607,
      "loss": 5.6255,
      "mean_token_accuracy": 0.13794530928134918,
      "num_tokens": 9266065.0,
      "step": 4235
    },
    {
      "entropy": 5.764676427841186,
      "epoch": 0.42080190551806274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498830307497001,
      "loss": 5.5935,
      "mean_token_accuracy": 0.1467057980597019,
      "num_tokens": 9277022.0,
      "step": 4240
    },
    {
      "entropy": 5.767722940444946,
      "epoch": 0.42129813418023027,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498826696577332,
      "loss": 5.5964,
      "mean_token_accuracy": 0.1434573471546173,
      "num_tokens": 9287144.0,
      "step": 4245
    },
    {
      "entropy": 5.687915992736817,
      "epoch": 0.4217943628423978,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988230801072432,
      "loss": 5.5606,
      "mean_token_accuracy": 0.13668668568134307,
      "num_tokens": 9298559.0,
      "step": 4250
    },
    {
      "entropy": 5.696269369125366,
      "epoch": 0.4222905915045653,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988194580868244,
      "loss": 5.5627,
      "mean_token_accuracy": 0.1408204182982445,
      "num_tokens": 9309275.0,
      "step": 4255
    },
    {
      "entropy": 5.718939733505249,
      "epoch": 0.42278682016673286,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988158305161654,
      "loss": 5.5111,
      "mean_token_accuracy": 0.14722881466150284,
      "num_tokens": 9319442.0,
      "step": 4260
    },
    {
      "entropy": 5.632960462570191,
      "epoch": 0.42328304882890033,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498812197395356,
      "loss": 5.4946,
      "mean_token_accuracy": 0.15007592290639876,
      "num_tokens": 9329762.0,
      "step": 4265
    },
    {
      "entropy": 5.813821268081665,
      "epoch": 0.42377927749106786,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988085587244863,
      "loss": 5.7269,
      "mean_token_accuracy": 0.13791365101933478,
      "num_tokens": 9340417.0,
      "step": 4270
    },
    {
      "entropy": 5.80053334236145,
      "epoch": 0.4242755061532354,
      "grad_norm": 0.875,
      "learning_rate": 0.0004988049145036466,
      "loss": 5.5599,
      "mean_token_accuracy": 0.14606219977140428,
      "num_tokens": 9351538.0,
      "step": 4275
    },
    {
      "entropy": 5.613687324523926,
      "epoch": 0.4247717348154029,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988012647329273,
      "loss": 5.4993,
      "mean_token_accuracy": 0.15433092713356017,
      "num_tokens": 9360916.0,
      "step": 4280
    },
    {
      "entropy": 5.6723957538604735,
      "epoch": 0.42526796347757045,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498797609412419,
      "loss": 5.6178,
      "mean_token_accuracy": 0.14207035824656486,
      "num_tokens": 9372959.0,
      "step": 4285
    },
    {
      "entropy": 5.777748441696167,
      "epoch": 0.425764192139738,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987939485422122,
      "loss": 5.6252,
      "mean_token_accuracy": 0.14154006764292718,
      "num_tokens": 9383887.0,
      "step": 4290
    },
    {
      "entropy": 5.7544303894042965,
      "epoch": 0.4262604208019055,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004987902821223976,
      "loss": 5.6143,
      "mean_token_accuracy": 0.14055404886603357,
      "num_tokens": 9397735.0,
      "step": 4295
    },
    {
      "entropy": 5.644642162322998,
      "epoch": 0.42675664946407305,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987866101530663,
      "loss": 5.4468,
      "mean_token_accuracy": 0.15418429896235467,
      "num_tokens": 9407614.0,
      "step": 4300
    },
    {
      "entropy": 5.701617765426636,
      "epoch": 0.4272528781262406,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987829326343092,
      "loss": 5.5893,
      "mean_token_accuracy": 0.1373545914888382,
      "num_tokens": 9420077.0,
      "step": 4305
    },
    {
      "entropy": 5.695030784606933,
      "epoch": 0.4277491067884081,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987792495662177,
      "loss": 5.543,
      "mean_token_accuracy": 0.14468438029289246,
      "num_tokens": 9431719.0,
      "step": 4310
    },
    {
      "entropy": 5.7009850025177,
      "epoch": 0.42824533545057564,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498775560948883,
      "loss": 5.5228,
      "mean_token_accuracy": 0.14737678915262223,
      "num_tokens": 9441462.0,
      "step": 4315
    },
    {
      "entropy": 5.6392723560333256,
      "epoch": 0.42874156411274317,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987718667823965,
      "loss": 5.538,
      "mean_token_accuracy": 0.14649159610271453,
      "num_tokens": 9451968.0,
      "step": 4320
    },
    {
      "entropy": 5.720705366134643,
      "epoch": 0.4292377927749107,
      "grad_norm": 0.8125,
      "learning_rate": 0.00049876816706685,
      "loss": 5.5856,
      "mean_token_accuracy": 0.14179600328207015,
      "num_tokens": 9463643.0,
      "step": 4325
    },
    {
      "entropy": 5.790055942535401,
      "epoch": 0.42973402143707823,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498764461802335,
      "loss": 5.6283,
      "mean_token_accuracy": 0.14421869218349456,
      "num_tokens": 9476067.0,
      "step": 4330
    },
    {
      "entropy": 5.67179913520813,
      "epoch": 0.43023025009924576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987607509889437,
      "loss": 5.4631,
      "mean_token_accuracy": 0.15186649858951567,
      "num_tokens": 9485697.0,
      "step": 4335
    },
    {
      "entropy": 5.75130877494812,
      "epoch": 0.43072647876141323,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987570346267677,
      "loss": 5.6721,
      "mean_token_accuracy": 0.14101483449339866,
      "num_tokens": 9496104.0,
      "step": 4340
    },
    {
      "entropy": 5.809075260162354,
      "epoch": 0.43122270742358076,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987533127158996,
      "loss": 5.6563,
      "mean_token_accuracy": 0.13856788575649262,
      "num_tokens": 9507391.0,
      "step": 4345
    },
    {
      "entropy": 5.670343112945557,
      "epoch": 0.4317189360857483,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004987495852564314,
      "loss": 5.6039,
      "mean_token_accuracy": 0.1471710152924061,
      "num_tokens": 9518713.0,
      "step": 4350
    },
    {
      "entropy": 5.716785573959351,
      "epoch": 0.4322151647479158,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987458522484556,
      "loss": 5.5771,
      "mean_token_accuracy": 0.14671115949749947,
      "num_tokens": 9529930.0,
      "step": 4355
    },
    {
      "entropy": 5.644144868850708,
      "epoch": 0.43271139341008336,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987421136920648,
      "loss": 5.4758,
      "mean_token_accuracy": 0.15136663094162942,
      "num_tokens": 9539455.0,
      "step": 4360
    },
    {
      "entropy": 5.728163576126098,
      "epoch": 0.4332076220722509,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004987383695873517,
      "loss": 5.6137,
      "mean_token_accuracy": 0.14231019467115402,
      "num_tokens": 9551197.0,
      "step": 4365
    },
    {
      "entropy": 5.69749493598938,
      "epoch": 0.4337038507344184,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987346199344092,
      "loss": 5.5491,
      "mean_token_accuracy": 0.14035396799445152,
      "num_tokens": 9561240.0,
      "step": 4370
    },
    {
      "entropy": 5.734319019317627,
      "epoch": 0.43420007939658595,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987308647333303,
      "loss": 5.6372,
      "mean_token_accuracy": 0.1377876028418541,
      "num_tokens": 9572876.0,
      "step": 4375
    },
    {
      "entropy": 5.768622255325317,
      "epoch": 0.4346963080587535,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987271039842079,
      "loss": 5.5665,
      "mean_token_accuracy": 0.1467779129743576,
      "num_tokens": 9583400.0,
      "step": 4380
    },
    {
      "entropy": 5.673107862472534,
      "epoch": 0.435192536720921,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987233376871355,
      "loss": 5.493,
      "mean_token_accuracy": 0.14794401377439498,
      "num_tokens": 9594637.0,
      "step": 4385
    },
    {
      "entropy": 5.6723779201507565,
      "epoch": 0.43568876538308854,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987195658422063,
      "loss": 5.647,
      "mean_token_accuracy": 0.14258133322000505,
      "num_tokens": 9605564.0,
      "step": 4390
    },
    {
      "entropy": 5.743455123901367,
      "epoch": 0.43618499404525607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987157884495141,
      "loss": 5.5753,
      "mean_token_accuracy": 0.13976100087165833,
      "num_tokens": 9615601.0,
      "step": 4395
    },
    {
      "entropy": 5.653528308868408,
      "epoch": 0.4366812227074236,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987120055091523,
      "loss": 5.4955,
      "mean_token_accuracy": 0.1433832623064518,
      "num_tokens": 9626182.0,
      "step": 4400
    },
    {
      "entropy": 5.769513893127441,
      "epoch": 0.43717745136959113,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987082170212149,
      "loss": 5.6227,
      "mean_token_accuracy": 0.14208893850445747,
      "num_tokens": 9636290.0,
      "step": 4405
    },
    {
      "entropy": 5.696675395965576,
      "epoch": 0.43767368003175866,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987044229857957,
      "loss": 5.5269,
      "mean_token_accuracy": 0.1449810490012169,
      "num_tokens": 9648891.0,
      "step": 4410
    },
    {
      "entropy": 5.6389124393463135,
      "epoch": 0.43816990869392614,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004987006234029889,
      "loss": 5.6396,
      "mean_token_accuracy": 0.14224111512303353,
      "num_tokens": 9662452.0,
      "step": 4415
    },
    {
      "entropy": 5.661336040496826,
      "epoch": 0.43866613735609367,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986968182728885,
      "loss": 5.4873,
      "mean_token_accuracy": 0.14389461725950242,
      "num_tokens": 9672685.0,
      "step": 4420
    },
    {
      "entropy": 5.798376560211182,
      "epoch": 0.4391623660182612,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986930075955892,
      "loss": 5.6315,
      "mean_token_accuracy": 0.13783416748046876,
      "num_tokens": 9683904.0,
      "step": 4425
    },
    {
      "entropy": 5.704138517379761,
      "epoch": 0.4396585946804287,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986891913711853,
      "loss": 5.5142,
      "mean_token_accuracy": 0.1462846115231514,
      "num_tokens": 9695463.0,
      "step": 4430
    },
    {
      "entropy": 5.706958341598511,
      "epoch": 0.44015482334259626,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004986853695997715,
      "loss": 5.6094,
      "mean_token_accuracy": 0.1411213308572769,
      "num_tokens": 9706922.0,
      "step": 4435
    },
    {
      "entropy": 5.742683601379395,
      "epoch": 0.4406510520047638,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986815422814425,
      "loss": 5.6024,
      "mean_token_accuracy": 0.14004263654351234,
      "num_tokens": 9716591.0,
      "step": 4440
    },
    {
      "entropy": 5.763763570785523,
      "epoch": 0.4411472806669313,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986777094162932,
      "loss": 5.6597,
      "mean_token_accuracy": 0.1410846993327141,
      "num_tokens": 9727745.0,
      "step": 4445
    },
    {
      "entropy": 5.759910154342651,
      "epoch": 0.44164350932909885,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986738710044187,
      "loss": 5.6089,
      "mean_token_accuracy": 0.14154430478811264,
      "num_tokens": 9738663.0,
      "step": 4450
    },
    {
      "entropy": 5.648026704788208,
      "epoch": 0.4421397379912664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986700270459142,
      "loss": 5.476,
      "mean_token_accuracy": 0.1486654907464981,
      "num_tokens": 9749815.0,
      "step": 4455
    },
    {
      "entropy": 5.76146388053894,
      "epoch": 0.4426359666534339,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986661775408749,
      "loss": 5.5941,
      "mean_token_accuracy": 0.144724602997303,
      "num_tokens": 9760038.0,
      "step": 4460
    },
    {
      "entropy": 5.700668859481811,
      "epoch": 0.44313219531560144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986623224893965,
      "loss": 5.5852,
      "mean_token_accuracy": 0.14230740293860436,
      "num_tokens": 9770322.0,
      "step": 4465
    },
    {
      "entropy": 5.733216238021851,
      "epoch": 0.44362842397776897,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004986584618915744,
      "loss": 5.5823,
      "mean_token_accuracy": 0.14704113453626633,
      "num_tokens": 9781649.0,
      "step": 4470
    },
    {
      "entropy": 5.686500310897827,
      "epoch": 0.4441246526399365,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986545957475045,
      "loss": 5.5773,
      "mean_token_accuracy": 0.14150618463754655,
      "num_tokens": 9793146.0,
      "step": 4475
    },
    {
      "entropy": 5.739275360107422,
      "epoch": 0.44462088130210403,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986507240572825,
      "loss": 5.5114,
      "mean_token_accuracy": 0.144402676820755,
      "num_tokens": 9804645.0,
      "step": 4480
    },
    {
      "entropy": 5.685995817184448,
      "epoch": 0.44511710996427156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986468468210042,
      "loss": 5.5331,
      "mean_token_accuracy": 0.14840690791606903,
      "num_tokens": 9815589.0,
      "step": 4485
    },
    {
      "entropy": 5.611461114883423,
      "epoch": 0.44561333862643904,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986429640387663,
      "loss": 5.4695,
      "mean_token_accuracy": 0.1494643822312355,
      "num_tokens": 9826064.0,
      "step": 4490
    },
    {
      "entropy": 5.598761463165284,
      "epoch": 0.44610956728860657,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986390757106648,
      "loss": 5.5496,
      "mean_token_accuracy": 0.14843137860298156,
      "num_tokens": 9835682.0,
      "step": 4495
    },
    {
      "entropy": 5.694795370101929,
      "epoch": 0.4466057959507741,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986351818367959,
      "loss": 5.5743,
      "mean_token_accuracy": 0.14617726653814317,
      "num_tokens": 9846445.0,
      "step": 4500
    },
    {
      "entropy": 5.71155276298523,
      "epoch": 0.44710202461294163,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986312824172566,
      "loss": 5.5656,
      "mean_token_accuracy": 0.14476391151547433,
      "num_tokens": 9856537.0,
      "step": 4505
    },
    {
      "entropy": 5.81048173904419,
      "epoch": 0.44759825327510916,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986273774521433,
      "loss": 5.6567,
      "mean_token_accuracy": 0.13928472027182578,
      "num_tokens": 9869464.0,
      "step": 4510
    },
    {
      "entropy": 5.646250629425049,
      "epoch": 0.4480944819372767,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986234669415529,
      "loss": 5.4809,
      "mean_token_accuracy": 0.1462387427687645,
      "num_tokens": 9879896.0,
      "step": 4515
    },
    {
      "entropy": 5.673519086837769,
      "epoch": 0.4485907105994442,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004986195508855824,
      "loss": 5.5553,
      "mean_token_accuracy": 0.14665139093995094,
      "num_tokens": 9890261.0,
      "step": 4520
    },
    {
      "entropy": 5.744044971466065,
      "epoch": 0.44908693926161175,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498615629284329,
      "loss": 5.669,
      "mean_token_accuracy": 0.13816292807459832,
      "num_tokens": 9903447.0,
      "step": 4525
    },
    {
      "entropy": 5.746433544158935,
      "epoch": 0.4495831679237793,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004986117021378896,
      "loss": 5.5197,
      "mean_token_accuracy": 0.14785393327474594,
      "num_tokens": 9913665.0,
      "step": 4530
    },
    {
      "entropy": 5.731913995742798,
      "epoch": 0.4500793965859468,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000498607769446362,
      "loss": 5.5964,
      "mean_token_accuracy": 0.14285696744918824,
      "num_tokens": 9926313.0,
      "step": 4535
    },
    {
      "entropy": 5.703529024124146,
      "epoch": 0.45057562524811434,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986038312098435,
      "loss": 5.5404,
      "mean_token_accuracy": 0.14697111994028092,
      "num_tokens": 9936797.0,
      "step": 4540
    },
    {
      "entropy": 5.70205454826355,
      "epoch": 0.4510718539102819,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985998874284319,
      "loss": 5.5547,
      "mean_token_accuracy": 0.14567258358001708,
      "num_tokens": 9948623.0,
      "step": 4545
    },
    {
      "entropy": 5.678313684463501,
      "epoch": 0.4515680825724494,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985959381022248,
      "loss": 5.565,
      "mean_token_accuracy": 0.14417936652898788,
      "num_tokens": 9959097.0,
      "step": 4550
    },
    {
      "entropy": 5.7207825660705565,
      "epoch": 0.45206431123461693,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985919832313203,
      "loss": 5.6138,
      "mean_token_accuracy": 0.14531633555889129,
      "num_tokens": 9969098.0,
      "step": 4555
    },
    {
      "entropy": 5.703067064285278,
      "epoch": 0.45256053989678446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985880228158164,
      "loss": 5.5263,
      "mean_token_accuracy": 0.14786665961146356,
      "num_tokens": 9979427.0,
      "step": 4560
    },
    {
      "entropy": 5.689963865280151,
      "epoch": 0.45305676855895194,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985840568558112,
      "loss": 5.6794,
      "mean_token_accuracy": 0.14244562163949012,
      "num_tokens": 9990543.0,
      "step": 4565
    },
    {
      "entropy": 5.708375072479248,
      "epoch": 0.45355299722111947,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985800853514033,
      "loss": 5.572,
      "mean_token_accuracy": 0.13732469230890273,
      "num_tokens": 10001119.0,
      "step": 4570
    },
    {
      "entropy": 5.7897083282470705,
      "epoch": 0.454049225883287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985761083026911,
      "loss": 5.656,
      "mean_token_accuracy": 0.14233326613903047,
      "num_tokens": 10011721.0,
      "step": 4575
    },
    {
      "entropy": 5.726434135437012,
      "epoch": 0.45454545454545453,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985721257097731,
      "loss": 5.5743,
      "mean_token_accuracy": 0.14098384454846383,
      "num_tokens": 10025463.0,
      "step": 4580
    },
    {
      "entropy": 5.8132609844207765,
      "epoch": 0.45504168320762206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985681375727482,
      "loss": 5.7168,
      "mean_token_accuracy": 0.1391293689608574,
      "num_tokens": 10036250.0,
      "step": 4585
    },
    {
      "entropy": 5.745821142196656,
      "epoch": 0.4555379118697896,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004985641438917151,
      "loss": 5.5593,
      "mean_token_accuracy": 0.14779864326119424,
      "num_tokens": 10048344.0,
      "step": 4590
    },
    {
      "entropy": 5.774887943267823,
      "epoch": 0.4560341405319571,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985601446667731,
      "loss": 5.6213,
      "mean_token_accuracy": 0.13995709121227265,
      "num_tokens": 10059798.0,
      "step": 4595
    },
    {
      "entropy": 5.658098649978638,
      "epoch": 0.45653036919412465,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985561398980213,
      "loss": 5.5194,
      "mean_token_accuracy": 0.14487059712409972,
      "num_tokens": 10069905.0,
      "step": 4600
    },
    {
      "entropy": 5.645683002471924,
      "epoch": 0.4570265978562922,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985521295855588,
      "loss": 5.5158,
      "mean_token_accuracy": 0.1443955421447754,
      "num_tokens": 10081226.0,
      "step": 4605
    },
    {
      "entropy": 5.76915979385376,
      "epoch": 0.4575228265184597,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004985481137294853,
      "loss": 5.6261,
      "mean_token_accuracy": 0.14413290694355965,
      "num_tokens": 10092673.0,
      "step": 4610
    },
    {
      "entropy": 5.759901714324951,
      "epoch": 0.45801905518062724,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004985440923299002,
      "loss": 5.4896,
      "mean_token_accuracy": 0.14923679530620576,
      "num_tokens": 10103243.0,
      "step": 4615
    },
    {
      "entropy": 5.528211212158203,
      "epoch": 0.4585152838427948,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004985400653869034,
      "loss": 5.46,
      "mean_token_accuracy": 0.16036638915538787,
      "num_tokens": 10114936.0,
      "step": 4620
    },
    {
      "entropy": 5.631842374801636,
      "epoch": 0.4590115125049623,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985360329005945,
      "loss": 5.4564,
      "mean_token_accuracy": 0.15086652636528014,
      "num_tokens": 10125505.0,
      "step": 4625
    },
    {
      "entropy": 5.757396793365478,
      "epoch": 0.45950774116712984,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985319948710738,
      "loss": 5.6267,
      "mean_token_accuracy": 0.13638549596071242,
      "num_tokens": 10137032.0,
      "step": 4630
    },
    {
      "entropy": 5.669996023178101,
      "epoch": 0.46000396982929737,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985279512984413,
      "loss": 5.5791,
      "mean_token_accuracy": 0.14591586887836455,
      "num_tokens": 10147527.0,
      "step": 4635
    },
    {
      "entropy": 5.669521284103394,
      "epoch": 0.46050019849146484,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985239021827972,
      "loss": 5.5288,
      "mean_token_accuracy": 0.14494569450616837,
      "num_tokens": 10158848.0,
      "step": 4640
    },
    {
      "entropy": 5.707905673980713,
      "epoch": 0.46099642715363237,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498519847524242,
      "loss": 5.5997,
      "mean_token_accuracy": 0.14298787862062454,
      "num_tokens": 10169594.0,
      "step": 4645
    },
    {
      "entropy": 5.69180736541748,
      "epoch": 0.4614926558157999,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985157873228762,
      "loss": 5.5548,
      "mean_token_accuracy": 0.14860456734895705,
      "num_tokens": 10180843.0,
      "step": 4650
    },
    {
      "entropy": 5.695382595062256,
      "epoch": 0.46198888447796743,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004985117215788005,
      "loss": 5.6226,
      "mean_token_accuracy": 0.14679727405309678,
      "num_tokens": 10193226.0,
      "step": 4655
    },
    {
      "entropy": 5.685736560821534,
      "epoch": 0.46248511314013496,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985076502921158,
      "loss": 5.4746,
      "mean_token_accuracy": 0.1473032832145691,
      "num_tokens": 10203023.0,
      "step": 4660
    },
    {
      "entropy": 5.699276399612427,
      "epoch": 0.4629813418023025,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498503573462923,
      "loss": 5.5135,
      "mean_token_accuracy": 0.14992617070674896,
      "num_tokens": 10214814.0,
      "step": 4665
    },
    {
      "entropy": 5.622752380371094,
      "epoch": 0.46347757046447,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984994910913232,
      "loss": 5.5232,
      "mean_token_accuracy": 0.14511056244373322,
      "num_tokens": 10224961.0,
      "step": 4670
    },
    {
      "entropy": 5.636011457443237,
      "epoch": 0.46397379912663755,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984954031774175,
      "loss": 5.4807,
      "mean_token_accuracy": 0.15218631625175477,
      "num_tokens": 10235923.0,
      "step": 4675
    },
    {
      "entropy": 5.681984519958496,
      "epoch": 0.4644700277888051,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984913097213075,
      "loss": 5.5932,
      "mean_token_accuracy": 0.1420750930905342,
      "num_tokens": 10246065.0,
      "step": 4680
    },
    {
      "entropy": 5.741244649887085,
      "epoch": 0.4649662564509726,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984872107230945,
      "loss": 5.4947,
      "mean_token_accuracy": 0.14651544988155366,
      "num_tokens": 10257085.0,
      "step": 4685
    },
    {
      "entropy": 5.755229854583741,
      "epoch": 0.46546248511314015,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984831061828803,
      "loss": 5.5751,
      "mean_token_accuracy": 0.1444445937871933,
      "num_tokens": 10267192.0,
      "step": 4690
    },
    {
      "entropy": 5.641280698776245,
      "epoch": 0.4659587137753077,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984789961007667,
      "loss": 5.5309,
      "mean_token_accuracy": 0.14865996837615966,
      "num_tokens": 10278580.0,
      "step": 4695
    },
    {
      "entropy": 5.705250406265259,
      "epoch": 0.4664549424374752,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984748804768554,
      "loss": 5.5564,
      "mean_token_accuracy": 0.14724098071455954,
      "num_tokens": 10290034.0,
      "step": 4700
    },
    {
      "entropy": 5.731843042373657,
      "epoch": 0.46695117109964274,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984707593112487,
      "loss": 5.5172,
      "mean_token_accuracy": 0.14083585292100906,
      "num_tokens": 10301279.0,
      "step": 4705
    },
    {
      "entropy": 5.674729061126709,
      "epoch": 0.46744739976181027,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984666326040487,
      "loss": 5.4771,
      "mean_token_accuracy": 0.15075616538524628,
      "num_tokens": 10311981.0,
      "step": 4710
    },
    {
      "entropy": 5.6466064453125,
      "epoch": 0.46794362842397774,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984625003553578,
      "loss": 5.5069,
      "mean_token_accuracy": 0.1453600063920021,
      "num_tokens": 10323315.0,
      "step": 4715
    },
    {
      "entropy": 5.701808929443359,
      "epoch": 0.4684398570861453,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004984583625652783,
      "loss": 5.5716,
      "mean_token_accuracy": 0.140718287974596,
      "num_tokens": 10333736.0,
      "step": 4720
    },
    {
      "entropy": 5.719180154800415,
      "epoch": 0.4689360857483128,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498454219233913,
      "loss": 5.6317,
      "mean_token_accuracy": 0.14207804650068284,
      "num_tokens": 10344736.0,
      "step": 4725
    },
    {
      "entropy": 5.759516382217408,
      "epoch": 0.46943231441048033,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984500703613645,
      "loss": 5.5763,
      "mean_token_accuracy": 0.1392699249088764,
      "num_tokens": 10355914.0,
      "step": 4730
    },
    {
      "entropy": 5.668118715286255,
      "epoch": 0.46992854307264786,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984459159477358,
      "loss": 5.4785,
      "mean_token_accuracy": 0.1454298198223114,
      "num_tokens": 10366748.0,
      "step": 4735
    },
    {
      "entropy": 5.671790218353271,
      "epoch": 0.4704247717348154,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984417559931298,
      "loss": 5.6053,
      "mean_token_accuracy": 0.14055315181612968,
      "num_tokens": 10378993.0,
      "step": 4740
    },
    {
      "entropy": 5.689825773239136,
      "epoch": 0.4709210003969829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984375904976498,
      "loss": 5.487,
      "mean_token_accuracy": 0.14609419852495192,
      "num_tokens": 10388813.0,
      "step": 4745
    },
    {
      "entropy": 5.676311349868774,
      "epoch": 0.47141722905915046,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984334194613989,
      "loss": 5.4789,
      "mean_token_accuracy": 0.14860565662384034,
      "num_tokens": 10399264.0,
      "step": 4750
    },
    {
      "entropy": 5.61346321105957,
      "epoch": 0.471913457721318,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984292428844809,
      "loss": 5.469,
      "mean_token_accuracy": 0.14876034259796142,
      "num_tokens": 10408580.0,
      "step": 4755
    },
    {
      "entropy": 5.5821356773376465,
      "epoch": 0.4724096863834855,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984250607669989,
      "loss": 5.4377,
      "mean_token_accuracy": 0.14673529118299483,
      "num_tokens": 10418246.0,
      "step": 4760
    },
    {
      "entropy": 5.685560178756714,
      "epoch": 0.47290591504565305,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984208731090569,
      "loss": 5.487,
      "mean_token_accuracy": 0.15245441794395448,
      "num_tokens": 10428068.0,
      "step": 4765
    },
    {
      "entropy": 5.624057722091675,
      "epoch": 0.4734021437078206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984166799107587,
      "loss": 5.5277,
      "mean_token_accuracy": 0.14296067208051683,
      "num_tokens": 10438402.0,
      "step": 4770
    },
    {
      "entropy": 5.662654256820678,
      "epoch": 0.4738983723699881,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984124811722081,
      "loss": 5.5162,
      "mean_token_accuracy": 0.14940229058265686,
      "num_tokens": 10449280.0,
      "step": 4775
    },
    {
      "entropy": 5.714998626708985,
      "epoch": 0.47439460103215564,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984082768935096,
      "loss": 5.6328,
      "mean_token_accuracy": 0.14610829949378967,
      "num_tokens": 10460688.0,
      "step": 4780
    },
    {
      "entropy": 5.666269540786743,
      "epoch": 0.47489082969432317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984040670747672,
      "loss": 5.5279,
      "mean_token_accuracy": 0.14058421105146407,
      "num_tokens": 10471652.0,
      "step": 4785
    },
    {
      "entropy": 5.599174976348877,
      "epoch": 0.47538705835649064,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983998517160852,
      "loss": 5.4922,
      "mean_token_accuracy": 0.1436099410057068,
      "num_tokens": 10481907.0,
      "step": 4790
    },
    {
      "entropy": 5.729784965515137,
      "epoch": 0.4758832870186582,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004983956308175684,
      "loss": 5.5387,
      "mean_token_accuracy": 0.14547334015369415,
      "num_tokens": 10494490.0,
      "step": 4795
    },
    {
      "entropy": 5.620134162902832,
      "epoch": 0.4763795156808257,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983914043793212,
      "loss": 5.4653,
      "mean_token_accuracy": 0.1539872482419014,
      "num_tokens": 10505331.0,
      "step": 4800
    },
    {
      "entropy": 5.712967491149902,
      "epoch": 0.47687574434299324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983871724014487,
      "loss": 5.612,
      "mean_token_accuracy": 0.13693344444036484,
      "num_tokens": 10516282.0,
      "step": 4805
    },
    {
      "entropy": 5.664428806304931,
      "epoch": 0.47737197300516077,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983829348840556,
      "loss": 5.4365,
      "mean_token_accuracy": 0.1529388427734375,
      "num_tokens": 10527454.0,
      "step": 4810
    },
    {
      "entropy": 5.63366436958313,
      "epoch": 0.4778682016673283,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498378691827247,
      "loss": 5.5222,
      "mean_token_accuracy": 0.14255161061882973,
      "num_tokens": 10537609.0,
      "step": 4815
    },
    {
      "entropy": 5.648998689651489,
      "epoch": 0.4783644303294958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983744432311282,
      "loss": 5.4923,
      "mean_token_accuracy": 0.14607268571853638,
      "num_tokens": 10548875.0,
      "step": 4820
    },
    {
      "entropy": 5.6902382373809814,
      "epoch": 0.47886065899166336,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983701890958046,
      "loss": 5.4543,
      "mean_token_accuracy": 0.15149110779166222,
      "num_tokens": 10559778.0,
      "step": 4825
    },
    {
      "entropy": 5.604547739028931,
      "epoch": 0.4793568876538309,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983659294213815,
      "loss": 5.4863,
      "mean_token_accuracy": 0.14867570102214814,
      "num_tokens": 10571517.0,
      "step": 4830
    },
    {
      "entropy": 5.6619298458099365,
      "epoch": 0.4798531163159984,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983616642079649,
      "loss": 5.4662,
      "mean_token_accuracy": 0.14753953516483306,
      "num_tokens": 10582209.0,
      "step": 4835
    },
    {
      "entropy": 5.641030645370483,
      "epoch": 0.48034934497816595,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983573934556601,
      "loss": 5.436,
      "mean_token_accuracy": 0.15206027925014495,
      "num_tokens": 10592164.0,
      "step": 4840
    },
    {
      "entropy": 5.583185815811158,
      "epoch": 0.4808455736403335,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983531171645734,
      "loss": 5.4405,
      "mean_token_accuracy": 0.15364081263542176,
      "num_tokens": 10602175.0,
      "step": 4845
    },
    {
      "entropy": 5.711642169952393,
      "epoch": 0.481341802302501,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983488353348106,
      "loss": 5.4162,
      "mean_token_accuracy": 0.15797716081142427,
      "num_tokens": 10612808.0,
      "step": 4850
    },
    {
      "entropy": 5.677546453475952,
      "epoch": 0.48183803096466854,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983445479664779,
      "loss": 5.5385,
      "mean_token_accuracy": 0.14271382763981819,
      "num_tokens": 10622918.0,
      "step": 4855
    },
    {
      "entropy": 5.690401935577393,
      "epoch": 0.48233425962683607,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983402550596817,
      "loss": 5.602,
      "mean_token_accuracy": 0.14286706522107123,
      "num_tokens": 10634607.0,
      "step": 4860
    },
    {
      "entropy": 5.630452585220337,
      "epoch": 0.48283048828900355,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004983359566145285,
      "loss": 5.4818,
      "mean_token_accuracy": 0.1472335323691368,
      "num_tokens": 10644951.0,
      "step": 4865
    },
    {
      "entropy": 5.661438131332398,
      "epoch": 0.4833267169511711,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983316526311248,
      "loss": 5.524,
      "mean_token_accuracy": 0.14634662717580796,
      "num_tokens": 10655839.0,
      "step": 4870
    },
    {
      "entropy": 5.667823553085327,
      "epoch": 0.4838229456133386,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004983273431095773,
      "loss": 5.5461,
      "mean_token_accuracy": 0.14504371136426925,
      "num_tokens": 10667642.0,
      "step": 4875
    },
    {
      "entropy": 5.632654666900635,
      "epoch": 0.48431917427550614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983230280499929,
      "loss": 5.58,
      "mean_token_accuracy": 0.1438919983804226,
      "num_tokens": 10677870.0,
      "step": 4880
    },
    {
      "entropy": 5.6506288051605225,
      "epoch": 0.48481540293767367,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983187074524787,
      "loss": 5.4897,
      "mean_token_accuracy": 0.1483173742890358,
      "num_tokens": 10688547.0,
      "step": 4885
    },
    {
      "entropy": 5.674701833724976,
      "epoch": 0.4853116315998412,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983143813171418,
      "loss": 5.514,
      "mean_token_accuracy": 0.14478935301303864,
      "num_tokens": 10699565.0,
      "step": 4890
    },
    {
      "entropy": 5.697041797637939,
      "epoch": 0.48580786026200873,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004983100496440895,
      "loss": 5.6283,
      "mean_token_accuracy": 0.14346733689308167,
      "num_tokens": 10711611.0,
      "step": 4895
    },
    {
      "entropy": 5.649517774581909,
      "epoch": 0.48630408892417626,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498305712433429,
      "loss": 5.4593,
      "mean_token_accuracy": 0.14551541954278946,
      "num_tokens": 10722719.0,
      "step": 4900
    },
    {
      "entropy": 5.735891628265381,
      "epoch": 0.4868003175863438,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498301369685268,
      "loss": 5.576,
      "mean_token_accuracy": 0.14439354538917543,
      "num_tokens": 10735308.0,
      "step": 4905
    },
    {
      "entropy": 5.591550827026367,
      "epoch": 0.4872965462485113,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982970213997142,
      "loss": 5.4746,
      "mean_token_accuracy": 0.1539837121963501,
      "num_tokens": 10745849.0,
      "step": 4910
    },
    {
      "entropy": 5.5551293849945065,
      "epoch": 0.48779277491067885,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982926675768756,
      "loss": 5.4438,
      "mean_token_accuracy": 0.15486387461423873,
      "num_tokens": 10755702.0,
      "step": 4915
    },
    {
      "entropy": 5.710547971725464,
      "epoch": 0.4882890035728464,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004982883082168599,
      "loss": 5.5553,
      "mean_token_accuracy": 0.14071914777159691,
      "num_tokens": 10766586.0,
      "step": 4920
    },
    {
      "entropy": 5.66877007484436,
      "epoch": 0.4887852322350139,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982839433197752,
      "loss": 5.5925,
      "mean_token_accuracy": 0.14218895584344865,
      "num_tokens": 10778634.0,
      "step": 4925
    },
    {
      "entropy": 5.675059652328491,
      "epoch": 0.48928146089718144,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049827957288573,
      "loss": 5.5957,
      "mean_token_accuracy": 0.14439964443445205,
      "num_tokens": 10789807.0,
      "step": 4930
    },
    {
      "entropy": 5.673707056045532,
      "epoch": 0.48977768955934897,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982751969148324,
      "loss": 5.492,
      "mean_token_accuracy": 0.1484350860118866,
      "num_tokens": 10801481.0,
      "step": 4935
    },
    {
      "entropy": 5.682024383544922,
      "epoch": 0.49027391822151645,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004982708154071912,
      "loss": 5.5913,
      "mean_token_accuracy": 0.13909316658973694,
      "num_tokens": 10813787.0,
      "step": 4940
    },
    {
      "entropy": 5.601347398757935,
      "epoch": 0.490770146883684,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982664283629147,
      "loss": 5.486,
      "mean_token_accuracy": 0.1466942921280861,
      "num_tokens": 10824698.0,
      "step": 4945
    },
    {
      "entropy": 5.646316432952881,
      "epoch": 0.4912663755458515,
      "grad_norm": 1.125,
      "learning_rate": 0.0004982620357821118,
      "loss": 5.4917,
      "mean_token_accuracy": 0.1437968857586384,
      "num_tokens": 10836308.0,
      "step": 4950
    },
    {
      "entropy": 5.709817743301391,
      "epoch": 0.49176260420801904,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982576376648918,
      "loss": 5.474,
      "mean_token_accuracy": 0.1472250908613205,
      "num_tokens": 10846844.0,
      "step": 4955
    },
    {
      "entropy": 5.608026838302612,
      "epoch": 0.49225883287018657,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982532340113633,
      "loss": 5.3713,
      "mean_token_accuracy": 0.15622817128896713,
      "num_tokens": 10857097.0,
      "step": 4960
    },
    {
      "entropy": 5.505818176269531,
      "epoch": 0.4927550615323541,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982488248216357,
      "loss": 5.3874,
      "mean_token_accuracy": 0.15339502841234207,
      "num_tokens": 10868005.0,
      "step": 4965
    },
    {
      "entropy": 5.636719417572022,
      "epoch": 0.49325129019452163,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982444100958182,
      "loss": 5.4806,
      "mean_token_accuracy": 0.1524288535118103,
      "num_tokens": 10878708.0,
      "step": 4970
    },
    {
      "entropy": 5.710105848312378,
      "epoch": 0.49374751885668916,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982399898340205,
      "loss": 5.4872,
      "mean_token_accuracy": 0.15132090598344802,
      "num_tokens": 10889340.0,
      "step": 4975
    },
    {
      "entropy": 5.6360618591308596,
      "epoch": 0.4942437475188567,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498235564036352,
      "loss": 5.5277,
      "mean_token_accuracy": 0.14438705891370773,
      "num_tokens": 10901182.0,
      "step": 4980
    },
    {
      "entropy": 5.627745485305786,
      "epoch": 0.4947399761810242,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982311327029226,
      "loss": 5.5047,
      "mean_token_accuracy": 0.14128261804580688,
      "num_tokens": 10911898.0,
      "step": 4985
    },
    {
      "entropy": 5.600978088378906,
      "epoch": 0.49523620484319175,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982266958338421,
      "loss": 5.5015,
      "mean_token_accuracy": 0.14600002095103265,
      "num_tokens": 10922848.0,
      "step": 4990
    },
    {
      "entropy": 5.60031361579895,
      "epoch": 0.4957324335053593,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982222534292205,
      "loss": 5.4152,
      "mean_token_accuracy": 0.15231944024562835,
      "num_tokens": 10933252.0,
      "step": 4995
    },
    {
      "entropy": 5.653515481948853,
      "epoch": 0.4962286621675268,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982178054891681,
      "loss": 5.5287,
      "mean_token_accuracy": 0.14584684669971465,
      "num_tokens": 10944240.0,
      "step": 5000
    },
    {
      "entropy": 5.676952075958252,
      "epoch": 0.49672489082969434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982133520137951,
      "loss": 5.5048,
      "mean_token_accuracy": 0.1452569544315338,
      "num_tokens": 10954233.0,
      "step": 5005
    },
    {
      "entropy": 5.78747444152832,
      "epoch": 0.4972211194918619,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982088930032119,
      "loss": 5.5642,
      "mean_token_accuracy": 0.14112753197550773,
      "num_tokens": 10965252.0,
      "step": 5010
    },
    {
      "entropy": 5.739939165115357,
      "epoch": 0.4977173481540294,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982044284575293,
      "loss": 5.565,
      "mean_token_accuracy": 0.1478274330496788,
      "num_tokens": 10974362.0,
      "step": 5015
    },
    {
      "entropy": 5.661859178543091,
      "epoch": 0.4982135768161969,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981999583768576,
      "loss": 5.6835,
      "mean_token_accuracy": 0.14331025034189224,
      "num_tokens": 10987481.0,
      "step": 5020
    },
    {
      "entropy": 5.717976188659668,
      "epoch": 0.4987098054783644,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498195482761308,
      "loss": 5.4934,
      "mean_token_accuracy": 0.1482692539691925,
      "num_tokens": 10997987.0,
      "step": 5025
    },
    {
      "entropy": 5.705264520645142,
      "epoch": 0.49920603414053194,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981910016109914,
      "loss": 5.6016,
      "mean_token_accuracy": 0.13954204469919204,
      "num_tokens": 11008349.0,
      "step": 5030
    },
    {
      "entropy": 5.680003976821899,
      "epoch": 0.49970226280269947,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981865149260189,
      "loss": 5.4527,
      "mean_token_accuracy": 0.15156907960772514,
      "num_tokens": 11017925.0,
      "step": 5035
    },
    {
      "entropy": 5.7168131351470945,
      "epoch": 0.5001984914648671,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981820227065016,
      "loss": 5.5177,
      "mean_token_accuracy": 0.14543583542108535,
      "num_tokens": 11028710.0,
      "step": 5040
    },
    {
      "entropy": 5.643894672393799,
      "epoch": 0.5006947201270345,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981775249525513,
      "loss": 5.519,
      "mean_token_accuracy": 0.1456161230802536,
      "num_tokens": 11039827.0,
      "step": 5045
    },
    {
      "entropy": 5.646371173858642,
      "epoch": 0.5011909487892021,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981730216642791,
      "loss": 5.5482,
      "mean_token_accuracy": 0.14045156687498092,
      "num_tokens": 11050875.0,
      "step": 5050
    },
    {
      "entropy": 5.711731767654419,
      "epoch": 0.5016871774513696,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004981685128417969,
      "loss": 5.5751,
      "mean_token_accuracy": 0.14403237849473954,
      "num_tokens": 11063530.0,
      "step": 5055
    },
    {
      "entropy": 5.709342193603516,
      "epoch": 0.5021834061135371,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004981639984852165,
      "loss": 5.5797,
      "mean_token_accuracy": 0.1439053326845169,
      "num_tokens": 11074644.0,
      "step": 5060
    },
    {
      "entropy": 5.6850138187408445,
      "epoch": 0.5026796347757047,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981594785946498,
      "loss": 5.5317,
      "mean_token_accuracy": 0.1438378646969795,
      "num_tokens": 11086204.0,
      "step": 5065
    },
    {
      "entropy": 5.63051438331604,
      "epoch": 0.5031758634378721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981549531702089,
      "loss": 5.5084,
      "mean_token_accuracy": 0.14871567264199256,
      "num_tokens": 11096547.0,
      "step": 5070
    },
    {
      "entropy": 5.695504570007325,
      "epoch": 0.5036720921000397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498150422212006,
      "loss": 5.4931,
      "mean_token_accuracy": 0.14339740574359894,
      "num_tokens": 11106692.0,
      "step": 5075
    },
    {
      "entropy": 5.70011682510376,
      "epoch": 0.5041683207622072,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981458857201534,
      "loss": 5.4555,
      "mean_token_accuracy": 0.14496204182505606,
      "num_tokens": 11118065.0,
      "step": 5080
    },
    {
      "entropy": 5.587588691711426,
      "epoch": 0.5046645494243748,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981413436947639,
      "loss": 5.4376,
      "mean_token_accuracy": 0.14765066653490067,
      "num_tokens": 11127795.0,
      "step": 5085
    },
    {
      "entropy": 5.530546379089356,
      "epoch": 0.5051607780865423,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981367961359496,
      "loss": 5.4375,
      "mean_token_accuracy": 0.1470058299601078,
      "num_tokens": 11138936.0,
      "step": 5090
    },
    {
      "entropy": 5.8068459033966064,
      "epoch": 0.5056570067487098,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981322430438239,
      "loss": 5.614,
      "mean_token_accuracy": 0.14049707353115082,
      "num_tokens": 11150671.0,
      "step": 5095
    },
    {
      "entropy": 5.6806396484375,
      "epoch": 0.5061532354108773,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981276844184991,
      "loss": 5.4908,
      "mean_token_accuracy": 0.15141792446374894,
      "num_tokens": 11161413.0,
      "step": 5100
    },
    {
      "entropy": 5.614329624176025,
      "epoch": 0.5066494640730449,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981231202600886,
      "loss": 5.5698,
      "mean_token_accuracy": 0.15129174441099166,
      "num_tokens": 11171773.0,
      "step": 5105
    },
    {
      "entropy": 5.65161509513855,
      "epoch": 0.5071456927352124,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004981185505687056,
      "loss": 5.5875,
      "mean_token_accuracy": 0.13981531485915183,
      "num_tokens": 11183373.0,
      "step": 5110
    },
    {
      "entropy": 5.614207744598389,
      "epoch": 0.50764192139738,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004981139753444632,
      "loss": 5.4051,
      "mean_token_accuracy": 0.15036127269268035,
      "num_tokens": 11195535.0,
      "step": 5115
    },
    {
      "entropy": 5.722874021530151,
      "epoch": 0.5081381500595474,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981093945874752,
      "loss": 5.6044,
      "mean_token_accuracy": 0.14495106786489487,
      "num_tokens": 11206937.0,
      "step": 5120
    },
    {
      "entropy": 5.670475959777832,
      "epoch": 0.508634378721715,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004981048082978547,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1461784139275551,
      "num_tokens": 11218179.0,
      "step": 5125
    },
    {
      "entropy": 5.687043333053589,
      "epoch": 0.5091306073838825,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981002164757158,
      "loss": 5.5546,
      "mean_token_accuracy": 0.13834378197789193,
      "num_tokens": 11229444.0,
      "step": 5130
    },
    {
      "entropy": 5.710991334915161,
      "epoch": 0.50962683604605,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980956191211723,
      "loss": 5.5495,
      "mean_token_accuracy": 0.14733220040798187,
      "num_tokens": 11240869.0,
      "step": 5135
    },
    {
      "entropy": 5.6909466743469235,
      "epoch": 0.5101230647082176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980910162343382,
      "loss": 5.5002,
      "mean_token_accuracy": 0.14717749282717704,
      "num_tokens": 11251110.0,
      "step": 5140
    },
    {
      "entropy": 5.594380617141724,
      "epoch": 0.510619293370385,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980864078153275,
      "loss": 5.5729,
      "mean_token_accuracy": 0.1491304263472557,
      "num_tokens": 11263522.0,
      "step": 5145
    },
    {
      "entropy": 5.630468463897705,
      "epoch": 0.5111155220325526,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980817938642548,
      "loss": 5.3525,
      "mean_token_accuracy": 0.1591050311923027,
      "num_tokens": 11273703.0,
      "step": 5150
    },
    {
      "entropy": 5.7466590881347654,
      "epoch": 0.5116117506947201,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498077174381234,
      "loss": 5.5845,
      "mean_token_accuracy": 0.14163955748081208,
      "num_tokens": 11285834.0,
      "step": 5155
    },
    {
      "entropy": 5.665092039108276,
      "epoch": 0.5121079793568877,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980725493663802,
      "loss": 5.4993,
      "mean_token_accuracy": 0.15115323513746262,
      "num_tokens": 11297382.0,
      "step": 5160
    },
    {
      "entropy": 5.634912586212158,
      "epoch": 0.5126042080190552,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004980679188198078,
      "loss": 5.4796,
      "mean_token_accuracy": 0.14199309945106506,
      "num_tokens": 11308869.0,
      "step": 5165
    },
    {
      "entropy": 5.630853271484375,
      "epoch": 0.5131004366812227,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980632827416317,
      "loss": 5.5351,
      "mean_token_accuracy": 0.15014643371105194,
      "num_tokens": 11319273.0,
      "step": 5170
    },
    {
      "entropy": 5.712605810165405,
      "epoch": 0.5135966653433902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980586411319669,
      "loss": 5.4823,
      "mean_token_accuracy": 0.145989890396595,
      "num_tokens": 11329996.0,
      "step": 5175
    },
    {
      "entropy": 5.739603662490845,
      "epoch": 0.5140928940055578,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004980539939909283,
      "loss": 5.4701,
      "mean_token_accuracy": 0.14686063602566718,
      "num_tokens": 11341338.0,
      "step": 5180
    },
    {
      "entropy": 5.564398336410522,
      "epoch": 0.5145891226677253,
      "grad_norm": 0.875,
      "learning_rate": 0.0004980493413186314,
      "loss": 5.4412,
      "mean_token_accuracy": 0.14763062447309494,
      "num_tokens": 11352546.0,
      "step": 5185
    },
    {
      "entropy": 5.648761749267578,
      "epoch": 0.5150853513298929,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980446831151914,
      "loss": 5.5408,
      "mean_token_accuracy": 0.1417682185769081,
      "num_tokens": 11363418.0,
      "step": 5190
    },
    {
      "entropy": 5.727824640274048,
      "epoch": 0.5155815799920603,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980400193807239,
      "loss": 5.5767,
      "mean_token_accuracy": 0.14090561643242835,
      "num_tokens": 11373715.0,
      "step": 5195
    },
    {
      "entropy": 5.681086826324463,
      "epoch": 0.5160778086542279,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980353501153446,
      "loss": 5.5,
      "mean_token_accuracy": 0.1466321639716625,
      "num_tokens": 11385881.0,
      "step": 5200
    },
    {
      "entropy": 5.661572599411011,
      "epoch": 0.5165740373163954,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980306753191691,
      "loss": 5.468,
      "mean_token_accuracy": 0.1451452910900116,
      "num_tokens": 11395917.0,
      "step": 5205
    },
    {
      "entropy": 5.672505140304565,
      "epoch": 0.5170702659785629,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980259949923135,
      "loss": 5.4589,
      "mean_token_accuracy": 0.1444351777434349,
      "num_tokens": 11407210.0,
      "step": 5210
    },
    {
      "entropy": 5.6683014869689945,
      "epoch": 0.5175664946407305,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980213091348938,
      "loss": 5.5184,
      "mean_token_accuracy": 0.14767149537801744,
      "num_tokens": 11418715.0,
      "step": 5215
    },
    {
      "entropy": 5.675573825836182,
      "epoch": 0.5180627233028979,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980166177470262,
      "loss": 5.4815,
      "mean_token_accuracy": 0.1486961156129837,
      "num_tokens": 11428964.0,
      "step": 5220
    },
    {
      "entropy": 5.644686031341553,
      "epoch": 0.5185589519650655,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980119208288271,
      "loss": 5.4358,
      "mean_token_accuracy": 0.14670464396476746,
      "num_tokens": 11439644.0,
      "step": 5225
    },
    {
      "entropy": 5.612020254135132,
      "epoch": 0.519055180627233,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980072183804128,
      "loss": 5.5284,
      "mean_token_accuracy": 0.14413978010416031,
      "num_tokens": 11450737.0,
      "step": 5230
    },
    {
      "entropy": 5.660719966888427,
      "epoch": 0.5195514092894006,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980025104019,
      "loss": 5.5101,
      "mean_token_accuracy": 0.1532495766878128,
      "num_tokens": 11461054.0,
      "step": 5235
    },
    {
      "entropy": 5.618498039245606,
      "epoch": 0.520047637951568,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979977968934056,
      "loss": 5.48,
      "mean_token_accuracy": 0.1454416662454605,
      "num_tokens": 11470900.0,
      "step": 5240
    },
    {
      "entropy": 5.599879503250122,
      "epoch": 0.5205438666137356,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979930778550463,
      "loss": 5.4485,
      "mean_token_accuracy": 0.14997834861278533,
      "num_tokens": 11480551.0,
      "step": 5245
    },
    {
      "entropy": 5.619562816619873,
      "epoch": 0.5210400952759031,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979883532869392,
      "loss": 5.4151,
      "mean_token_accuracy": 0.15181739181280135,
      "num_tokens": 11491317.0,
      "step": 5250
    },
    {
      "entropy": 5.612633514404297,
      "epoch": 0.5215363239380707,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979836231892013,
      "loss": 5.5276,
      "mean_token_accuracy": 0.1496187038719654,
      "num_tokens": 11502975.0,
      "step": 5255
    },
    {
      "entropy": 5.651110553741455,
      "epoch": 0.5220325526002382,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979788875619501,
      "loss": 5.5099,
      "mean_token_accuracy": 0.14345327392220497,
      "num_tokens": 11514803.0,
      "step": 5260
    },
    {
      "entropy": 5.711249494552613,
      "epoch": 0.5225287812624058,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497974146405303,
      "loss": 5.5243,
      "mean_token_accuracy": 0.14779631793498993,
      "num_tokens": 11525782.0,
      "step": 5265
    },
    {
      "entropy": 5.725921487808227,
      "epoch": 0.5230250099245732,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979693997193776,
      "loss": 5.5637,
      "mean_token_accuracy": 0.14164871126413345,
      "num_tokens": 11535857.0,
      "step": 5270
    },
    {
      "entropy": 5.683490753173828,
      "epoch": 0.5235212385867408,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979646475042914,
      "loss": 5.5801,
      "mean_token_accuracy": 0.14225950986146926,
      "num_tokens": 11547385.0,
      "step": 5275
    },
    {
      "entropy": 5.683238220214844,
      "epoch": 0.5240174672489083,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979598897601625,
      "loss": 5.462,
      "mean_token_accuracy": 0.15227275267243384,
      "num_tokens": 11559382.0,
      "step": 5280
    },
    {
      "entropy": 5.6968690872192385,
      "epoch": 0.5245136959110758,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979551264871088,
      "loss": 5.5315,
      "mean_token_accuracy": 0.1423482984304428,
      "num_tokens": 11570096.0,
      "step": 5285
    },
    {
      "entropy": 5.604145908355713,
      "epoch": 0.5250099245732434,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979503576852482,
      "loss": 5.4257,
      "mean_token_accuracy": 0.15198830962181092,
      "num_tokens": 11580543.0,
      "step": 5290
    },
    {
      "entropy": 5.608744812011719,
      "epoch": 0.5255061532354108,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979455833546994,
      "loss": 5.4048,
      "mean_token_accuracy": 0.15378104001283646,
      "num_tokens": 11591132.0,
      "step": 5295
    },
    {
      "entropy": 5.65214729309082,
      "epoch": 0.5260023818975784,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979408034955804,
      "loss": 5.5749,
      "mean_token_accuracy": 0.144743674993515,
      "num_tokens": 11602684.0,
      "step": 5300
    },
    {
      "entropy": 5.657083559036255,
      "epoch": 0.5264986105597459,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004979360181080099,
      "loss": 5.522,
      "mean_token_accuracy": 0.15236675292253493,
      "num_tokens": 11614009.0,
      "step": 5305
    },
    {
      "entropy": 5.6740587711334225,
      "epoch": 0.5269948392219135,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979312271921065,
      "loss": 5.6297,
      "mean_token_accuracy": 0.14404670298099517,
      "num_tokens": 11624577.0,
      "step": 5310
    },
    {
      "entropy": 5.729754543304443,
      "epoch": 0.527491067884081,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000497926430747989,
      "loss": 5.5326,
      "mean_token_accuracy": 0.1465068504214287,
      "num_tokens": 11636382.0,
      "step": 5315
    },
    {
      "entropy": 5.604878950119018,
      "epoch": 0.5279872965462485,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979216287757764,
      "loss": 5.4887,
      "mean_token_accuracy": 0.14911949038505554,
      "num_tokens": 11648606.0,
      "step": 5320
    },
    {
      "entropy": 5.668821477890015,
      "epoch": 0.528483525208416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497916821275588,
      "loss": 5.455,
      "mean_token_accuracy": 0.1496219277381897,
      "num_tokens": 11658191.0,
      "step": 5325
    },
    {
      "entropy": 5.76564884185791,
      "epoch": 0.5289797538705836,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004979120082475424,
      "loss": 5.5496,
      "mean_token_accuracy": 0.14533232972025872,
      "num_tokens": 11668955.0,
      "step": 5330
    },
    {
      "entropy": 5.60173659324646,
      "epoch": 0.5294759825327511,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979071896917595,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1412031590938568,
      "num_tokens": 11680634.0,
      "step": 5335
    },
    {
      "entropy": 5.720175409317017,
      "epoch": 0.5299722111949187,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979023656083586,
      "loss": 5.5596,
      "mean_token_accuracy": 0.14906308352947234,
      "num_tokens": 11692155.0,
      "step": 5340
    },
    {
      "entropy": 5.674312019348145,
      "epoch": 0.5304684398570861,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978975359974594,
      "loss": 5.5054,
      "mean_token_accuracy": 0.1465664565563202,
      "num_tokens": 11702968.0,
      "step": 5345
    },
    {
      "entropy": 5.731066370010376,
      "epoch": 0.5309646685192537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978927008591816,
      "loss": 5.5663,
      "mean_token_accuracy": 0.14366454184055327,
      "num_tokens": 11714329.0,
      "step": 5350
    },
    {
      "entropy": 5.572608184814453,
      "epoch": 0.5314608971814212,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978878601936451,
      "loss": 5.3614,
      "mean_token_accuracy": 0.15541550219058992,
      "num_tokens": 11724071.0,
      "step": 5355
    },
    {
      "entropy": 5.637559318542481,
      "epoch": 0.5319571258435887,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978830140009698,
      "loss": 5.556,
      "mean_token_accuracy": 0.1455766476690769,
      "num_tokens": 11735939.0,
      "step": 5360
    },
    {
      "entropy": 5.675500440597534,
      "epoch": 0.5324533545057563,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978781622812762,
      "loss": 5.5845,
      "mean_token_accuracy": 0.1446147307753563,
      "num_tokens": 11747965.0,
      "step": 5365
    },
    {
      "entropy": 5.713941192626953,
      "epoch": 0.5329495831679237,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978733050346844,
      "loss": 5.4754,
      "mean_token_accuracy": 0.1473374918103218,
      "num_tokens": 11758206.0,
      "step": 5370
    },
    {
      "entropy": 5.664413547515869,
      "epoch": 0.5334458118300913,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004978684422613149,
      "loss": 5.5079,
      "mean_token_accuracy": 0.14940823540091513,
      "num_tokens": 11770158.0,
      "step": 5375
    },
    {
      "entropy": 5.513971376419067,
      "epoch": 0.5339420404922588,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978635739612882,
      "loss": 5.3556,
      "mean_token_accuracy": 0.15853073000907897,
      "num_tokens": 11780332.0,
      "step": 5380
    },
    {
      "entropy": 5.582099056243896,
      "epoch": 0.5344382691544264,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004978587001347252,
      "loss": 5.4897,
      "mean_token_accuracy": 0.1498791128396988,
      "num_tokens": 11791178.0,
      "step": 5385
    },
    {
      "entropy": 5.657362174987793,
      "epoch": 0.5349344978165939,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978538207817465,
      "loss": 5.4485,
      "mean_token_accuracy": 0.14950233101844787,
      "num_tokens": 11800701.0,
      "step": 5390
    },
    {
      "entropy": 5.616297245025635,
      "epoch": 0.5354307264787614,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004978489359024734,
      "loss": 5.4566,
      "mean_token_accuracy": 0.15169827863574029,
      "num_tokens": 11811727.0,
      "step": 5395
    },
    {
      "entropy": 5.636269569396973,
      "epoch": 0.5359269551409289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978440454970268,
      "loss": 5.5457,
      "mean_token_accuracy": 0.14859212636947633,
      "num_tokens": 11822892.0,
      "step": 5400
    },
    {
      "entropy": 5.6795073509216305,
      "epoch": 0.5364231838030965,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978391495655281,
      "loss": 5.505,
      "mean_token_accuracy": 0.14875393807888032,
      "num_tokens": 11834547.0,
      "step": 5405
    },
    {
      "entropy": 5.65238094329834,
      "epoch": 0.536919412465264,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978342481080987,
      "loss": 5.4854,
      "mean_token_accuracy": 0.15107561349868776,
      "num_tokens": 11845667.0,
      "step": 5410
    },
    {
      "entropy": 5.635183668136596,
      "epoch": 0.5374156411274316,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004978293411248601,
      "loss": 5.4149,
      "mean_token_accuracy": 0.15212117433547973,
      "num_tokens": 11856228.0,
      "step": 5415
    },
    {
      "entropy": 5.609660243988037,
      "epoch": 0.537911869789599,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004978244286159339,
      "loss": 5.4669,
      "mean_token_accuracy": 0.14964728355407714,
      "num_tokens": 11867491.0,
      "step": 5420
    },
    {
      "entropy": 5.676966953277588,
      "epoch": 0.5384080984517666,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978195105814421,
      "loss": 5.5615,
      "mean_token_accuracy": 0.14637720137834548,
      "num_tokens": 11878635.0,
      "step": 5425
    },
    {
      "entropy": 5.686103820800781,
      "epoch": 0.5389043271139341,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978145870215065,
      "loss": 5.4823,
      "mean_token_accuracy": 0.15092322900891303,
      "num_tokens": 11890204.0,
      "step": 5430
    },
    {
      "entropy": 5.641102361679077,
      "epoch": 0.5394005557761016,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978096579362493,
      "loss": 5.5251,
      "mean_token_accuracy": 0.15008656457066535,
      "num_tokens": 11901049.0,
      "step": 5435
    },
    {
      "entropy": 5.63104019165039,
      "epoch": 0.5398967844382692,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978047233257927,
      "loss": 5.4979,
      "mean_token_accuracy": 0.14436167031526564,
      "num_tokens": 11912957.0,
      "step": 5440
    },
    {
      "entropy": 5.6869189739227295,
      "epoch": 0.5403930131004366,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977997831902591,
      "loss": 5.5123,
      "mean_token_accuracy": 0.1438940078020096,
      "num_tokens": 11924583.0,
      "step": 5445
    },
    {
      "entropy": 5.549581527709961,
      "epoch": 0.5408892417626042,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004977948375297709,
      "loss": 5.4474,
      "mean_token_accuracy": 0.14686232954263687,
      "num_tokens": 11934898.0,
      "step": 5450
    },
    {
      "entropy": 5.602767658233643,
      "epoch": 0.5413854704247717,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977898863444509,
      "loss": 5.4299,
      "mean_token_accuracy": 0.1516907647252083,
      "num_tokens": 11944577.0,
      "step": 5455
    },
    {
      "entropy": 5.606003665924073,
      "epoch": 0.5418816990869393,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977849296344217,
      "loss": 5.484,
      "mean_token_accuracy": 0.15010611712932587,
      "num_tokens": 11956184.0,
      "step": 5460
    },
    {
      "entropy": 5.632574987411499,
      "epoch": 0.5423779277491068,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004977799673998063,
      "loss": 5.5211,
      "mean_token_accuracy": 0.14659062027931213,
      "num_tokens": 11966929.0,
      "step": 5465
    },
    {
      "entropy": 5.707142353057861,
      "epoch": 0.5428741564112743,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004977749996407278,
      "loss": 5.4547,
      "mean_token_accuracy": 0.15533087998628617,
      "num_tokens": 11978166.0,
      "step": 5470
    },
    {
      "entropy": 5.65168251991272,
      "epoch": 0.5433703850734418,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977700263573094,
      "loss": 5.538,
      "mean_token_accuracy": 0.1417289935052395,
      "num_tokens": 11988719.0,
      "step": 5475
    },
    {
      "entropy": 5.660910892486572,
      "epoch": 0.5438666137356094,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977650475496743,
      "loss": 5.5014,
      "mean_token_accuracy": 0.14837072789669037,
      "num_tokens": 11999877.0,
      "step": 5480
    },
    {
      "entropy": 5.63805513381958,
      "epoch": 0.5443628423977769,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977600632179459,
      "loss": 5.4895,
      "mean_token_accuracy": 0.1501492217183113,
      "num_tokens": 12011123.0,
      "step": 5485
    },
    {
      "entropy": 5.645816135406494,
      "epoch": 0.5448590710599445,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004977550733622482,
      "loss": 5.4539,
      "mean_token_accuracy": 0.15027592182159424,
      "num_tokens": 12021494.0,
      "step": 5490
    },
    {
      "entropy": 5.718552923202514,
      "epoch": 0.5453552997221119,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977500779827045,
      "loss": 5.5147,
      "mean_token_accuracy": 0.14032889828085898,
      "num_tokens": 12032780.0,
      "step": 5495
    },
    {
      "entropy": 5.600382137298584,
      "epoch": 0.5458515283842795,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977450770794389,
      "loss": 5.5251,
      "mean_token_accuracy": 0.14582949876785278,
      "num_tokens": 12043910.0,
      "step": 5500
    },
    {
      "entropy": 5.633700132369995,
      "epoch": 0.546347757046447,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004977400706525755,
      "loss": 5.4334,
      "mean_token_accuracy": 0.15561349242925643,
      "num_tokens": 12053722.0,
      "step": 5505
    },
    {
      "entropy": 5.599662733078003,
      "epoch": 0.5468439857086145,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977350587022381,
      "loss": 5.5363,
      "mean_token_accuracy": 0.14684804230928422,
      "num_tokens": 12065316.0,
      "step": 5510
    },
    {
      "entropy": 5.639979267120362,
      "epoch": 0.5473402143707821,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977300412285513,
      "loss": 5.4756,
      "mean_token_accuracy": 0.1470431424677372,
      "num_tokens": 12076095.0,
      "step": 5515
    },
    {
      "entropy": 5.662402057647705,
      "epoch": 0.5478364430329495,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004977250182316394,
      "loss": 5.4406,
      "mean_token_accuracy": 0.15188678801059724,
      "num_tokens": 12086960.0,
      "step": 5520
    },
    {
      "entropy": 5.5803542137146,
      "epoch": 0.5483326716951171,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497719989711627,
      "loss": 5.4378,
      "mean_token_accuracy": 0.15385584980249406,
      "num_tokens": 12097534.0,
      "step": 5525
    },
    {
      "entropy": 5.587559366226197,
      "epoch": 0.5488289003572846,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977149556686387,
      "loss": 5.3945,
      "mean_token_accuracy": 0.1565391704440117,
      "num_tokens": 12110348.0,
      "step": 5530
    },
    {
      "entropy": 5.599958848953247,
      "epoch": 0.5493251290194522,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977099161027996,
      "loss": 5.4053,
      "mean_token_accuracy": 0.14727725833654404,
      "num_tokens": 12121171.0,
      "step": 5535
    },
    {
      "entropy": 5.624782991409302,
      "epoch": 0.5498213576816197,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977048710142342,
      "loss": 5.5625,
      "mean_token_accuracy": 0.14204036593437194,
      "num_tokens": 12133054.0,
      "step": 5540
    },
    {
      "entropy": 5.640163373947144,
      "epoch": 0.5503175863437872,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976998204030681,
      "loss": 5.4535,
      "mean_token_accuracy": 0.15206095278263093,
      "num_tokens": 12144278.0,
      "step": 5545
    },
    {
      "entropy": 5.69926962852478,
      "epoch": 0.5508138150059547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976947642694262,
      "loss": 5.5589,
      "mean_token_accuracy": 0.14523016959428786,
      "num_tokens": 12154446.0,
      "step": 5550
    },
    {
      "entropy": 5.674716663360596,
      "epoch": 0.5513100436681223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976897026134341,
      "loss": 5.4454,
      "mean_token_accuracy": 0.1498950183391571,
      "num_tokens": 12165176.0,
      "step": 5555
    },
    {
      "entropy": 5.6429126262664795,
      "epoch": 0.5518062723302898,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976846354352172,
      "loss": 5.4673,
      "mean_token_accuracy": 0.14666500836610794,
      "num_tokens": 12176329.0,
      "step": 5560
    },
    {
      "entropy": 5.694870233535767,
      "epoch": 0.5523025009924574,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976795627349012,
      "loss": 5.4877,
      "mean_token_accuracy": 0.14380425065755845,
      "num_tokens": 12186927.0,
      "step": 5565
    },
    {
      "entropy": 5.698277950286865,
      "epoch": 0.5527987296546248,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976744845126118,
      "loss": 5.517,
      "mean_token_accuracy": 0.15412544310092927,
      "num_tokens": 12196632.0,
      "step": 5570
    },
    {
      "entropy": 5.620586776733399,
      "epoch": 0.5532949583167924,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976694007684751,
      "loss": 5.4737,
      "mean_token_accuracy": 0.14799923151731492,
      "num_tokens": 12207099.0,
      "step": 5575
    },
    {
      "entropy": 5.701907730102539,
      "epoch": 0.5537911869789599,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497664311502617,
      "loss": 5.5328,
      "mean_token_accuracy": 0.14305446594953536,
      "num_tokens": 12218595.0,
      "step": 5580
    },
    {
      "entropy": 5.640064001083374,
      "epoch": 0.5542874156411274,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004976592167151638,
      "loss": 5.4908,
      "mean_token_accuracy": 0.14687163084745408,
      "num_tokens": 12229280.0,
      "step": 5585
    },
    {
      "entropy": 5.598618268966675,
      "epoch": 0.554783644303295,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976541164062418,
      "loss": 5.4603,
      "mean_token_accuracy": 0.14839887768030166,
      "num_tokens": 12240079.0,
      "step": 5590
    },
    {
      "entropy": 5.583456420898438,
      "epoch": 0.5552798729654624,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976490105759774,
      "loss": 5.4222,
      "mean_token_accuracy": 0.15392844825983049,
      "num_tokens": 12250083.0,
      "step": 5595
    },
    {
      "entropy": 5.559115409851074,
      "epoch": 0.55577610162763,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976438992244976,
      "loss": 5.3652,
      "mean_token_accuracy": 0.1540672928094864,
      "num_tokens": 12260868.0,
      "step": 5600
    },
    {
      "entropy": 5.579015827178955,
      "epoch": 0.5562723302897975,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976387823519286,
      "loss": 5.4278,
      "mean_token_accuracy": 0.1523951455950737,
      "num_tokens": 12272268.0,
      "step": 5605
    },
    {
      "entropy": 5.638259315490723,
      "epoch": 0.5567685589519651,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976336599583975,
      "loss": 5.4662,
      "mean_token_accuracy": 0.14606835395097734,
      "num_tokens": 12283347.0,
      "step": 5610
    },
    {
      "entropy": 5.615521335601807,
      "epoch": 0.5572647876141326,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976285320440315,
      "loss": 5.4864,
      "mean_token_accuracy": 0.14265448227524757,
      "num_tokens": 12294639.0,
      "step": 5615
    },
    {
      "entropy": 5.619390964508057,
      "epoch": 0.5577610162763001,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976233986089577,
      "loss": 5.4556,
      "mean_token_accuracy": 0.1584750235080719,
      "num_tokens": 12304573.0,
      "step": 5620
    },
    {
      "entropy": 5.66530966758728,
      "epoch": 0.5582572449384676,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976182596533031,
      "loss": 5.4949,
      "mean_token_accuracy": 0.14809588938951493,
      "num_tokens": 12314559.0,
      "step": 5625
    },
    {
      "entropy": 5.624449920654297,
      "epoch": 0.5587534736006352,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976131151771955,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1499591127038002,
      "num_tokens": 12324433.0,
      "step": 5630
    },
    {
      "entropy": 5.606946229934692,
      "epoch": 0.5592497022628027,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976079651807623,
      "loss": 5.5623,
      "mean_token_accuracy": 0.13970149755477906,
      "num_tokens": 12335039.0,
      "step": 5635
    },
    {
      "entropy": 5.681036949157715,
      "epoch": 0.5597459309249703,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976028096641313,
      "loss": 5.443,
      "mean_token_accuracy": 0.15265071988105774,
      "num_tokens": 12345917.0,
      "step": 5640
    },
    {
      "entropy": 5.668826580047607,
      "epoch": 0.5602421595871377,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975976486274302,
      "loss": 5.5031,
      "mean_token_accuracy": 0.15012277215719222,
      "num_tokens": 12357362.0,
      "step": 5645
    },
    {
      "entropy": 5.615872049331665,
      "epoch": 0.5607383882493053,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975924820707872,
      "loss": 5.4883,
      "mean_token_accuracy": 0.14900902062654495,
      "num_tokens": 12368012.0,
      "step": 5650
    },
    {
      "entropy": 5.671743774414063,
      "epoch": 0.5612346169114728,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004975873099943302,
      "loss": 5.5216,
      "mean_token_accuracy": 0.1491253674030304,
      "num_tokens": 12378745.0,
      "step": 5655
    },
    {
      "entropy": 5.617655992507935,
      "epoch": 0.5617308455736403,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975821323981875,
      "loss": 5.4054,
      "mean_token_accuracy": 0.15520506650209426,
      "num_tokens": 12388821.0,
      "step": 5660
    },
    {
      "entropy": 5.748631334304809,
      "epoch": 0.5622270742358079,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004975769492824877,
      "loss": 5.5503,
      "mean_token_accuracy": 0.1412370964884758,
      "num_tokens": 12399798.0,
      "step": 5665
    },
    {
      "entropy": 5.725917530059815,
      "epoch": 0.5627233028979753,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975717606473589,
      "loss": 5.4661,
      "mean_token_accuracy": 0.14498765617609025,
      "num_tokens": 12410887.0,
      "step": 5670
    },
    {
      "entropy": 5.63080472946167,
      "epoch": 0.5632195315601429,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975665664929302,
      "loss": 5.4997,
      "mean_token_accuracy": 0.15183598101139067,
      "num_tokens": 12422430.0,
      "step": 5675
    },
    {
      "entropy": 5.5743249416351315,
      "epoch": 0.5637157602223104,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975613668193303,
      "loss": 5.4612,
      "mean_token_accuracy": 0.14651041626930236,
      "num_tokens": 12432930.0,
      "step": 5680
    },
    {
      "entropy": 5.644167137145996,
      "epoch": 0.564211988884478,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497556161626688,
      "loss": 5.5081,
      "mean_token_accuracy": 0.14547248035669327,
      "num_tokens": 12444117.0,
      "step": 5685
    },
    {
      "entropy": 5.639104509353638,
      "epoch": 0.5647082175466455,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004975509509151325,
      "loss": 5.4528,
      "mean_token_accuracy": 0.15139108449220656,
      "num_tokens": 12454619.0,
      "step": 5690
    },
    {
      "entropy": 5.610605144500733,
      "epoch": 0.565204446208813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975457346847929,
      "loss": 5.4907,
      "mean_token_accuracy": 0.15113139301538467,
      "num_tokens": 12464666.0,
      "step": 5695
    },
    {
      "entropy": 5.653555965423584,
      "epoch": 0.5657006748709805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975405129357986,
      "loss": 5.5235,
      "mean_token_accuracy": 0.15141180157661438,
      "num_tokens": 12474923.0,
      "step": 5700
    },
    {
      "entropy": 5.6175727367401125,
      "epoch": 0.5661969035331481,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975352856682792,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1477993756532669,
      "num_tokens": 12485505.0,
      "step": 5705
    },
    {
      "entropy": 5.603899240493774,
      "epoch": 0.5666931321953156,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004975300528823642,
      "loss": 5.5117,
      "mean_token_accuracy": 0.1497193917632103,
      "num_tokens": 12497489.0,
      "step": 5710
    },
    {
      "entropy": 5.623446035385132,
      "epoch": 0.5671893608574832,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004975248145781834,
      "loss": 5.4571,
      "mean_token_accuracy": 0.15412696450948715,
      "num_tokens": 12508117.0,
      "step": 5715
    },
    {
      "entropy": 5.62113618850708,
      "epoch": 0.5676855895196506,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975195707558667,
      "loss": 5.4604,
      "mean_token_accuracy": 0.1502758279442787,
      "num_tokens": 12519090.0,
      "step": 5720
    },
    {
      "entropy": 5.661679029464722,
      "epoch": 0.5681818181818182,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975143214155443,
      "loss": 5.5929,
      "mean_token_accuracy": 0.13893014937639236,
      "num_tokens": 12530427.0,
      "step": 5725
    },
    {
      "entropy": 5.5774599552154545,
      "epoch": 0.5686780468439857,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004975090665573459,
      "loss": 5.3854,
      "mean_token_accuracy": 0.15700648427009584,
      "num_tokens": 12542142.0,
      "step": 5730
    },
    {
      "entropy": 5.6237077713012695,
      "epoch": 0.5691742755061532,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004975038061814023,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1490109980106354,
      "num_tokens": 12552716.0,
      "step": 5735
    },
    {
      "entropy": 5.601270580291748,
      "epoch": 0.5696705041683208,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974985402878437,
      "loss": 5.4309,
      "mean_token_accuracy": 0.14649184048175812,
      "num_tokens": 12562659.0,
      "step": 5740
    },
    {
      "entropy": 5.63472318649292,
      "epoch": 0.5701667328304882,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974932688768007,
      "loss": 5.3812,
      "mean_token_accuracy": 0.16051386147737504,
      "num_tokens": 12573982.0,
      "step": 5745
    },
    {
      "entropy": 5.59278564453125,
      "epoch": 0.5706629614926558,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974879919484042,
      "loss": 5.3771,
      "mean_token_accuracy": 0.16215328127145767,
      "num_tokens": 12584402.0,
      "step": 5750
    },
    {
      "entropy": 5.600103855133057,
      "epoch": 0.5711591901548233,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974827095027848,
      "loss": 5.4799,
      "mean_token_accuracy": 0.14797588139772416,
      "num_tokens": 12595342.0,
      "step": 5755
    },
    {
      "entropy": 5.556008243560791,
      "epoch": 0.5716554188169909,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004974774215400736,
      "loss": 5.4138,
      "mean_token_accuracy": 0.1532015845179558,
      "num_tokens": 12605634.0,
      "step": 5760
    },
    {
      "entropy": 5.6285937309265135,
      "epoch": 0.5721516474791584,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004974721280604019,
      "loss": 5.4425,
      "mean_token_accuracy": 0.15120814219117165,
      "num_tokens": 12618000.0,
      "step": 5765
    },
    {
      "entropy": 5.541146898269654,
      "epoch": 0.572647876141326,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004974668290639007,
      "loss": 5.2707,
      "mean_token_accuracy": 0.16064726561307907,
      "num_tokens": 12629744.0,
      "step": 5770
    },
    {
      "entropy": 5.541376972198487,
      "epoch": 0.5731441048034934,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974615245507014,
      "loss": 5.3864,
      "mean_token_accuracy": 0.1517522320151329,
      "num_tokens": 12639752.0,
      "step": 5775
    },
    {
      "entropy": 5.586554098129272,
      "epoch": 0.573640333465661,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974562145209359,
      "loss": 5.4134,
      "mean_token_accuracy": 0.15405119508504866,
      "num_tokens": 12650219.0,
      "step": 5780
    },
    {
      "entropy": 5.645528554916382,
      "epoch": 0.5741365621278285,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004974508989747354,
      "loss": 5.5267,
      "mean_token_accuracy": 0.14913564845919608,
      "num_tokens": 12661589.0,
      "step": 5785
    },
    {
      "entropy": 5.650177383422852,
      "epoch": 0.5746327907899961,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004974455779122321,
      "loss": 5.5231,
      "mean_token_accuracy": 0.1432395339012146,
      "num_tokens": 12674794.0,
      "step": 5790
    },
    {
      "entropy": 5.671242523193359,
      "epoch": 0.5751290194521635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974402513335577,
      "loss": 5.4289,
      "mean_token_accuracy": 0.15152156949043274,
      "num_tokens": 12684573.0,
      "step": 5795
    },
    {
      "entropy": 5.618958425521851,
      "epoch": 0.5756252481143311,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974349192388444,
      "loss": 5.4241,
      "mean_token_accuracy": 0.15097050070762635,
      "num_tokens": 12695319.0,
      "step": 5800
    },
    {
      "entropy": 5.613748025894165,
      "epoch": 0.5761214767764986,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004974295816282245,
      "loss": 5.446,
      "mean_token_accuracy": 0.15115070939064026,
      "num_tokens": 12706007.0,
      "step": 5805
    },
    {
      "entropy": 5.643027639389038,
      "epoch": 0.5766177054386661,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974242385018303,
      "loss": 5.4498,
      "mean_token_accuracy": 0.15285826176404954,
      "num_tokens": 12716213.0,
      "step": 5810
    },
    {
      "entropy": 5.631270408630371,
      "epoch": 0.5771139341008337,
      "grad_norm": 0.828125,
      "learning_rate": 0.000497418889859794,
      "loss": 5.4739,
      "mean_token_accuracy": 0.14576493203639984,
      "num_tokens": 12727643.0,
      "step": 5815
    },
    {
      "entropy": 5.654403162002564,
      "epoch": 0.5776101627630011,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974135357022488,
      "loss": 5.4814,
      "mean_token_accuracy": 0.14785733744502066,
      "num_tokens": 12738535.0,
      "step": 5820
    },
    {
      "entropy": 5.621318721771241,
      "epoch": 0.5781063914251687,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974081760293269,
      "loss": 5.4587,
      "mean_token_accuracy": 0.15592710971832274,
      "num_tokens": 12750419.0,
      "step": 5825
    },
    {
      "entropy": 5.598451280593872,
      "epoch": 0.5786026200873362,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974028108411617,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1496580272912979,
      "num_tokens": 12759827.0,
      "step": 5830
    },
    {
      "entropy": 5.597668027877807,
      "epoch": 0.5790988487495038,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004973974401378859,
      "loss": 5.4591,
      "mean_token_accuracy": 0.15755970627069474,
      "num_tokens": 12770328.0,
      "step": 5835
    },
    {
      "entropy": 5.694442558288574,
      "epoch": 0.5795950774116713,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973920639196329,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1516176737844944,
      "num_tokens": 12781788.0,
      "step": 5840
    },
    {
      "entropy": 5.528429794311523,
      "epoch": 0.5800913060738389,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973866821865358,
      "loss": 5.3581,
      "mean_token_accuracy": 0.15554836839437486,
      "num_tokens": 12792134.0,
      "step": 5845
    },
    {
      "entropy": 5.546235418319702,
      "epoch": 0.5805875347360063,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973812949387283,
      "loss": 5.4052,
      "mean_token_accuracy": 0.15124022364616393,
      "num_tokens": 12801810.0,
      "step": 5850
    },
    {
      "entropy": 5.715386962890625,
      "epoch": 0.5810837633981739,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004973759021763439,
      "loss": 5.5558,
      "mean_token_accuracy": 0.14452450424432756,
      "num_tokens": 12814412.0,
      "step": 5855
    },
    {
      "entropy": 5.603662347793579,
      "epoch": 0.5815799920603414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973705038995162,
      "loss": 5.4815,
      "mean_token_accuracy": 0.14441258907318116,
      "num_tokens": 12826344.0,
      "step": 5860
    },
    {
      "entropy": 5.599206399917603,
      "epoch": 0.582076220722509,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973651001083792,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1475580006837845,
      "num_tokens": 12837434.0,
      "step": 5865
    },
    {
      "entropy": 5.657679271697998,
      "epoch": 0.5825724493846764,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973596908030669,
      "loss": 5.4672,
      "mean_token_accuracy": 0.14360740557312965,
      "num_tokens": 12848681.0,
      "step": 5870
    },
    {
      "entropy": 5.635468912124634,
      "epoch": 0.583068678046844,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973542759837135,
      "loss": 5.5863,
      "mean_token_accuracy": 0.14105776473879814,
      "num_tokens": 12860300.0,
      "step": 5875
    },
    {
      "entropy": 5.650388908386231,
      "epoch": 0.5835649067090115,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497348855650453,
      "loss": 5.4602,
      "mean_token_accuracy": 0.15065626204013824,
      "num_tokens": 12871079.0,
      "step": 5880
    },
    {
      "entropy": 5.555815362930298,
      "epoch": 0.584061135371179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004973434298034201,
      "loss": 5.4585,
      "mean_token_accuracy": 0.1567874312400818,
      "num_tokens": 12881168.0,
      "step": 5885
    },
    {
      "entropy": 5.6398046016693115,
      "epoch": 0.5845573640333466,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973379984427492,
      "loss": 5.4982,
      "mean_token_accuracy": 0.14602236449718475,
      "num_tokens": 12891912.0,
      "step": 5890
    },
    {
      "entropy": 5.579549646377563,
      "epoch": 0.585053592695514,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497332561568575,
      "loss": 5.4486,
      "mean_token_accuracy": 0.14765242114663124,
      "num_tokens": 12902984.0,
      "step": 5895
    },
    {
      "entropy": 5.659204006195068,
      "epoch": 0.5855498213576816,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004973271191810325,
      "loss": 5.4919,
      "mean_token_accuracy": 0.15171285420656205,
      "num_tokens": 12914635.0,
      "step": 5900
    },
    {
      "entropy": 5.649992895126343,
      "epoch": 0.5860460500198491,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973216712802564,
      "loss": 5.3997,
      "mean_token_accuracy": 0.1511367067694664,
      "num_tokens": 12924575.0,
      "step": 5905
    },
    {
      "entropy": 5.576608276367187,
      "epoch": 0.5865422786820167,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497316217866382,
      "loss": 5.3563,
      "mean_token_accuracy": 0.15268706902861595,
      "num_tokens": 12935404.0,
      "step": 5910
    },
    {
      "entropy": 5.572375726699829,
      "epoch": 0.5870385073441842,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973107589395444,
      "loss": 5.4282,
      "mean_token_accuracy": 0.16134506613016128,
      "num_tokens": 12946772.0,
      "step": 5915
    },
    {
      "entropy": 5.623561143875122,
      "epoch": 0.5875347360063518,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973052944998791,
      "loss": 5.4937,
      "mean_token_accuracy": 0.14911405146121978,
      "num_tokens": 12956651.0,
      "step": 5920
    },
    {
      "entropy": 5.667007160186768,
      "epoch": 0.5880309646685192,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004972998245475214,
      "loss": 5.5477,
      "mean_token_accuracy": 0.1457376264035702,
      "num_tokens": 12968206.0,
      "step": 5925
    },
    {
      "entropy": 5.710270595550537,
      "epoch": 0.5885271933306868,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497294349082607,
      "loss": 5.5005,
      "mean_token_accuracy": 0.15173040181398392,
      "num_tokens": 12980050.0,
      "step": 5930
    },
    {
      "entropy": 5.610395860671997,
      "epoch": 0.5890234219928543,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972888681052719,
      "loss": 5.4908,
      "mean_token_accuracy": 0.14804925322532653,
      "num_tokens": 12990514.0,
      "step": 5935
    },
    {
      "entropy": 5.663093852996826,
      "epoch": 0.5895196506550219,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972833816156518,
      "loss": 5.5201,
      "mean_token_accuracy": 0.147402523458004,
      "num_tokens": 13000803.0,
      "step": 5940
    },
    {
      "entropy": 5.575763416290283,
      "epoch": 0.5900158793171894,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972778896138828,
      "loss": 5.4036,
      "mean_token_accuracy": 0.1546789139509201,
      "num_tokens": 13010877.0,
      "step": 5945
    },
    {
      "entropy": 5.572369384765625,
      "epoch": 0.5905121079793569,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004972723921001011,
      "loss": 5.4299,
      "mean_token_accuracy": 0.15311885476112366,
      "num_tokens": 13021906.0,
      "step": 5950
    },
    {
      "entropy": 5.649492931365967,
      "epoch": 0.5910083366415244,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004972668890744431,
      "loss": 5.5016,
      "mean_token_accuracy": 0.1487954691052437,
      "num_tokens": 13033977.0,
      "step": 5955
    },
    {
      "entropy": 5.718819284439087,
      "epoch": 0.5915045653036919,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972613805370452,
      "loss": 5.5562,
      "mean_token_accuracy": 0.1481661319732666,
      "num_tokens": 13044078.0,
      "step": 5960
    },
    {
      "entropy": 5.6608216762542725,
      "epoch": 0.5920007939658595,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972558664880439,
      "loss": 5.4965,
      "mean_token_accuracy": 0.14956050366163254,
      "num_tokens": 13055779.0,
      "step": 5965
    },
    {
      "entropy": 5.583147382736206,
      "epoch": 0.592497022628027,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000497250346927576,
      "loss": 5.403,
      "mean_token_accuracy": 0.1515258476138115,
      "num_tokens": 13066727.0,
      "step": 5970
    },
    {
      "entropy": 5.587754631042481,
      "epoch": 0.5929932512901945,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972448218557785,
      "loss": 5.359,
      "mean_token_accuracy": 0.15223066061735152,
      "num_tokens": 13077678.0,
      "step": 5975
    },
    {
      "entropy": 5.636417245864868,
      "epoch": 0.593489479952362,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972392912727882,
      "loss": 5.4694,
      "mean_token_accuracy": 0.14147043600678444,
      "num_tokens": 13088612.0,
      "step": 5980
    },
    {
      "entropy": 5.623294258117676,
      "epoch": 0.5939857086145296,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972337551787425,
      "loss": 5.4749,
      "mean_token_accuracy": 0.1500108703970909,
      "num_tokens": 13100960.0,
      "step": 5985
    },
    {
      "entropy": 5.569150495529175,
      "epoch": 0.5944819372766971,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972282135737785,
      "loss": 5.4197,
      "mean_token_accuracy": 0.15503498613834382,
      "num_tokens": 13110512.0,
      "step": 5990
    },
    {
      "entropy": 5.602548456192016,
      "epoch": 0.5949781659388647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972226664580337,
      "loss": 5.3897,
      "mean_token_accuracy": 0.15015229508280753,
      "num_tokens": 13120569.0,
      "step": 5995
    },
    {
      "entropy": 5.649443578720093,
      "epoch": 0.5954743946010321,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972171138316456,
      "loss": 5.4942,
      "mean_token_accuracy": 0.14562277495861053,
      "num_tokens": 13132006.0,
      "step": 6000
    },
    {
      "epoch": 0.5954743946010321,
      "eval_entropy": 5.422436697292169,
      "eval_loss": 5.511370658874512,
      "eval_mean_token_accuracy": 0.15275962095832452,
      "eval_num_tokens": 13132006.0,
      "eval_runtime": 20.2578,
      "eval_samples_per_second": 1662.025,
      "eval_steps_per_second": 207.772,
      "step": 6000
    },
    {
      "entropy": 5.573627519607544,
      "epoch": 0.5959706232631997,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972115556947518,
      "loss": 5.3859,
      "mean_token_accuracy": 0.1525780588388443,
      "num_tokens": 13143602.0,
      "step": 6005
    },
    {
      "entropy": 5.613383007049561,
      "epoch": 0.5964668519253672,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972059920474904,
      "loss": 5.4619,
      "mean_token_accuracy": 0.14913110584020614,
      "num_tokens": 13154134.0,
      "step": 6010
    },
    {
      "entropy": 5.627599906921387,
      "epoch": 0.5969630805875348,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972004228899991,
      "loss": 5.4624,
      "mean_token_accuracy": 0.15466451346874238,
      "num_tokens": 13163460.0,
      "step": 6015
    },
    {
      "entropy": 5.685395574569702,
      "epoch": 0.5974593092497023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971948482224162,
      "loss": 5.5288,
      "mean_token_accuracy": 0.15045984089374542,
      "num_tokens": 13173199.0,
      "step": 6020
    },
    {
      "entropy": 5.620848274230957,
      "epoch": 0.5979555379118698,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004971892680448798,
      "loss": 5.4496,
      "mean_token_accuracy": 0.1518327549099922,
      "num_tokens": 13183713.0,
      "step": 6025
    },
    {
      "entropy": 5.620371580123901,
      "epoch": 0.5984517665740373,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971836823575283,
      "loss": 5.3985,
      "mean_token_accuracy": 0.14758674502372743,
      "num_tokens": 13194483.0,
      "step": 6030
    },
    {
      "entropy": 5.702054691314697,
      "epoch": 0.5989479952362048,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971780911605003,
      "loss": 5.5054,
      "mean_token_accuracy": 0.14447234943509102,
      "num_tokens": 13206876.0,
      "step": 6035
    },
    {
      "entropy": 5.593558311462402,
      "epoch": 0.5994442238983724,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971724944539343,
      "loss": 5.4523,
      "mean_token_accuracy": 0.14900655075907707,
      "num_tokens": 13219196.0,
      "step": 6040
    },
    {
      "entropy": 5.5757408618927,
      "epoch": 0.5999404525605398,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004971668922379692,
      "loss": 5.475,
      "mean_token_accuracy": 0.14579279869794845,
      "num_tokens": 13230004.0,
      "step": 6045
    },
    {
      "entropy": 5.698142099380493,
      "epoch": 0.6004366812227074,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497161284512744,
      "loss": 5.462,
      "mean_token_accuracy": 0.15078376680612565,
      "num_tokens": 13240760.0,
      "step": 6050
    },
    {
      "entropy": 5.607710647583008,
      "epoch": 0.6009329098848749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971556712783976,
      "loss": 5.3324,
      "mean_token_accuracy": 0.16373973339796066,
      "num_tokens": 13250789.0,
      "step": 6055
    },
    {
      "entropy": 5.562711429595947,
      "epoch": 0.6014291385470425,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971500525350692,
      "loss": 5.4796,
      "mean_token_accuracy": 0.1504553124308586,
      "num_tokens": 13261502.0,
      "step": 6060
    },
    {
      "entropy": 5.641306829452515,
      "epoch": 0.60192536720921,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004971444282828983,
      "loss": 5.4831,
      "mean_token_accuracy": 0.15213162153959275,
      "num_tokens": 13272657.0,
      "step": 6065
    },
    {
      "entropy": 5.691198205947876,
      "epoch": 0.6024215958713776,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004971387985220242,
      "loss": 5.4936,
      "mean_token_accuracy": 0.14803667441010476,
      "num_tokens": 13285418.0,
      "step": 6070
    },
    {
      "entropy": 5.613701295852661,
      "epoch": 0.602917824533545,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971331632525866,
      "loss": 5.4352,
      "mean_token_accuracy": 0.1513635665178299,
      "num_tokens": 13296640.0,
      "step": 6075
    },
    {
      "entropy": 5.6365883350372314,
      "epoch": 0.6034140531957126,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497127522474725,
      "loss": 5.4874,
      "mean_token_accuracy": 0.15728661566972732,
      "num_tokens": 13307418.0,
      "step": 6080
    },
    {
      "entropy": 5.6334559440612795,
      "epoch": 0.6039102818578801,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971218761885797,
      "loss": 5.3789,
      "mean_token_accuracy": 0.15841340124607087,
      "num_tokens": 13317637.0,
      "step": 6085
    },
    {
      "entropy": 5.578281593322754,
      "epoch": 0.6044065105200477,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971162243942904,
      "loss": 5.465,
      "mean_token_accuracy": 0.15196214914321898,
      "num_tokens": 13329472.0,
      "step": 6090
    },
    {
      "entropy": 5.661563444137573,
      "epoch": 0.6049027391822152,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004971105670919975,
      "loss": 5.4588,
      "mean_token_accuracy": 0.14547687917947769,
      "num_tokens": 13340797.0,
      "step": 6095
    },
    {
      "entropy": 5.618959856033325,
      "epoch": 0.6053989678443827,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497104904281841,
      "loss": 5.3697,
      "mean_token_accuracy": 0.14840306639671325,
      "num_tokens": 13350389.0,
      "step": 6100
    },
    {
      "entropy": 5.519381332397461,
      "epoch": 0.6058951965065502,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970992359639616,
      "loss": 5.4029,
      "mean_token_accuracy": 0.15417401790618895,
      "num_tokens": 13361763.0,
      "step": 6105
    },
    {
      "entropy": 5.600857448577881,
      "epoch": 0.6063914251687177,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970935621384996,
      "loss": 5.419,
      "mean_token_accuracy": 0.1529591992497444,
      "num_tokens": 13372229.0,
      "step": 6110
    },
    {
      "entropy": 5.731998157501221,
      "epoch": 0.6068876538308853,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004970878828055959,
      "loss": 5.5794,
      "mean_token_accuracy": 0.1385640799999237,
      "num_tokens": 13384568.0,
      "step": 6115
    },
    {
      "entropy": 5.577160215377807,
      "epoch": 0.6073838824930528,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970821979653912,
      "loss": 5.406,
      "mean_token_accuracy": 0.15079507678747178,
      "num_tokens": 13396093.0,
      "step": 6120
    },
    {
      "entropy": 5.6627274513244625,
      "epoch": 0.6078801111552203,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970765076180265,
      "loss": 5.575,
      "mean_token_accuracy": 0.14542568773031234,
      "num_tokens": 13407061.0,
      "step": 6125
    },
    {
      "entropy": 5.523578262329101,
      "epoch": 0.6083763398173878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970708117636431,
      "loss": 5.3021,
      "mean_token_accuracy": 0.1549241930246353,
      "num_tokens": 13417310.0,
      "step": 6130
    },
    {
      "entropy": 5.6133385181427,
      "epoch": 0.6088725684795554,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497065110402382,
      "loss": 5.4296,
      "mean_token_accuracy": 0.14508086442947388,
      "num_tokens": 13427187.0,
      "step": 6135
    },
    {
      "entropy": 5.603012371063232,
      "epoch": 0.6093687971417229,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970594035343846,
      "loss": 5.3929,
      "mean_token_accuracy": 0.15680171102285384,
      "num_tokens": 13439345.0,
      "step": 6140
    },
    {
      "entropy": 5.6177229404449465,
      "epoch": 0.6098650258038905,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970536911597926,
      "loss": 5.4327,
      "mean_token_accuracy": 0.15486124008893967,
      "num_tokens": 13448733.0,
      "step": 6145
    },
    {
      "entropy": 5.5633306980133055,
      "epoch": 0.6103612544660579,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970479732787476,
      "loss": 5.3739,
      "mean_token_accuracy": 0.1564572237432003,
      "num_tokens": 13460384.0,
      "step": 6150
    },
    {
      "entropy": 5.593255662918091,
      "epoch": 0.6108574831282255,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970422498913911,
      "loss": 5.4762,
      "mean_token_accuracy": 0.1437652476131916,
      "num_tokens": 13471573.0,
      "step": 6155
    },
    {
      "entropy": 5.581676054000854,
      "epoch": 0.611353711790393,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970365209978655,
      "loss": 5.4188,
      "mean_token_accuracy": 0.14645414799451828,
      "num_tokens": 13481550.0,
      "step": 6160
    },
    {
      "entropy": 5.605168294906616,
      "epoch": 0.6118499404525606,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004970307865983124,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1477748289704323,
      "num_tokens": 13491980.0,
      "step": 6165
    },
    {
      "entropy": 5.616139888763428,
      "epoch": 0.6123461691147281,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004970250466928744,
      "loss": 5.4799,
      "mean_token_accuracy": 0.15071297883987428,
      "num_tokens": 13503563.0,
      "step": 6170
    },
    {
      "entropy": 5.580693626403809,
      "epoch": 0.6128423977768956,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970193012816936,
      "loss": 5.4451,
      "mean_token_accuracy": 0.1581176519393921,
      "num_tokens": 13514123.0,
      "step": 6175
    },
    {
      "entropy": 5.572427034378052,
      "epoch": 0.6133386264390631,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004970135503649125,
      "loss": 5.4319,
      "mean_token_accuracy": 0.1552307575941086,
      "num_tokens": 13524772.0,
      "step": 6180
    },
    {
      "entropy": 5.522135257720947,
      "epoch": 0.6138348551012306,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970077939426738,
      "loss": 5.3449,
      "mean_token_accuracy": 0.1645152226090431,
      "num_tokens": 13535423.0,
      "step": 6185
    },
    {
      "entropy": 5.634279537200928,
      "epoch": 0.6143310837633982,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970020320151202,
      "loss": 5.4364,
      "mean_token_accuracy": 0.14737826734781265,
      "num_tokens": 13546088.0,
      "step": 6190
    },
    {
      "entropy": 5.600497722625732,
      "epoch": 0.6148273124255657,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004969962645823946,
      "loss": 5.3783,
      "mean_token_accuracy": 0.15073917508125306,
      "num_tokens": 13557035.0,
      "step": 6195
    },
    {
      "entropy": 5.653165054321289,
      "epoch": 0.6153235410877332,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049699049164464,
      "loss": 5.5357,
      "mean_token_accuracy": 0.149424447119236,
      "num_tokens": 13568346.0,
      "step": 6200
    },
    {
      "entropy": 5.64992127418518,
      "epoch": 0.6158197697499007,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969847132019995,
      "loss": 5.3988,
      "mean_token_accuracy": 0.14995585829019548,
      "num_tokens": 13579475.0,
      "step": 6205
    },
    {
      "entropy": 5.528518295288086,
      "epoch": 0.6163159984120683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969789292546164,
      "loss": 5.4579,
      "mean_token_accuracy": 0.151688289642334,
      "num_tokens": 13591944.0,
      "step": 6210
    },
    {
      "entropy": 5.556851005554199,
      "epoch": 0.6168122270742358,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004969731398026343,
      "loss": 5.4266,
      "mean_token_accuracy": 0.15476437509059907,
      "num_tokens": 13603464.0,
      "step": 6215
    },
    {
      "entropy": 5.682210302352905,
      "epoch": 0.6173084557364034,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004969673448461966,
      "loss": 5.4124,
      "mean_token_accuracy": 0.15477975383400916,
      "num_tokens": 13614670.0,
      "step": 6220
    },
    {
      "entropy": 5.598567533493042,
      "epoch": 0.6178046843985708,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000496961544385447,
      "loss": 5.4947,
      "mean_token_accuracy": 0.14840939193964003,
      "num_tokens": 13624963.0,
      "step": 6225
    },
    {
      "entropy": 5.594861650466919,
      "epoch": 0.6183009130607384,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004969557384205295,
      "loss": 5.486,
      "mean_token_accuracy": 0.14785611778497695,
      "num_tokens": 13636950.0,
      "step": 6230
    },
    {
      "entropy": 5.635034656524658,
      "epoch": 0.6187971417229059,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004969499269515878,
      "loss": 5.3985,
      "mean_token_accuracy": 0.15143660604953765,
      "num_tokens": 13647462.0,
      "step": 6235
    },
    {
      "entropy": 5.555815505981445,
      "epoch": 0.6192933703850735,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969441099787663,
      "loss": 5.361,
      "mean_token_accuracy": 0.1597166582942009,
      "num_tokens": 13657855.0,
      "step": 6240
    },
    {
      "entropy": 5.579755067825317,
      "epoch": 0.619789599047241,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496938287502209,
      "loss": 5.484,
      "mean_token_accuracy": 0.15499924570322038,
      "num_tokens": 13668942.0,
      "step": 6245
    },
    {
      "entropy": 5.585437917709351,
      "epoch": 0.6202858277094085,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969324595220605,
      "loss": 5.4137,
      "mean_token_accuracy": 0.1568784758448601,
      "num_tokens": 13678515.0,
      "step": 6250
    },
    {
      "entropy": 5.68235936164856,
      "epoch": 0.620782056371576,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004969266260384652,
      "loss": 5.4688,
      "mean_token_accuracy": 0.14953088015317917,
      "num_tokens": 13690413.0,
      "step": 6255
    },
    {
      "entropy": 5.654050397872925,
      "epoch": 0.6212782850337435,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969207870515678,
      "loss": 5.4367,
      "mean_token_accuracy": 0.15436980724334717,
      "num_tokens": 13701302.0,
      "step": 6260
    },
    {
      "entropy": 5.670112943649292,
      "epoch": 0.6217745136959111,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000496914942561513,
      "loss": 5.5758,
      "mean_token_accuracy": 0.1467592351138592,
      "num_tokens": 13711335.0,
      "step": 6265
    },
    {
      "entropy": 5.606876993179322,
      "epoch": 0.6222707423580786,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969090925684459,
      "loss": 5.3978,
      "mean_token_accuracy": 0.1508271038532257,
      "num_tokens": 13721470.0,
      "step": 6270
    },
    {
      "entropy": 5.584316396713257,
      "epoch": 0.6227669710202461,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004969032370725114,
      "loss": 5.4103,
      "mean_token_accuracy": 0.14955755472183227,
      "num_tokens": 13731397.0,
      "step": 6275
    },
    {
      "entropy": 5.588260984420776,
      "epoch": 0.6232631996824136,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968973760738548,
      "loss": 5.3771,
      "mean_token_accuracy": 0.15707318484783173,
      "num_tokens": 13741868.0,
      "step": 6280
    },
    {
      "entropy": 5.571529197692871,
      "epoch": 0.6237594283445812,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968915095726216,
      "loss": 5.4398,
      "mean_token_accuracy": 0.1547071374952793,
      "num_tokens": 13751538.0,
      "step": 6285
    },
    {
      "entropy": 5.561303520202637,
      "epoch": 0.6242556570067487,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004968856375689568,
      "loss": 5.4256,
      "mean_token_accuracy": 0.1495334230363369,
      "num_tokens": 13762639.0,
      "step": 6290
    },
    {
      "entropy": 5.588180255889893,
      "epoch": 0.6247518856689163,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004968797600630065,
      "loss": 5.3457,
      "mean_token_accuracy": 0.1600557804107666,
      "num_tokens": 13772721.0,
      "step": 6295
    },
    {
      "entropy": 5.598163175582886,
      "epoch": 0.6252481143310837,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968738770549162,
      "loss": 5.4853,
      "mean_token_accuracy": 0.14522723704576493,
      "num_tokens": 13783520.0,
      "step": 6300
    },
    {
      "entropy": 5.743756866455078,
      "epoch": 0.6257443429932513,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968679885448318,
      "loss": 5.5722,
      "mean_token_accuracy": 0.15225543230772018,
      "num_tokens": 13793814.0,
      "step": 6305
    },
    {
      "entropy": 5.714323902130127,
      "epoch": 0.6262405716554188,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004968620945328995,
      "loss": 5.5838,
      "mean_token_accuracy": 0.14561453983187675,
      "num_tokens": 13806298.0,
      "step": 6310
    },
    {
      "entropy": 5.580462455749512,
      "epoch": 0.6267368003175864,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968561950192652,
      "loss": 5.4104,
      "mean_token_accuracy": 0.1472599647939205,
      "num_tokens": 13816585.0,
      "step": 6315
    },
    {
      "entropy": 5.61671552658081,
      "epoch": 0.6272330289797539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968502900040753,
      "loss": 5.4166,
      "mean_token_accuracy": 0.14943759143352509,
      "num_tokens": 13827670.0,
      "step": 6320
    },
    {
      "entropy": 5.661916255950928,
      "epoch": 0.6277292576419214,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968443794874764,
      "loss": 5.3387,
      "mean_token_accuracy": 0.15909873098134994,
      "num_tokens": 13838601.0,
      "step": 6325
    },
    {
      "entropy": 5.703682804107666,
      "epoch": 0.6282254863040889,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004968384634696149,
      "loss": 5.5305,
      "mean_token_accuracy": 0.1406520277261734,
      "num_tokens": 13849869.0,
      "step": 6330
    },
    {
      "entropy": 5.615419912338257,
      "epoch": 0.6287217149662565,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968325419506375,
      "loss": 5.4702,
      "mean_token_accuracy": 0.14650243520736694,
      "num_tokens": 13861930.0,
      "step": 6335
    },
    {
      "entropy": 5.518544912338257,
      "epoch": 0.629217943628424,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968266149306909,
      "loss": 5.3331,
      "mean_token_accuracy": 0.16056502014398574,
      "num_tokens": 13873795.0,
      "step": 6340
    },
    {
      "entropy": 5.594861268997192,
      "epoch": 0.6297141722905915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968206824099224,
      "loss": 5.3894,
      "mean_token_accuracy": 0.15113001614809035,
      "num_tokens": 13883479.0,
      "step": 6345
    },
    {
      "entropy": 5.602398157119751,
      "epoch": 0.630210400952759,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004968147443884789,
      "loss": 5.4465,
      "mean_token_accuracy": 0.14664751142263413,
      "num_tokens": 13894981.0,
      "step": 6350
    },
    {
      "entropy": 5.58678822517395,
      "epoch": 0.6307066296149265,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968088008665078,
      "loss": 5.4266,
      "mean_token_accuracy": 0.15518708080053328,
      "num_tokens": 13906414.0,
      "step": 6355
    },
    {
      "entropy": 5.610052633285522,
      "epoch": 0.6312028582770941,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968028518441563,
      "loss": 5.4296,
      "mean_token_accuracy": 0.150247460603714,
      "num_tokens": 13917500.0,
      "step": 6360
    },
    {
      "entropy": 5.621783208847046,
      "epoch": 0.6316990869392616,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967968973215721,
      "loss": 5.425,
      "mean_token_accuracy": 0.15201058536767958,
      "num_tokens": 13927843.0,
      "step": 6365
    },
    {
      "entropy": 5.538600969314575,
      "epoch": 0.6321953156014292,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967909372989027,
      "loss": 5.3312,
      "mean_token_accuracy": 0.15451033264398575,
      "num_tokens": 13937939.0,
      "step": 6370
    },
    {
      "entropy": 5.582266139984131,
      "epoch": 0.6326915442635966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967849717762959,
      "loss": 5.4332,
      "mean_token_accuracy": 0.1549287348985672,
      "num_tokens": 13951374.0,
      "step": 6375
    },
    {
      "entropy": 5.573072290420532,
      "epoch": 0.6331877729257642,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967790007538999,
      "loss": 5.3643,
      "mean_token_accuracy": 0.15843954533338547,
      "num_tokens": 13963739.0,
      "step": 6380
    },
    {
      "entropy": 5.495598268508911,
      "epoch": 0.6336840015879317,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967730242318624,
      "loss": 5.3779,
      "mean_token_accuracy": 0.15630589425563812,
      "num_tokens": 13974375.0,
      "step": 6385
    },
    {
      "entropy": 5.5434901237487795,
      "epoch": 0.6341802302500993,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496767042210332,
      "loss": 5.3388,
      "mean_token_accuracy": 0.15538083761930466,
      "num_tokens": 13984379.0,
      "step": 6390
    },
    {
      "entropy": 5.580129766464234,
      "epoch": 0.6346764589122668,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967610546894566,
      "loss": 5.3526,
      "mean_token_accuracy": 0.15347034186124803,
      "num_tokens": 13994638.0,
      "step": 6395
    },
    {
      "entropy": 5.617799615859985,
      "epoch": 0.6351726875744343,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496755061669385,
      "loss": 5.4792,
      "mean_token_accuracy": 0.1480176031589508,
      "num_tokens": 14005835.0,
      "step": 6400
    },
    {
      "entropy": 5.558763599395752,
      "epoch": 0.6356689162366018,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004967490631502656,
      "loss": 5.3644,
      "mean_token_accuracy": 0.15586011707782746,
      "num_tokens": 14017549.0,
      "step": 6405
    },
    {
      "entropy": 5.576519632339478,
      "epoch": 0.6361651448987694,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967430591322474,
      "loss": 5.4274,
      "mean_token_accuracy": 0.15214964523911476,
      "num_tokens": 14028417.0,
      "step": 6410
    },
    {
      "entropy": 5.5468714237213135,
      "epoch": 0.6366613735609369,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967370496154791,
      "loss": 5.3589,
      "mean_token_accuracy": 0.15957570523023606,
      "num_tokens": 14038535.0,
      "step": 6415
    },
    {
      "entropy": 5.632842397689819,
      "epoch": 0.6371576022231044,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967310346001098,
      "loss": 5.4272,
      "mean_token_accuracy": 0.15456130504608154,
      "num_tokens": 14050082.0,
      "step": 6420
    },
    {
      "entropy": 5.596773958206176,
      "epoch": 0.6376538308852719,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967250140862885,
      "loss": 5.4554,
      "mean_token_accuracy": 0.15169748216867446,
      "num_tokens": 14061419.0,
      "step": 6425
    },
    {
      "entropy": 5.625483798980713,
      "epoch": 0.6381500595474394,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004967189880741647,
      "loss": 5.4217,
      "mean_token_accuracy": 0.15191691145300865,
      "num_tokens": 14073360.0,
      "step": 6430
    },
    {
      "entropy": 5.610320281982422,
      "epoch": 0.638646288209607,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967129565638878,
      "loss": 5.457,
      "mean_token_accuracy": 0.15132322907447815,
      "num_tokens": 14083665.0,
      "step": 6435
    },
    {
      "entropy": 5.5531909465789795,
      "epoch": 0.6391425168717745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967069195556073,
      "loss": 5.4025,
      "mean_token_accuracy": 0.1516549125313759,
      "num_tokens": 14094968.0,
      "step": 6440
    },
    {
      "entropy": 5.602369070053101,
      "epoch": 0.6396387455339421,
      "grad_norm": 1.0,
      "learning_rate": 0.000496700877049473,
      "loss": 5.4244,
      "mean_token_accuracy": 0.15252992063760756,
      "num_tokens": 14105325.0,
      "step": 6445
    },
    {
      "entropy": 5.626211166381836,
      "epoch": 0.6401349741961095,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966948290456345,
      "loss": 5.4431,
      "mean_token_accuracy": 0.15126971304416656,
      "num_tokens": 14115764.0,
      "step": 6450
    },
    {
      "entropy": 5.544282245635986,
      "epoch": 0.6406312028582771,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966887755442421,
      "loss": 5.404,
      "mean_token_accuracy": 0.1501368209719658,
      "num_tokens": 14126389.0,
      "step": 6455
    },
    {
      "entropy": 5.612423992156982,
      "epoch": 0.6411274315204446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966827165454456,
      "loss": 5.4013,
      "mean_token_accuracy": 0.15302912294864654,
      "num_tokens": 14136241.0,
      "step": 6460
    },
    {
      "entropy": 5.59951376914978,
      "epoch": 0.6416236601826122,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966766520493956,
      "loss": 5.3783,
      "mean_token_accuracy": 0.14901676028966904,
      "num_tokens": 14147641.0,
      "step": 6465
    },
    {
      "entropy": 5.544262027740478,
      "epoch": 0.6421198888447797,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966705820562422,
      "loss": 5.3834,
      "mean_token_accuracy": 0.16056301146745683,
      "num_tokens": 14158780.0,
      "step": 6470
    },
    {
      "entropy": 5.541188764572143,
      "epoch": 0.6426161175069472,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004966645065661361,
      "loss": 5.3562,
      "mean_token_accuracy": 0.15511102229356766,
      "num_tokens": 14170672.0,
      "step": 6475
    },
    {
      "entropy": 5.589229774475098,
      "epoch": 0.6431123461691147,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966584255792279,
      "loss": 5.2743,
      "mean_token_accuracy": 0.15746648907661437,
      "num_tokens": 14179835.0,
      "step": 6480
    },
    {
      "entropy": 5.553354358673095,
      "epoch": 0.6436085748312823,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966523390956683,
      "loss": 5.4193,
      "mean_token_accuracy": 0.1556967630982399,
      "num_tokens": 14190624.0,
      "step": 6485
    },
    {
      "entropy": 5.630648517608643,
      "epoch": 0.6441048034934498,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966462471156083,
      "loss": 5.4001,
      "mean_token_accuracy": 0.1567710518836975,
      "num_tokens": 14201350.0,
      "step": 6490
    },
    {
      "entropy": 5.602198457717895,
      "epoch": 0.6446010321556173,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966401496391989,
      "loss": 5.3401,
      "mean_token_accuracy": 0.15134084671735765,
      "num_tokens": 14212837.0,
      "step": 6495
    },
    {
      "entropy": 5.52803053855896,
      "epoch": 0.6450972608177848,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966340466665915,
      "loss": 5.3896,
      "mean_token_accuracy": 0.15534065216779708,
      "num_tokens": 14223479.0,
      "step": 6500
    },
    {
      "entropy": 5.564976787567138,
      "epoch": 0.6455934894799523,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004966279381979373,
      "loss": 5.4066,
      "mean_token_accuracy": 0.1574067384004593,
      "num_tokens": 14233735.0,
      "step": 6505
    },
    {
      "entropy": 5.731327676773072,
      "epoch": 0.6460897181421199,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004966218242333879,
      "loss": 5.5182,
      "mean_token_accuracy": 0.15012315511703492,
      "num_tokens": 14246092.0,
      "step": 6510
    },
    {
      "entropy": 5.634858226776123,
      "epoch": 0.6465859468042874,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966157047730945,
      "loss": 5.4625,
      "mean_token_accuracy": 0.1493428483605385,
      "num_tokens": 14257652.0,
      "step": 6515
    },
    {
      "entropy": 5.644312047958374,
      "epoch": 0.647082175466455,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004966095798172095,
      "loss": 5.4177,
      "mean_token_accuracy": 0.1520577237010002,
      "num_tokens": 14268033.0,
      "step": 6520
    },
    {
      "entropy": 5.681197929382324,
      "epoch": 0.6475784041286224,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966034493658843,
      "loss": 5.6035,
      "mean_token_accuracy": 0.14491873234510422,
      "num_tokens": 14279224.0,
      "step": 6525
    },
    {
      "entropy": 5.742174053192139,
      "epoch": 0.64807463279079,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965973134192711,
      "loss": 5.5101,
      "mean_token_accuracy": 0.14795609340071678,
      "num_tokens": 14288908.0,
      "step": 6530
    },
    {
      "entropy": 5.569133996963501,
      "epoch": 0.6485708614529575,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000496591171977522,
      "loss": 5.4686,
      "mean_token_accuracy": 0.14720422625541688,
      "num_tokens": 14299654.0,
      "step": 6535
    },
    {
      "entropy": 5.572997951507569,
      "epoch": 0.6490670901151251,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004965850250407893,
      "loss": 5.4,
      "mean_token_accuracy": 0.15774376839399337,
      "num_tokens": 14310689.0,
      "step": 6540
    },
    {
      "entropy": 5.583591318130493,
      "epoch": 0.6495633187772926,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965788726092256,
      "loss": 5.4173,
      "mean_token_accuracy": 0.1523214116692543,
      "num_tokens": 14321913.0,
      "step": 6545
    },
    {
      "entropy": 5.608176183700562,
      "epoch": 0.6500595474394602,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004965727146829832,
      "loss": 5.3655,
      "mean_token_accuracy": 0.1537041187286377,
      "num_tokens": 14332869.0,
      "step": 6550
    },
    {
      "entropy": 5.570441865921021,
      "epoch": 0.6505557761016276,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004965665512622148,
      "loss": 5.3251,
      "mean_token_accuracy": 0.16085829287767411,
      "num_tokens": 14344363.0,
      "step": 6555
    },
    {
      "entropy": 5.499426937103271,
      "epoch": 0.6510520047637952,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965603823470735,
      "loss": 5.378,
      "mean_token_accuracy": 0.16010768413543702,
      "num_tokens": 14355149.0,
      "step": 6560
    },
    {
      "entropy": 5.6788856983184814,
      "epoch": 0.6515482334259627,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965542079377122,
      "loss": 5.4713,
      "mean_token_accuracy": 0.14611513912677765,
      "num_tokens": 14366833.0,
      "step": 6565
    },
    {
      "entropy": 5.612633800506591,
      "epoch": 0.6520444620881302,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965480280342839,
      "loss": 5.4038,
      "mean_token_accuracy": 0.15392701625823973,
      "num_tokens": 14376945.0,
      "step": 6570
    },
    {
      "entropy": 5.566470623016357,
      "epoch": 0.6525406907502977,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004965418426369419,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1530841112136841,
      "num_tokens": 14388376.0,
      "step": 6575
    },
    {
      "entropy": 5.608035755157471,
      "epoch": 0.6530369194124652,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965356517458395,
      "loss": 5.3687,
      "mean_token_accuracy": 0.15580754205584527,
      "num_tokens": 14398826.0,
      "step": 6580
    },
    {
      "entropy": 5.636288452148437,
      "epoch": 0.6535331480746328,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965294553611303,
      "loss": 5.451,
      "mean_token_accuracy": 0.15208111554384232,
      "num_tokens": 14410758.0,
      "step": 6585
    },
    {
      "entropy": 5.595287704467774,
      "epoch": 0.6540293767368003,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496523253482968,
      "loss": 5.3883,
      "mean_token_accuracy": 0.15530412644147873,
      "num_tokens": 14420097.0,
      "step": 6590
    },
    {
      "entropy": 5.568693971633911,
      "epoch": 0.6545256053989679,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965170461115065,
      "loss": 5.4329,
      "mean_token_accuracy": 0.1518382251262665,
      "num_tokens": 14430998.0,
      "step": 6595
    },
    {
      "entropy": 5.625651693344116,
      "epoch": 0.6550218340611353,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965108332468994,
      "loss": 5.3465,
      "mean_token_accuracy": 0.16162973493337632,
      "num_tokens": 14442234.0,
      "step": 6600
    },
    {
      "entropy": 5.680300855636597,
      "epoch": 0.6555180627233029,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965046148893011,
      "loss": 5.6349,
      "mean_token_accuracy": 0.1367014691233635,
      "num_tokens": 14453270.0,
      "step": 6605
    },
    {
      "entropy": 5.583677101135254,
      "epoch": 0.6560142913854704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964983910388654,
      "loss": 5.3828,
      "mean_token_accuracy": 0.1576682485640049,
      "num_tokens": 14463818.0,
      "step": 6610
    },
    {
      "entropy": 5.64986891746521,
      "epoch": 0.656510520047638,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964921616957471,
      "loss": 5.4774,
      "mean_token_accuracy": 0.1540710985660553,
      "num_tokens": 14476342.0,
      "step": 6615
    },
    {
      "entropy": 5.530845737457275,
      "epoch": 0.6570067487098055,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004964859268601005,
      "loss": 5.3242,
      "mean_token_accuracy": 0.15631238520145416,
      "num_tokens": 14488683.0,
      "step": 6620
    },
    {
      "entropy": 5.546418905258179,
      "epoch": 0.657502977371973,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964796865320801,
      "loss": 5.3223,
      "mean_token_accuracy": 0.1593739539384842,
      "num_tokens": 14499650.0,
      "step": 6625
    },
    {
      "entropy": 5.557725524902343,
      "epoch": 0.6579992060341405,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004964734407118406,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1473859116435051,
      "num_tokens": 14511390.0,
      "step": 6630
    },
    {
      "entropy": 5.594749879837036,
      "epoch": 0.6584954346963081,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004964671893995372,
      "loss": 5.4442,
      "mean_token_accuracy": 0.14892742931842803,
      "num_tokens": 14522973.0,
      "step": 6635
    },
    {
      "entropy": 5.600301265716553,
      "epoch": 0.6589916633584756,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004964609325953245,
      "loss": 5.4165,
      "mean_token_accuracy": 0.15884668976068497,
      "num_tokens": 14534202.0,
      "step": 6640
    },
    {
      "entropy": 5.615841436386108,
      "epoch": 0.6594878920206431,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964546702993581,
      "loss": 5.4381,
      "mean_token_accuracy": 0.1449699208140373,
      "num_tokens": 14545855.0,
      "step": 6645
    },
    {
      "entropy": 5.6032805919647215,
      "epoch": 0.6599841206828106,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496448402511793,
      "loss": 5.3717,
      "mean_token_accuracy": 0.16163230836391448,
      "num_tokens": 14556716.0,
      "step": 6650
    },
    {
      "entropy": 5.540382003784179,
      "epoch": 0.6604803493449781,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004964421292327847,
      "loss": 5.2902,
      "mean_token_accuracy": 0.16156066209077835,
      "num_tokens": 14566680.0,
      "step": 6655
    },
    {
      "entropy": 5.553192901611328,
      "epoch": 0.6609765780071457,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004964358504624887,
      "loss": 5.4265,
      "mean_token_accuracy": 0.15593043565750123,
      "num_tokens": 14578106.0,
      "step": 6660
    },
    {
      "entropy": 5.519644784927368,
      "epoch": 0.6614728066693132,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964295662010608,
      "loss": 5.353,
      "mean_token_accuracy": 0.15554747432470323,
      "num_tokens": 14588601.0,
      "step": 6665
    },
    {
      "entropy": 5.657138395309448,
      "epoch": 0.6619690353314808,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964232764486568,
      "loss": 5.3514,
      "mean_token_accuracy": 0.15236941277980803,
      "num_tokens": 14599686.0,
      "step": 6670
    },
    {
      "entropy": 5.616515159606934,
      "epoch": 0.6624652639936482,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964169812054326,
      "loss": 5.459,
      "mean_token_accuracy": 0.151620514690876,
      "num_tokens": 14610704.0,
      "step": 6675
    },
    {
      "entropy": 5.537838172912598,
      "epoch": 0.6629614926558158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964106804715444,
      "loss": 5.3782,
      "mean_token_accuracy": 0.15941275507211686,
      "num_tokens": 14621004.0,
      "step": 6680
    },
    {
      "entropy": 5.554213762283325,
      "epoch": 0.6634577213179833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964043742471486,
      "loss": 5.3639,
      "mean_token_accuracy": 0.16116376370191574,
      "num_tokens": 14631584.0,
      "step": 6685
    },
    {
      "entropy": 5.567893409729004,
      "epoch": 0.6639539499801509,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963980625324012,
      "loss": 5.4255,
      "mean_token_accuracy": 0.1525879457592964,
      "num_tokens": 14642687.0,
      "step": 6690
    },
    {
      "entropy": 5.584773206710816,
      "epoch": 0.6644501786423184,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496391745327459,
      "loss": 5.386,
      "mean_token_accuracy": 0.15478010177612306,
      "num_tokens": 14652569.0,
      "step": 6695
    },
    {
      "entropy": 5.678552532196045,
      "epoch": 0.664946407304486,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004963854226324785,
      "loss": 5.4397,
      "mean_token_accuracy": 0.15262390524148942,
      "num_tokens": 14664018.0,
      "step": 6700
    },
    {
      "entropy": 5.542888545989991,
      "epoch": 0.6654426359666534,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963790944476167,
      "loss": 5.3158,
      "mean_token_accuracy": 0.16195368468761445,
      "num_tokens": 14673416.0,
      "step": 6705
    },
    {
      "entropy": 5.485993099212647,
      "epoch": 0.665938864628821,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963727607730303,
      "loss": 5.3658,
      "mean_token_accuracy": 0.16045239716768264,
      "num_tokens": 14685393.0,
      "step": 6710
    },
    {
      "entropy": 5.639457654953003,
      "epoch": 0.6664350932909885,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963664216088764,
      "loss": 5.4786,
      "mean_token_accuracy": 0.1556263357400894,
      "num_tokens": 14697133.0,
      "step": 6715
    },
    {
      "entropy": 5.650945854187012,
      "epoch": 0.666931321953156,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004963600769553124,
      "loss": 5.4666,
      "mean_token_accuracy": 0.15382171422243118,
      "num_tokens": 14708340.0,
      "step": 6720
    },
    {
      "entropy": 5.583352375030517,
      "epoch": 0.6674275506153236,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963537268124953,
      "loss": 5.43,
      "mean_token_accuracy": 0.1598139151930809,
      "num_tokens": 14719195.0,
      "step": 6725
    },
    {
      "entropy": 5.547114419937134,
      "epoch": 0.667923779277491,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004963473711805829,
      "loss": 5.4116,
      "mean_token_accuracy": 0.155468587577343,
      "num_tokens": 14730999.0,
      "step": 6730
    },
    {
      "entropy": 5.599129486083984,
      "epoch": 0.6684200079396586,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963410100597326,
      "loss": 5.3471,
      "mean_token_accuracy": 0.1635589048266411,
      "num_tokens": 14741548.0,
      "step": 6735
    },
    {
      "entropy": 5.589991855621338,
      "epoch": 0.6689162366018261,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004963346434501021,
      "loss": 5.4607,
      "mean_token_accuracy": 0.15114777982234956,
      "num_tokens": 14753257.0,
      "step": 6740
    },
    {
      "entropy": 5.587455224990845,
      "epoch": 0.6694124652639937,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004963282713518495,
      "loss": 5.4612,
      "mean_token_accuracy": 0.14838118553161622,
      "num_tokens": 14765340.0,
      "step": 6745
    },
    {
      "entropy": 5.613324213027954,
      "epoch": 0.6699086939261611,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963218937651326,
      "loss": 5.4474,
      "mean_token_accuracy": 0.14588816463947296,
      "num_tokens": 14775650.0,
      "step": 6750
    },
    {
      "entropy": 5.575215244293213,
      "epoch": 0.6704049225883287,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963155106901098,
      "loss": 5.3778,
      "mean_token_accuracy": 0.15298025012016297,
      "num_tokens": 14786991.0,
      "step": 6755
    },
    {
      "entropy": 5.556736421585083,
      "epoch": 0.6709011512504962,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963091221269391,
      "loss": 5.3658,
      "mean_token_accuracy": 0.1534993276000023,
      "num_tokens": 14798992.0,
      "step": 6760
    },
    {
      "entropy": 5.611054706573486,
      "epoch": 0.6713973799126638,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000496302728075779,
      "loss": 5.4636,
      "mean_token_accuracy": 0.15155581384897232,
      "num_tokens": 14809691.0,
      "step": 6765
    },
    {
      "entropy": 5.554108190536499,
      "epoch": 0.6718936085748313,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962963285367882,
      "loss": 5.285,
      "mean_token_accuracy": 0.16349877864122392,
      "num_tokens": 14820217.0,
      "step": 6770
    },
    {
      "entropy": 5.584902906417847,
      "epoch": 0.6723898372369989,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962899235101253,
      "loss": 5.496,
      "mean_token_accuracy": 0.15075885355472565,
      "num_tokens": 14831030.0,
      "step": 6775
    },
    {
      "entropy": 5.545218658447266,
      "epoch": 0.6728860658991663,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004962835129959491,
      "loss": 5.3484,
      "mean_token_accuracy": 0.16029660254716874,
      "num_tokens": 14841422.0,
      "step": 6780
    },
    {
      "entropy": 5.602077770233154,
      "epoch": 0.6733822945613339,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004962770969944186,
      "loss": 5.3533,
      "mean_token_accuracy": 0.15501683801412583,
      "num_tokens": 14852492.0,
      "step": 6785
    },
    {
      "entropy": 5.548709440231323,
      "epoch": 0.6738785232235014,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000496270675505693,
      "loss": 5.3784,
      "mean_token_accuracy": 0.15387199223041534,
      "num_tokens": 14862933.0,
      "step": 6790
    },
    {
      "entropy": 5.53840651512146,
      "epoch": 0.6743747518856689,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962642485299314,
      "loss": 5.3455,
      "mean_token_accuracy": 0.15669092684984207,
      "num_tokens": 14873718.0,
      "step": 6795
    },
    {
      "entropy": 5.604877185821533,
      "epoch": 0.6748709805478365,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962578160672931,
      "loss": 5.4162,
      "mean_token_accuracy": 0.15211003497242928,
      "num_tokens": 14886113.0,
      "step": 6800
    },
    {
      "entropy": 5.591918420791626,
      "epoch": 0.6753672092100039,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962513781179379,
      "loss": 5.4541,
      "mean_token_accuracy": 0.15024468451738357,
      "num_tokens": 14897162.0,
      "step": 6805
    },
    {
      "entropy": 5.62456202507019,
      "epoch": 0.6758634378721715,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004962449346820251,
      "loss": 5.4217,
      "mean_token_accuracy": 0.15283632203936576,
      "num_tokens": 14908387.0,
      "step": 6810
    },
    {
      "entropy": 5.611571502685547,
      "epoch": 0.676359666534339,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004962384857597148,
      "loss": 5.4525,
      "mean_token_accuracy": 0.14822545796632766,
      "num_tokens": 14919163.0,
      "step": 6815
    },
    {
      "entropy": 5.570850801467896,
      "epoch": 0.6768558951965066,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962320313511668,
      "loss": 5.3046,
      "mean_token_accuracy": 0.15631168186664582,
      "num_tokens": 14929431.0,
      "step": 6820
    },
    {
      "entropy": 5.535919570922852,
      "epoch": 0.677352123858674,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962255714565409,
      "loss": 5.3492,
      "mean_token_accuracy": 0.15995125025510787,
      "num_tokens": 14940395.0,
      "step": 6825
    },
    {
      "entropy": 5.456424379348755,
      "epoch": 0.6778483525208416,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962191060759978,
      "loss": 5.3259,
      "mean_token_accuracy": 0.16244170516729356,
      "num_tokens": 14952268.0,
      "step": 6830
    },
    {
      "entropy": 5.67572135925293,
      "epoch": 0.6783445811830091,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004962126352096973,
      "loss": 5.4589,
      "mean_token_accuracy": 0.14956776052713394,
      "num_tokens": 14964880.0,
      "step": 6835
    },
    {
      "entropy": 5.607454681396485,
      "epoch": 0.6788408098451767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962061588578004,
      "loss": 5.4007,
      "mean_token_accuracy": 0.15093457996845244,
      "num_tokens": 14975512.0,
      "step": 6840
    },
    {
      "entropy": 5.538913869857788,
      "epoch": 0.6793370385073442,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961996770204673,
      "loss": 5.402,
      "mean_token_accuracy": 0.15925564169883727,
      "num_tokens": 14986373.0,
      "step": 6845
    },
    {
      "entropy": 5.5862802028656,
      "epoch": 0.6798332671695118,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004961931896978588,
      "loss": 5.3502,
      "mean_token_accuracy": 0.15667034089565277,
      "num_tokens": 14997449.0,
      "step": 6850
    },
    {
      "entropy": 5.546631526947022,
      "epoch": 0.6803294958316792,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961866968901359,
      "loss": 5.3088,
      "mean_token_accuracy": 0.15922699347138405,
      "num_tokens": 15008424.0,
      "step": 6855
    },
    {
      "entropy": 5.587251996994018,
      "epoch": 0.6808257244938468,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961801985974595,
      "loss": 5.4526,
      "mean_token_accuracy": 0.15401181876659392,
      "num_tokens": 15018933.0,
      "step": 6860
    },
    {
      "entropy": 5.631841087341309,
      "epoch": 0.6813219531560143,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961736948199907,
      "loss": 5.4743,
      "mean_token_accuracy": 0.14421806782484053,
      "num_tokens": 15030973.0,
      "step": 6865
    },
    {
      "entropy": 5.556496238708496,
      "epoch": 0.6818181818181818,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004961671855578909,
      "loss": 5.3356,
      "mean_token_accuracy": 0.1502171352505684,
      "num_tokens": 15042412.0,
      "step": 6870
    },
    {
      "entropy": 5.575684309005737,
      "epoch": 0.6823144104803494,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004961606708113216,
      "loss": 5.3379,
      "mean_token_accuracy": 0.1604937121272087,
      "num_tokens": 15053301.0,
      "step": 6875
    },
    {
      "entropy": 5.481336832046509,
      "epoch": 0.6828106391425168,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004961541505804441,
      "loss": 5.3121,
      "mean_token_accuracy": 0.155144040286541,
      "num_tokens": 15064753.0,
      "step": 6880
    },
    {
      "entropy": 5.5831273078918455,
      "epoch": 0.6833068678046844,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961476248654202,
      "loss": 5.4044,
      "mean_token_accuracy": 0.15218231976032257,
      "num_tokens": 15074991.0,
      "step": 6885
    },
    {
      "entropy": 5.583978176116943,
      "epoch": 0.6838030964668519,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961410936664118,
      "loss": 5.4522,
      "mean_token_accuracy": 0.14528602063655854,
      "num_tokens": 15084567.0,
      "step": 6890
    },
    {
      "entropy": 5.613344430923462,
      "epoch": 0.6842993251290195,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961345569835807,
      "loss": 5.455,
      "mean_token_accuracy": 0.15477517545223235,
      "num_tokens": 15096222.0,
      "step": 6895
    },
    {
      "entropy": 5.660449028015137,
      "epoch": 0.684795553791187,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961280148170892,
      "loss": 5.4177,
      "mean_token_accuracy": 0.15607869625091553,
      "num_tokens": 15107019.0,
      "step": 6900
    },
    {
      "entropy": 5.59068489074707,
      "epoch": 0.6852917824533545,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961214671670994,
      "loss": 5.3089,
      "mean_token_accuracy": 0.15733786001801492,
      "num_tokens": 15117695.0,
      "step": 6905
    },
    {
      "entropy": 5.488613224029541,
      "epoch": 0.685788011115522,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961149140337737,
      "loss": 5.3781,
      "mean_token_accuracy": 0.15287320464849471,
      "num_tokens": 15130115.0,
      "step": 6910
    },
    {
      "entropy": 5.576217126846314,
      "epoch": 0.6862842397776896,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961083554172745,
      "loss": 5.3428,
      "mean_token_accuracy": 0.15954004749655723,
      "num_tokens": 15140626.0,
      "step": 6915
    },
    {
      "entropy": 5.618510913848877,
      "epoch": 0.6867804684398571,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961017913177646,
      "loss": 5.429,
      "mean_token_accuracy": 0.1515742838382721,
      "num_tokens": 15151586.0,
      "step": 6920
    },
    {
      "entropy": 5.581950426101685,
      "epoch": 0.6872766971020247,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960952217354066,
      "loss": 5.3653,
      "mean_token_accuracy": 0.15943562984466553,
      "num_tokens": 15162662.0,
      "step": 6925
    },
    {
      "entropy": 5.541510677337646,
      "epoch": 0.6877729257641921,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960886466703636,
      "loss": 5.3562,
      "mean_token_accuracy": 0.1531212091445923,
      "num_tokens": 15172691.0,
      "step": 6930
    },
    {
      "entropy": 5.573947763442993,
      "epoch": 0.6882691544263597,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004960820661227987,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1461188554763794,
      "num_tokens": 15184655.0,
      "step": 6935
    },
    {
      "entropy": 5.638881874084473,
      "epoch": 0.6887653830885272,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004960754800928748,
      "loss": 5.419,
      "mean_token_accuracy": 0.15994274914264678,
      "num_tokens": 15196031.0,
      "step": 6940
    },
    {
      "entropy": 5.645197343826294,
      "epoch": 0.6892616117506947,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960688885807553,
      "loss": 5.4474,
      "mean_token_accuracy": 0.15324670523405076,
      "num_tokens": 15206094.0,
      "step": 6945
    },
    {
      "entropy": 5.5976806640625,
      "epoch": 0.6897578404128623,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004960622915866038,
      "loss": 5.4994,
      "mean_token_accuracy": 0.15229054018855095,
      "num_tokens": 15216758.0,
      "step": 6950
    },
    {
      "entropy": 5.5646665573120115,
      "epoch": 0.6902540690750297,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960556891105839,
      "loss": 5.375,
      "mean_token_accuracy": 0.15717837810516358,
      "num_tokens": 15226594.0,
      "step": 6955
    },
    {
      "entropy": 5.488271903991699,
      "epoch": 0.6907502977371973,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960490811528591,
      "loss": 5.2456,
      "mean_token_accuracy": 0.15999548733234406,
      "num_tokens": 15237225.0,
      "step": 6960
    },
    {
      "entropy": 5.598249912261963,
      "epoch": 0.6912465263993648,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004960424677135935,
      "loss": 5.3848,
      "mean_token_accuracy": 0.15727042257785798,
      "num_tokens": 15248215.0,
      "step": 6965
    },
    {
      "entropy": 5.528061389923096,
      "epoch": 0.6917427550615324,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960358487929511,
      "loss": 5.3614,
      "mean_token_accuracy": 0.1582505911588669,
      "num_tokens": 15258056.0,
      "step": 6970
    },
    {
      "entropy": 5.527980709075928,
      "epoch": 0.6922389837236999,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960292243910958,
      "loss": 5.2742,
      "mean_token_accuracy": 0.1645634390413761,
      "num_tokens": 15268438.0,
      "step": 6975
    },
    {
      "entropy": 5.624602890014648,
      "epoch": 0.6927352123858674,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496022594508192,
      "loss": 5.4432,
      "mean_token_accuracy": 0.15136538296937943,
      "num_tokens": 15279308.0,
      "step": 6980
    },
    {
      "entropy": 5.545179796218872,
      "epoch": 0.6932314410480349,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960159591444041,
      "loss": 5.4041,
      "mean_token_accuracy": 0.1536109194159508,
      "num_tokens": 15290804.0,
      "step": 6985
    },
    {
      "entropy": 5.6160321712493895,
      "epoch": 0.6937276697102025,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960093182998968,
      "loss": 5.3732,
      "mean_token_accuracy": 0.15154838487505912,
      "num_tokens": 15302543.0,
      "step": 6990
    },
    {
      "entropy": 5.5771231174469,
      "epoch": 0.69422389837237,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960026719748346,
      "loss": 5.3697,
      "mean_token_accuracy": 0.1599016435444355,
      "num_tokens": 15312841.0,
      "step": 6995
    },
    {
      "entropy": 5.547192859649658,
      "epoch": 0.6947201270345376,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004959960201693822,
      "loss": 5.3733,
      "mean_token_accuracy": 0.15269829481840133,
      "num_tokens": 15323940.0,
      "step": 7000
    },
    {
      "entropy": 5.655949544906616,
      "epoch": 0.695216355696705,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959893628837048,
      "loss": 5.5048,
      "mean_token_accuracy": 0.1433499850332737,
      "num_tokens": 15335094.0,
      "step": 7005
    },
    {
      "entropy": 5.625252151489258,
      "epoch": 0.6957125843588726,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959827001179675,
      "loss": 5.4481,
      "mean_token_accuracy": 0.14871692061424255,
      "num_tokens": 15346383.0,
      "step": 7010
    },
    {
      "entropy": 5.571574783325195,
      "epoch": 0.6962088130210401,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959760318723353,
      "loss": 5.3402,
      "mean_token_accuracy": 0.1649706944823265,
      "num_tokens": 15356243.0,
      "step": 7015
    },
    {
      "entropy": 5.553963088989258,
      "epoch": 0.6967050416832076,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004959693581469736,
      "loss": 5.4356,
      "mean_token_accuracy": 0.1529014952480793,
      "num_tokens": 15367706.0,
      "step": 7020
    },
    {
      "entropy": 5.585008764266968,
      "epoch": 0.6972012703453752,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959626789420481,
      "loss": 5.3015,
      "mean_token_accuracy": 0.1597673237323761,
      "num_tokens": 15377208.0,
      "step": 7025
    },
    {
      "entropy": 5.537755393981934,
      "epoch": 0.6976974990075426,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004959559942577243,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1594744563102722,
      "num_tokens": 15387901.0,
      "step": 7030
    },
    {
      "entropy": 5.558951044082642,
      "epoch": 0.6981937276697102,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959493040941678,
      "loss": 5.3555,
      "mean_token_accuracy": 0.1569546103477478,
      "num_tokens": 15398562.0,
      "step": 7035
    },
    {
      "entropy": 5.570696449279785,
      "epoch": 0.6986899563318777,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959426084515448,
      "loss": 5.2944,
      "mean_token_accuracy": 0.16199443489313126,
      "num_tokens": 15409498.0,
      "step": 7040
    },
    {
      "entropy": 5.569762754440307,
      "epoch": 0.6991861849940453,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004959359073300212,
      "loss": 5.3151,
      "mean_token_accuracy": 0.16169282048940659,
      "num_tokens": 15419737.0,
      "step": 7045
    },
    {
      "entropy": 5.581260728836059,
      "epoch": 0.6996824136562128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959292007297631,
      "loss": 5.4568,
      "mean_token_accuracy": 0.1515838697552681,
      "num_tokens": 15431384.0,
      "step": 7050
    },
    {
      "entropy": 5.5438906192779545,
      "epoch": 0.7001786423183803,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495922488650937,
      "loss": 5.3453,
      "mean_token_accuracy": 0.15134536921977998,
      "num_tokens": 15441253.0,
      "step": 7055
    },
    {
      "entropy": 5.492924785614013,
      "epoch": 0.7006748709805478,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959157710937092,
      "loss": 5.3032,
      "mean_token_accuracy": 0.15900238305330278,
      "num_tokens": 15452388.0,
      "step": 7060
    },
    {
      "entropy": 5.549806785583496,
      "epoch": 0.7011710996427154,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004959090480582462,
      "loss": 5.4027,
      "mean_token_accuracy": 0.1584886997938156,
      "num_tokens": 15463192.0,
      "step": 7065
    },
    {
      "entropy": 5.559626722335816,
      "epoch": 0.7016673283048829,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004959023195447148,
      "loss": 5.3551,
      "mean_token_accuracy": 0.1578907012939453,
      "num_tokens": 15473978.0,
      "step": 7070
    },
    {
      "entropy": 5.5646559715271,
      "epoch": 0.7021635569670505,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495895585553282,
      "loss": 5.4126,
      "mean_token_accuracy": 0.1546211361885071,
      "num_tokens": 15484466.0,
      "step": 7075
    },
    {
      "entropy": 5.555582189559937,
      "epoch": 0.7026597856292179,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958888460841147,
      "loss": 5.3161,
      "mean_token_accuracy": 0.15918228775262833,
      "num_tokens": 15495407.0,
      "step": 7080
    },
    {
      "entropy": 5.556728410720825,
      "epoch": 0.7031560142913855,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958821011373798,
      "loss": 5.3147,
      "mean_token_accuracy": 0.15541466027498246,
      "num_tokens": 15506012.0,
      "step": 7085
    },
    {
      "entropy": 5.566561698913574,
      "epoch": 0.703652242953553,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004958753507132449,
      "loss": 5.4571,
      "mean_token_accuracy": 0.1495395615696907,
      "num_tokens": 15516829.0,
      "step": 7090
    },
    {
      "entropy": 5.662977123260498,
      "epoch": 0.7041484716157205,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004958685948118772,
      "loss": 5.4334,
      "mean_token_accuracy": 0.14862314313650132,
      "num_tokens": 15528611.0,
      "step": 7095
    },
    {
      "entropy": 5.5765410423278805,
      "epoch": 0.7046447002778881,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958618334334443,
      "loss": 5.375,
      "mean_token_accuracy": 0.15312562584877015,
      "num_tokens": 15540010.0,
      "step": 7100
    },
    {
      "entropy": 5.632585763931274,
      "epoch": 0.7051409289400555,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958550665781138,
      "loss": 5.4256,
      "mean_token_accuracy": 0.15335492491722108,
      "num_tokens": 15550670.0,
      "step": 7105
    },
    {
      "entropy": 5.608358001708984,
      "epoch": 0.7056371576022231,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004958482942460536,
      "loss": 5.3897,
      "mean_token_accuracy": 0.153936704993248,
      "num_tokens": 15562222.0,
      "step": 7110
    },
    {
      "entropy": 5.563803768157959,
      "epoch": 0.7061333862643906,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004958415164374316,
      "loss": 5.3811,
      "mean_token_accuracy": 0.15262936279177666,
      "num_tokens": 15573217.0,
      "step": 7115
    },
    {
      "entropy": 5.487719488143921,
      "epoch": 0.7066296149265582,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004958347331524158,
      "loss": 5.2369,
      "mean_token_accuracy": 0.1654719203710556,
      "num_tokens": 15584043.0,
      "step": 7120
    },
    {
      "entropy": 5.501652574539184,
      "epoch": 0.7071258435887257,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958279443911745,
      "loss": 5.35,
      "mean_token_accuracy": 0.15733844563364982,
      "num_tokens": 15595844.0,
      "step": 7125
    },
    {
      "entropy": 5.594591426849365,
      "epoch": 0.7076220722508932,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495821150153876,
      "loss": 5.4561,
      "mean_token_accuracy": 0.14959362596273423,
      "num_tokens": 15606748.0,
      "step": 7130
    },
    {
      "entropy": 5.556677150726318,
      "epoch": 0.7081183009130607,
      "grad_norm": 0.84375,
      "learning_rate": 0.000495814350440689,
      "loss": 5.3463,
      "mean_token_accuracy": 0.1580582931637764,
      "num_tokens": 15617415.0,
      "step": 7135
    },
    {
      "entropy": 5.61547327041626,
      "epoch": 0.7086145295752283,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004958075452517816,
      "loss": 5.4226,
      "mean_token_accuracy": 0.15163422524929046,
      "num_tokens": 15628953.0,
      "step": 7140
    },
    {
      "entropy": 5.525249433517456,
      "epoch": 0.7091107582373958,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958007345873232,
      "loss": 5.301,
      "mean_token_accuracy": 0.15187306553125382,
      "num_tokens": 15638937.0,
      "step": 7145
    },
    {
      "entropy": 5.552455282211303,
      "epoch": 0.7096069868995634,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957939184474821,
      "loss": 5.4758,
      "mean_token_accuracy": 0.15658174604177474,
      "num_tokens": 15650522.0,
      "step": 7150
    },
    {
      "entropy": 5.555604314804077,
      "epoch": 0.7101032155617308,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957870968324277,
      "loss": 5.4405,
      "mean_token_accuracy": 0.14736189246177672,
      "num_tokens": 15661899.0,
      "step": 7155
    },
    {
      "entropy": 5.607872104644775,
      "epoch": 0.7105994442238984,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957802697423291,
      "loss": 5.3121,
      "mean_token_accuracy": 0.15969704389572142,
      "num_tokens": 15673234.0,
      "step": 7160
    },
    {
      "entropy": 5.543533754348755,
      "epoch": 0.7110956728860659,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004957734371773554,
      "loss": 5.332,
      "mean_token_accuracy": 0.15109097957611084,
      "num_tokens": 15684567.0,
      "step": 7165
    },
    {
      "entropy": 5.526414585113526,
      "epoch": 0.7115919015482334,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957665991376763,
      "loss": 5.4649,
      "mean_token_accuracy": 0.15397563129663466,
      "num_tokens": 15695726.0,
      "step": 7170
    },
    {
      "entropy": 5.611112976074219,
      "epoch": 0.712088130210401,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004957597556234613,
      "loss": 5.4744,
      "mean_token_accuracy": 0.15540014952421188,
      "num_tokens": 15706598.0,
      "step": 7175
    },
    {
      "entropy": 5.604140853881836,
      "epoch": 0.7125843588725684,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957529066348799,
      "loss": 5.342,
      "mean_token_accuracy": 0.15936760753393173,
      "num_tokens": 15717915.0,
      "step": 7180
    },
    {
      "entropy": 5.604084444046021,
      "epoch": 0.713080587534736,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004957460521721022,
      "loss": 5.4075,
      "mean_token_accuracy": 0.15279852151870726,
      "num_tokens": 15730466.0,
      "step": 7185
    },
    {
      "entropy": 5.572271633148193,
      "epoch": 0.7135768161969035,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000495739192235298,
      "loss": 5.3883,
      "mean_token_accuracy": 0.15851826816797257,
      "num_tokens": 15741944.0,
      "step": 7190
    },
    {
      "entropy": 5.597372150421142,
      "epoch": 0.7140730448590711,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957323268246376,
      "loss": 5.4298,
      "mean_token_accuracy": 0.1541032537817955,
      "num_tokens": 15753839.0,
      "step": 7195
    },
    {
      "entropy": 5.614315319061279,
      "epoch": 0.7145692735212386,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957254559402911,
      "loss": 5.3704,
      "mean_token_accuracy": 0.15371397510170937,
      "num_tokens": 15765443.0,
      "step": 7200
    },
    {
      "entropy": 5.5829390525817875,
      "epoch": 0.7150655021834061,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957185795824288,
      "loss": 5.2999,
      "mean_token_accuracy": 0.15983096659183502,
      "num_tokens": 15775134.0,
      "step": 7205
    },
    {
      "entropy": 5.498440790176391,
      "epoch": 0.7155617308455736,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957116977512214,
      "loss": 5.2779,
      "mean_token_accuracy": 0.15592044442892075,
      "num_tokens": 15784900.0,
      "step": 7210
    },
    {
      "entropy": 5.563574981689453,
      "epoch": 0.7160579595077412,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957048104468395,
      "loss": 5.3958,
      "mean_token_accuracy": 0.15710543990135192,
      "num_tokens": 15796383.0,
      "step": 7215
    },
    {
      "entropy": 5.572084712982178,
      "epoch": 0.7165541881699087,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004956979176694539,
      "loss": 5.3935,
      "mean_token_accuracy": 0.15849161446094512,
      "num_tokens": 15806879.0,
      "step": 7220
    },
    {
      "entropy": 5.621825218200684,
      "epoch": 0.7170504168320763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956910194192355,
      "loss": 5.4874,
      "mean_token_accuracy": 0.1558576539158821,
      "num_tokens": 15818298.0,
      "step": 7225
    },
    {
      "entropy": 5.513638877868653,
      "epoch": 0.7175466454942437,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956841156963553,
      "loss": 5.3313,
      "mean_token_accuracy": 0.1556086167693138,
      "num_tokens": 15828288.0,
      "step": 7230
    },
    {
      "entropy": 5.539648342132568,
      "epoch": 0.7180428741564113,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956772065009846,
      "loss": 5.3289,
      "mean_token_accuracy": 0.15712707936763765,
      "num_tokens": 15838275.0,
      "step": 7235
    },
    {
      "entropy": 5.662791728973389,
      "epoch": 0.7185391028185788,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004956702918332948,
      "loss": 5.5076,
      "mean_token_accuracy": 0.14859764873981476,
      "num_tokens": 15850477.0,
      "step": 7240
    },
    {
      "entropy": 5.632599115371704,
      "epoch": 0.7190353314807463,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956633716934572,
      "loss": 5.4476,
      "mean_token_accuracy": 0.1546854555606842,
      "num_tokens": 15861620.0,
      "step": 7245
    },
    {
      "entropy": 5.576498508453369,
      "epoch": 0.7195315601429139,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004956564460816433,
      "loss": 5.3115,
      "mean_token_accuracy": 0.1618914484977722,
      "num_tokens": 15872069.0,
      "step": 7250
    },
    {
      "entropy": 5.584980821609497,
      "epoch": 0.7200277888050813,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956495149980252,
      "loss": 5.3452,
      "mean_token_accuracy": 0.15717238336801528,
      "num_tokens": 15883164.0,
      "step": 7255
    },
    {
      "entropy": 5.5276686668396,
      "epoch": 0.7205240174672489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956425784427745,
      "loss": 5.4802,
      "mean_token_accuracy": 0.1512013278901577,
      "num_tokens": 15893574.0,
      "step": 7260
    },
    {
      "entropy": 5.548484802246094,
      "epoch": 0.7210202461294164,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956356364160634,
      "loss": 5.3591,
      "mean_token_accuracy": 0.15895131528377532,
      "num_tokens": 15904999.0,
      "step": 7265
    },
    {
      "entropy": 5.699073362350464,
      "epoch": 0.721516474791584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956286889180639,
      "loss": 5.4136,
      "mean_token_accuracy": 0.15923498421907425,
      "num_tokens": 15915839.0,
      "step": 7270
    },
    {
      "entropy": 5.580031013488769,
      "epoch": 0.7220127034537515,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956217359489483,
      "loss": 5.3371,
      "mean_token_accuracy": 0.1577318489551544,
      "num_tokens": 15927335.0,
      "step": 7275
    },
    {
      "entropy": 5.516853332519531,
      "epoch": 0.722508932115919,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495614777508889,
      "loss": 5.3673,
      "mean_token_accuracy": 0.15981525480747222,
      "num_tokens": 15937640.0,
      "step": 7280
    },
    {
      "entropy": 5.5997645378112795,
      "epoch": 0.7230051607780865,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956078135980586,
      "loss": 5.3609,
      "mean_token_accuracy": 0.15928976237773895,
      "num_tokens": 15949044.0,
      "step": 7285
    },
    {
      "entropy": 5.618537569046021,
      "epoch": 0.7235013894402541,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004956008442166299,
      "loss": 5.4395,
      "mean_token_accuracy": 0.14896034598350524,
      "num_tokens": 15961415.0,
      "step": 7290
    },
    {
      "entropy": 5.491722917556762,
      "epoch": 0.7239976181024216,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004955938693647756,
      "loss": 5.3522,
      "mean_token_accuracy": 0.15638119876384735,
      "num_tokens": 15973689.0,
      "step": 7295
    },
    {
      "entropy": 5.544183731079102,
      "epoch": 0.7244938467645892,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955868890426686,
      "loss": 5.4049,
      "mean_token_accuracy": 0.1539571315050125,
      "num_tokens": 15986754.0,
      "step": 7300
    },
    {
      "entropy": 5.6224809169769285,
      "epoch": 0.7249900754267566,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955799032504821,
      "loss": 5.3359,
      "mean_token_accuracy": 0.1527300626039505,
      "num_tokens": 15997268.0,
      "step": 7305
    },
    {
      "entropy": 5.557440757751465,
      "epoch": 0.7254863040889242,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004955729119883893,
      "loss": 5.3285,
      "mean_token_accuracy": 0.15668768286705018,
      "num_tokens": 16008392.0,
      "step": 7310
    },
    {
      "entropy": 5.541127443313599,
      "epoch": 0.7259825327510917,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955659152565636,
      "loss": 5.4373,
      "mean_token_accuracy": 0.14914107024669648,
      "num_tokens": 16019674.0,
      "step": 7315
    },
    {
      "entropy": 5.57411823272705,
      "epoch": 0.7264787614132592,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955589130551784,
      "loss": 5.415,
      "mean_token_accuracy": 0.14987687915563583,
      "num_tokens": 16030095.0,
      "step": 7320
    },
    {
      "entropy": 5.550596857070923,
      "epoch": 0.7269749900754268,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955519053844076,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1609972432255745,
      "num_tokens": 16041039.0,
      "step": 7325
    },
    {
      "entropy": 5.555497407913208,
      "epoch": 0.7274712187375942,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004955448922444246,
      "loss": 5.354,
      "mean_token_accuracy": 0.1531379833817482,
      "num_tokens": 16051251.0,
      "step": 7330
    },
    {
      "entropy": 5.500340843200684,
      "epoch": 0.7279674473997618,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955378736354036,
      "loss": 5.2451,
      "mean_token_accuracy": 0.16135801672935485,
      "num_tokens": 16061578.0,
      "step": 7335
    },
    {
      "entropy": 5.5460480690002445,
      "epoch": 0.7284636760619293,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004955308495575185,
      "loss": 5.3333,
      "mean_token_accuracy": 0.15707673132419586,
      "num_tokens": 16073532.0,
      "step": 7340
    },
    {
      "entropy": 5.530425405502319,
      "epoch": 0.7289599047240969,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955238200109436,
      "loss": 5.3405,
      "mean_token_accuracy": 0.15347977727651596,
      "num_tokens": 16084689.0,
      "step": 7345
    },
    {
      "entropy": 5.555598449707031,
      "epoch": 0.7294561333862644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955167849958531,
      "loss": 5.3326,
      "mean_token_accuracy": 0.16167059391736985,
      "num_tokens": 16094598.0,
      "step": 7350
    },
    {
      "entropy": 5.525218915939331,
      "epoch": 0.729952362048432,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955097445124215,
      "loss": 5.2726,
      "mean_token_accuracy": 0.164432792365551,
      "num_tokens": 16105901.0,
      "step": 7355
    },
    {
      "entropy": 5.598638868331909,
      "epoch": 0.7304485907105994,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955026985608235,
      "loss": 5.4893,
      "mean_token_accuracy": 0.15626563280820846,
      "num_tokens": 16118060.0,
      "step": 7360
    },
    {
      "entropy": 5.5312505722045895,
      "epoch": 0.730944819372767,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004954956471412336,
      "loss": 5.2701,
      "mean_token_accuracy": 0.162357459962368,
      "num_tokens": 16129506.0,
      "step": 7365
    },
    {
      "entropy": 5.537882375717163,
      "epoch": 0.7314410480349345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954885902538269,
      "loss": 5.2994,
      "mean_token_accuracy": 0.16724557131528855,
      "num_tokens": 16140269.0,
      "step": 7370
    },
    {
      "entropy": 5.541881942749024,
      "epoch": 0.7319372766971021,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004954815278987781,
      "loss": 5.3838,
      "mean_token_accuracy": 0.15459475219249724,
      "num_tokens": 16151139.0,
      "step": 7375
    },
    {
      "entropy": 5.571033191680908,
      "epoch": 0.7324335053592695,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954744600762627,
      "loss": 5.301,
      "mean_token_accuracy": 0.16219995319843292,
      "num_tokens": 16162784.0,
      "step": 7380
    },
    {
      "entropy": 5.587934827804565,
      "epoch": 0.7329297340214371,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004954673867864555,
      "loss": 5.3706,
      "mean_token_accuracy": 0.15517609268426896,
      "num_tokens": 16174496.0,
      "step": 7385
    },
    {
      "entropy": 5.5697893619537355,
      "epoch": 0.7334259626836046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954603080295323,
      "loss": 5.417,
      "mean_token_accuracy": 0.1523167073726654,
      "num_tokens": 16186962.0,
      "step": 7390
    },
    {
      "entropy": 5.616978693008423,
      "epoch": 0.7339221913457721,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004954532238056686,
      "loss": 5.4634,
      "mean_token_accuracy": 0.1586115464568138,
      "num_tokens": 16196836.0,
      "step": 7395
    },
    {
      "entropy": 5.562237691879273,
      "epoch": 0.7344184200079397,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954461341150399,
      "loss": 5.412,
      "mean_token_accuracy": 0.14974036440253258,
      "num_tokens": 16207326.0,
      "step": 7400
    },
    {
      "entropy": 5.553341245651245,
      "epoch": 0.7349146486701071,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004954390389578222,
      "loss": 5.3512,
      "mean_token_accuracy": 0.15514750480651857,
      "num_tokens": 16218241.0,
      "step": 7405
    },
    {
      "entropy": 5.557055234909058,
      "epoch": 0.7354108773322747,
      "grad_norm": 1.125,
      "learning_rate": 0.0004954319383341912,
      "loss": 5.3811,
      "mean_token_accuracy": 0.1519709572196007,
      "num_tokens": 16228101.0,
      "step": 7410
    },
    {
      "entropy": 5.483792734146118,
      "epoch": 0.7359071059944422,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004954248322443232,
      "loss": 5.2217,
      "mean_token_accuracy": 0.1639699563384056,
      "num_tokens": 16239080.0,
      "step": 7415
    },
    {
      "entropy": 5.513272619247436,
      "epoch": 0.7364033346566098,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004954177206883943,
      "loss": 5.2789,
      "mean_token_accuracy": 0.16293613165616988,
      "num_tokens": 16250446.0,
      "step": 7420
    },
    {
      "entropy": 5.548108673095703,
      "epoch": 0.7368995633187773,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954106036665809,
      "loss": 5.3472,
      "mean_token_accuracy": 0.16110820472240447,
      "num_tokens": 16260414.0,
      "step": 7425
    },
    {
      "entropy": 5.53383059501648,
      "epoch": 0.7373957919809448,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954034811790594,
      "loss": 5.348,
      "mean_token_accuracy": 0.15963209569454193,
      "num_tokens": 16270925.0,
      "step": 7430
    },
    {
      "entropy": 5.590784549713135,
      "epoch": 0.7378920206431123,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953963532260067,
      "loss": 5.4139,
      "mean_token_accuracy": 0.15323833525180816,
      "num_tokens": 16282761.0,
      "step": 7435
    },
    {
      "entropy": 5.562132930755615,
      "epoch": 0.7383882493052799,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953892198075991,
      "loss": 5.2931,
      "mean_token_accuracy": 0.1577959895133972,
      "num_tokens": 16294901.0,
      "step": 7440
    },
    {
      "entropy": 5.480667400360107,
      "epoch": 0.7388844779674474,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953820809240138,
      "loss": 5.2213,
      "mean_token_accuracy": 0.16530861109495162,
      "num_tokens": 16306310.0,
      "step": 7445
    },
    {
      "entropy": 5.540656471252442,
      "epoch": 0.739380706629615,
      "grad_norm": 0.875,
      "learning_rate": 0.0004953749365754279,
      "loss": 5.4095,
      "mean_token_accuracy": 0.15136562734842302,
      "num_tokens": 16317325.0,
      "step": 7450
    },
    {
      "entropy": 5.56732702255249,
      "epoch": 0.7398769352917824,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953677867620184,
      "loss": 5.3632,
      "mean_token_accuracy": 0.15809557288885118,
      "num_tokens": 16328115.0,
      "step": 7455
    },
    {
      "entropy": 5.5416851997375485,
      "epoch": 0.74037316395395,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004953606314839626,
      "loss": 5.3817,
      "mean_token_accuracy": 0.15903314501047133,
      "num_tokens": 16340692.0,
      "step": 7460
    },
    {
      "entropy": 5.592165327072143,
      "epoch": 0.7408693926161175,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495353470741438,
      "loss": 5.3615,
      "mean_token_accuracy": 0.15553851202130317,
      "num_tokens": 16351704.0,
      "step": 7465
    },
    {
      "entropy": 5.545892381668091,
      "epoch": 0.741365621278285,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000495346304534622,
      "loss": 5.3758,
      "mean_token_accuracy": 0.1536679521203041,
      "num_tokens": 16363443.0,
      "step": 7470
    },
    {
      "entropy": 5.455718374252319,
      "epoch": 0.7418618499404526,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953391328636926,
      "loss": 5.2782,
      "mean_token_accuracy": 0.16107317805290222,
      "num_tokens": 16373913.0,
      "step": 7475
    },
    {
      "entropy": 5.525425672531128,
      "epoch": 0.74235807860262,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004953319557288275,
      "loss": 5.3635,
      "mean_token_accuracy": 0.1576432153582573,
      "num_tokens": 16384607.0,
      "step": 7480
    },
    {
      "entropy": 5.586904621124267,
      "epoch": 0.7428543072647876,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953247731302046,
      "loss": 5.3785,
      "mean_token_accuracy": 0.1606116384267807,
      "num_tokens": 16394769.0,
      "step": 7485
    },
    {
      "entropy": 5.4808979511260985,
      "epoch": 0.7433505359269551,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953175850680022,
      "loss": 5.2879,
      "mean_token_accuracy": 0.15461359769105912,
      "num_tokens": 16404906.0,
      "step": 7490
    },
    {
      "entropy": 5.520489168167114,
      "epoch": 0.7438467645891227,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953103915423983,
      "loss": 5.3829,
      "mean_token_accuracy": 0.15156420618295668,
      "num_tokens": 16415779.0,
      "step": 7495
    },
    {
      "entropy": 5.613870286941529,
      "epoch": 0.7443429932512902,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953031925535715,
      "loss": 5.4432,
      "mean_token_accuracy": 0.15039485096931457,
      "num_tokens": 16426612.0,
      "step": 7500
    },
    {
      "entropy": 5.600257110595703,
      "epoch": 0.7448392219134577,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004952959881017003,
      "loss": 5.4132,
      "mean_token_accuracy": 0.14676258116960525,
      "num_tokens": 16437168.0,
      "step": 7505
    },
    {
      "entropy": 5.565494775772095,
      "epoch": 0.7453354505756252,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952887781869631,
      "loss": 5.3064,
      "mean_token_accuracy": 0.15396720170974731,
      "num_tokens": 16448073.0,
      "step": 7510
    },
    {
      "entropy": 5.481198358535766,
      "epoch": 0.7458316792377928,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004952815628095391,
      "loss": 5.2279,
      "mean_token_accuracy": 0.1643837571144104,
      "num_tokens": 16458271.0,
      "step": 7515
    },
    {
      "entropy": 5.538802242279052,
      "epoch": 0.7463279078999603,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004952743419696069,
      "loss": 5.486,
      "mean_token_accuracy": 0.1504957288503647,
      "num_tokens": 16469667.0,
      "step": 7520
    },
    {
      "entropy": 5.588812017440796,
      "epoch": 0.7468241365621279,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004952671156673456,
      "loss": 5.3763,
      "mean_token_accuracy": 0.15851462110877038,
      "num_tokens": 16479449.0,
      "step": 7525
    },
    {
      "entropy": 5.6118762493133545,
      "epoch": 0.7473203652242953,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004952598839029345,
      "loss": 5.4574,
      "mean_token_accuracy": 0.15003828853368759,
      "num_tokens": 16490754.0,
      "step": 7530
    },
    {
      "entropy": 5.569642114639282,
      "epoch": 0.7478165938864629,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004952526466765529,
      "loss": 5.2885,
      "mean_token_accuracy": 0.15977183133363723,
      "num_tokens": 16500017.0,
      "step": 7535
    },
    {
      "entropy": 5.619054079055786,
      "epoch": 0.7483128225486304,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004952454039883804,
      "loss": 5.4295,
      "mean_token_accuracy": 0.15389198064804077,
      "num_tokens": 16512155.0,
      "step": 7540
    },
    {
      "entropy": 5.5996985912323,
      "epoch": 0.7488090512107979,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004952381558385963,
      "loss": 5.4056,
      "mean_token_accuracy": 0.15694097131490709,
      "num_tokens": 16522735.0,
      "step": 7545
    },
    {
      "entropy": 5.546477651596069,
      "epoch": 0.7493052798729655,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952309022273805,
      "loss": 5.2975,
      "mean_token_accuracy": 0.16060952693223954,
      "num_tokens": 16533031.0,
      "step": 7550
    },
    {
      "entropy": 5.537762117385864,
      "epoch": 0.7498015085351329,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004952236431549129,
      "loss": 5.376,
      "mean_token_accuracy": 0.1574877992272377,
      "num_tokens": 16544106.0,
      "step": 7555
    },
    {
      "entropy": 5.542153215408325,
      "epoch": 0.7502977371973005,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952163786213735,
      "loss": 5.3246,
      "mean_token_accuracy": 0.15493188053369522,
      "num_tokens": 16554585.0,
      "step": 7560
    },
    {
      "entropy": 5.545419073104858,
      "epoch": 0.750793965859468,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004952091086269425,
      "loss": 5.3079,
      "mean_token_accuracy": 0.160880246758461,
      "num_tokens": 16565660.0,
      "step": 7565
    },
    {
      "entropy": 5.600740814208985,
      "epoch": 0.7512901945216356,
      "grad_norm": 0.875,
      "learning_rate": 0.0004952018331718001,
      "loss": 5.4801,
      "mean_token_accuracy": 0.14994871169328688,
      "num_tokens": 16576381.0,
      "step": 7570
    },
    {
      "entropy": 5.595485544204712,
      "epoch": 0.7517864231838031,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951945522561268,
      "loss": 5.4211,
      "mean_token_accuracy": 0.15069512873888016,
      "num_tokens": 16587100.0,
      "step": 7575
    },
    {
      "entropy": 5.627595043182373,
      "epoch": 0.7522826518459707,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951872658801029,
      "loss": 5.3781,
      "mean_token_accuracy": 0.15624116063117982,
      "num_tokens": 16598195.0,
      "step": 7580
    },
    {
      "entropy": 5.583223342895508,
      "epoch": 0.7527788805081381,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004951799740439093,
      "loss": 5.333,
      "mean_token_accuracy": 0.16368968039751053,
      "num_tokens": 16609225.0,
      "step": 7585
    },
    {
      "entropy": 5.542636394500732,
      "epoch": 0.7532751091703057,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000495172676747727,
      "loss": 5.3676,
      "mean_token_accuracy": 0.16041516214609147,
      "num_tokens": 16620874.0,
      "step": 7590
    },
    {
      "entropy": 5.560567045211792,
      "epoch": 0.7537713378324732,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951653739917366,
      "loss": 5.3799,
      "mean_token_accuracy": 0.15980777144432068,
      "num_tokens": 16631758.0,
      "step": 7595
    },
    {
      "entropy": 5.589287662506104,
      "epoch": 0.7542675664946408,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951580657761192,
      "loss": 5.3928,
      "mean_token_accuracy": 0.15425439029932023,
      "num_tokens": 16642269.0,
      "step": 7600
    },
    {
      "entropy": 5.548319911956787,
      "epoch": 0.7547637951568082,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004951507521010563,
      "loss": 5.2546,
      "mean_token_accuracy": 0.1583784967660904,
      "num_tokens": 16653842.0,
      "step": 7605
    },
    {
      "entropy": 5.518681335449219,
      "epoch": 0.7552600238189758,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004951434329667293,
      "loss": 5.3646,
      "mean_token_accuracy": 0.15525074154138566,
      "num_tokens": 16664625.0,
      "step": 7610
    },
    {
      "entropy": 5.543423080444336,
      "epoch": 0.7557562524811433,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004951361083733192,
      "loss": 5.3734,
      "mean_token_accuracy": 0.16123099625110626,
      "num_tokens": 16677076.0,
      "step": 7615
    },
    {
      "entropy": 5.590677690505982,
      "epoch": 0.7562524811433108,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951287783210082,
      "loss": 5.3925,
      "mean_token_accuracy": 0.15301271378993989,
      "num_tokens": 16688310.0,
      "step": 7620
    },
    {
      "entropy": 5.622752904891968,
      "epoch": 0.7567487098054784,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004951214428099778,
      "loss": 5.486,
      "mean_token_accuracy": 0.15087427645921708,
      "num_tokens": 16699992.0,
      "step": 7625
    },
    {
      "entropy": 5.701030111312866,
      "epoch": 0.7572449384676458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951141018404099,
      "loss": 5.4006,
      "mean_token_accuracy": 0.1635703831911087,
      "num_tokens": 16710815.0,
      "step": 7630
    },
    {
      "entropy": 5.488446474075317,
      "epoch": 0.7577411671298134,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004951067554124865,
      "loss": 5.3261,
      "mean_token_accuracy": 0.16771110445261,
      "num_tokens": 16721656.0,
      "step": 7635
    },
    {
      "entropy": 5.452736043930054,
      "epoch": 0.7582373957919809,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049509940352639,
      "loss": 5.2577,
      "mean_token_accuracy": 0.1594707563519478,
      "num_tokens": 16730977.0,
      "step": 7640
    },
    {
      "entropy": 5.517125177383423,
      "epoch": 0.7587336244541485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004950920461823024,
      "loss": 5.3532,
      "mean_token_accuracy": 0.16074190959334372,
      "num_tokens": 16741569.0,
      "step": 7645
    },
    {
      "entropy": 5.561464452743531,
      "epoch": 0.759229853116316,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950846833804064,
      "loss": 5.2488,
      "mean_token_accuracy": 0.1626979038119316,
      "num_tokens": 16752300.0,
      "step": 7650
    },
    {
      "entropy": 5.552998352050781,
      "epoch": 0.7597260817784836,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004950773151208845,
      "loss": 5.3667,
      "mean_token_accuracy": 0.16301628947257996,
      "num_tokens": 16763202.0,
      "step": 7655
    },
    {
      "entropy": 5.5504326820373535,
      "epoch": 0.760222310440651,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004950699414039194,
      "loss": 5.3,
      "mean_token_accuracy": 0.15984521061182022,
      "num_tokens": 16773792.0,
      "step": 7660
    },
    {
      "entropy": 5.58971004486084,
      "epoch": 0.7607185391028186,
      "grad_norm": 0.875,
      "learning_rate": 0.000495062562229694,
      "loss": 5.3861,
      "mean_token_accuracy": 0.15393077507615088,
      "num_tokens": 16785669.0,
      "step": 7665
    },
    {
      "entropy": 5.538853883743286,
      "epoch": 0.7612147677649861,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004950551775983912,
      "loss": 5.308,
      "mean_token_accuracy": 0.15864354372024536,
      "num_tokens": 16797141.0,
      "step": 7670
    },
    {
      "entropy": 5.5318855285644535,
      "epoch": 0.7617109964271537,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950477875101942,
      "loss": 5.3443,
      "mean_token_accuracy": 0.15933319330215454,
      "num_tokens": 16807904.0,
      "step": 7675
    },
    {
      "entropy": 5.522061443328857,
      "epoch": 0.7622072250893211,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004950403919652863,
      "loss": 5.3706,
      "mean_token_accuracy": 0.15435705333948135,
      "num_tokens": 16817270.0,
      "step": 7680
    },
    {
      "entropy": 5.585412454605103,
      "epoch": 0.7627034537514887,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950329909638507,
      "loss": 5.3782,
      "mean_token_accuracy": 0.15556803941726685,
      "num_tokens": 16828770.0,
      "step": 7685
    },
    {
      "entropy": 5.6229020118713375,
      "epoch": 0.7631996824136562,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000495025584506071,
      "loss": 5.3878,
      "mean_token_accuracy": 0.15136945396661758,
      "num_tokens": 16839811.0,
      "step": 7690
    },
    {
      "entropy": 5.597045755386352,
      "epoch": 0.7636959110758237,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950181725921311,
      "loss": 5.3649,
      "mean_token_accuracy": 0.1557265967130661,
      "num_tokens": 16850180.0,
      "step": 7695
    },
    {
      "entropy": 5.555637502670288,
      "epoch": 0.7641921397379913,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950107552222145,
      "loss": 5.3022,
      "mean_token_accuracy": 0.15821422785520553,
      "num_tokens": 16862401.0,
      "step": 7700
    },
    {
      "entropy": 5.556877374649048,
      "epoch": 0.7646883684001587,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004950033323965052,
      "loss": 5.387,
      "mean_token_accuracy": 0.14840614870190622,
      "num_tokens": 16872733.0,
      "step": 7705
    },
    {
      "entropy": 5.597369623184204,
      "epoch": 0.7651845970623263,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949959041151875,
      "loss": 5.4385,
      "mean_token_accuracy": 0.15158104300498962,
      "num_tokens": 16883291.0,
      "step": 7710
    },
    {
      "entropy": 5.633507823944091,
      "epoch": 0.7656808257244938,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004949884703784453,
      "loss": 5.3883,
      "mean_token_accuracy": 0.1597304791212082,
      "num_tokens": 16894591.0,
      "step": 7715
    },
    {
      "entropy": 5.5499327182769775,
      "epoch": 0.7661770543866614,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000494981031186463,
      "loss": 5.3571,
      "mean_token_accuracy": 0.15928771048784257,
      "num_tokens": 16905239.0,
      "step": 7720
    },
    {
      "entropy": 5.515296554565429,
      "epoch": 0.7666732830488289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949735865394253,
      "loss": 5.2879,
      "mean_token_accuracy": 0.1597840592265129,
      "num_tokens": 16914711.0,
      "step": 7725
    },
    {
      "entropy": 5.643017721176148,
      "epoch": 0.7671695117109965,
      "grad_norm": 0.875,
      "learning_rate": 0.0004949661364375167,
      "loss": 5.3983,
      "mean_token_accuracy": 0.15871076434850692,
      "num_tokens": 16926184.0,
      "step": 7730
    },
    {
      "entropy": 5.5712502002716064,
      "epoch": 0.7676657403731639,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949586808809217,
      "loss": 5.3929,
      "mean_token_accuracy": 0.14740023016929626,
      "num_tokens": 16937138.0,
      "step": 7735
    },
    {
      "entropy": 5.585552787780761,
      "epoch": 0.7681619690353315,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004949512198698256,
      "loss": 5.3632,
      "mean_token_accuracy": 0.15606023520231246,
      "num_tokens": 16948797.0,
      "step": 7740
    },
    {
      "entropy": 5.669891786575318,
      "epoch": 0.768658197697499,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000494943753404413,
      "loss": 5.44,
      "mean_token_accuracy": 0.14690390303730966,
      "num_tokens": 16959407.0,
      "step": 7745
    },
    {
      "entropy": 5.500219058990479,
      "epoch": 0.7691544263596666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949362814848693,
      "loss": 5.3809,
      "mean_token_accuracy": 0.16069244593381882,
      "num_tokens": 16970600.0,
      "step": 7750
    },
    {
      "entropy": 5.569076681137085,
      "epoch": 0.769650655021834,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004949288041113799,
      "loss": 5.3758,
      "mean_token_accuracy": 0.15595488250255585,
      "num_tokens": 16980167.0,
      "step": 7755
    },
    {
      "entropy": 5.576169157028199,
      "epoch": 0.7701468836840016,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949213212841299,
      "loss": 5.294,
      "mean_token_accuracy": 0.16847391426563263,
      "num_tokens": 16992209.0,
      "step": 7760
    },
    {
      "entropy": 5.525841569900512,
      "epoch": 0.7706431123461691,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494913833003305,
      "loss": 5.358,
      "mean_token_accuracy": 0.15427246317267418,
      "num_tokens": 17003508.0,
      "step": 7765
    },
    {
      "entropy": 5.554324913024902,
      "epoch": 0.7711393410083366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494906339269091,
      "loss": 5.3571,
      "mean_token_accuracy": 0.16296836584806443,
      "num_tokens": 17013160.0,
      "step": 7770
    },
    {
      "entropy": 5.6100077629089355,
      "epoch": 0.7716355696705042,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948988400816736,
      "loss": 5.3026,
      "mean_token_accuracy": 0.16540822982788086,
      "num_tokens": 17023601.0,
      "step": 7775
    },
    {
      "entropy": 5.585769557952881,
      "epoch": 0.7721317983326716,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004948913354412388,
      "loss": 5.3466,
      "mean_token_accuracy": 0.1556863531470299,
      "num_tokens": 17035530.0,
      "step": 7780
    },
    {
      "entropy": 5.534577703475952,
      "epoch": 0.7726280269948392,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004948838253479727,
      "loss": 5.2912,
      "mean_token_accuracy": 0.16162253469228743,
      "num_tokens": 17046059.0,
      "step": 7785
    },
    {
      "entropy": 5.534445667266846,
      "epoch": 0.7731242556570067,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948763098020616,
      "loss": 5.3357,
      "mean_token_accuracy": 0.1601007729768753,
      "num_tokens": 17056231.0,
      "step": 7790
    },
    {
      "entropy": 5.543054056167603,
      "epoch": 0.7736204843191743,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004948687888036918,
      "loss": 5.3066,
      "mean_token_accuracy": 0.16334328353404998,
      "num_tokens": 17067144.0,
      "step": 7795
    },
    {
      "entropy": 5.547908449172974,
      "epoch": 0.7741167129813418,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948612623530497,
      "loss": 5.3662,
      "mean_token_accuracy": 0.15529667735099792,
      "num_tokens": 17077276.0,
      "step": 7800
    },
    {
      "entropy": 5.577593660354614,
      "epoch": 0.7746129416435094,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948537304503221,
      "loss": 5.4287,
      "mean_token_accuracy": 0.15559004247188568,
      "num_tokens": 17089022.0,
      "step": 7805
    },
    {
      "entropy": 5.59152398109436,
      "epoch": 0.7751091703056768,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004948461930956959,
      "loss": 5.3788,
      "mean_token_accuracy": 0.16160464957356452,
      "num_tokens": 17100657.0,
      "step": 7810
    },
    {
      "entropy": 5.585228300094604,
      "epoch": 0.7756053989678444,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004948386502893577,
      "loss": 5.3659,
      "mean_token_accuracy": 0.15911518186330795,
      "num_tokens": 17111203.0,
      "step": 7815
    },
    {
      "entropy": 5.630031967163086,
      "epoch": 0.7761016276300119,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948311020314946,
      "loss": 5.4192,
      "mean_token_accuracy": 0.15520685762166977,
      "num_tokens": 17121999.0,
      "step": 7820
    },
    {
      "entropy": 5.601098108291626,
      "epoch": 0.7765978562921795,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004948235483222941,
      "loss": 5.357,
      "mean_token_accuracy": 0.15619806572794914,
      "num_tokens": 17131872.0,
      "step": 7825
    },
    {
      "entropy": 5.527451896667481,
      "epoch": 0.777094084954347,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948159891619431,
      "loss": 5.3762,
      "mean_token_accuracy": 0.15864432007074356,
      "num_tokens": 17143108.0,
      "step": 7830
    },
    {
      "entropy": 5.543817949295044,
      "epoch": 0.7775903136165145,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004948084245506293,
      "loss": 5.3113,
      "mean_token_accuracy": 0.1571893125772476,
      "num_tokens": 17154034.0,
      "step": 7835
    },
    {
      "entropy": 5.591415119171143,
      "epoch": 0.778086542278682,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049480085448854,
      "loss": 5.4094,
      "mean_token_accuracy": 0.15748743265867232,
      "num_tokens": 17164526.0,
      "step": 7840
    },
    {
      "entropy": 5.537517261505127,
      "epoch": 0.7785827709408495,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947932789758635,
      "loss": 5.3096,
      "mean_token_accuracy": 0.16233795881271362,
      "num_tokens": 17175653.0,
      "step": 7845
    },
    {
      "entropy": 5.498408555984497,
      "epoch": 0.7790789996030171,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494785698012787,
      "loss": 5.2635,
      "mean_token_accuracy": 0.1598915249109268,
      "num_tokens": 17186781.0,
      "step": 7850
    },
    {
      "entropy": 5.510153532028198,
      "epoch": 0.7795752282651845,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004947781115994989,
      "loss": 5.3136,
      "mean_token_accuracy": 0.15809347480535507,
      "num_tokens": 17197375.0,
      "step": 7855
    },
    {
      "entropy": 5.574258756637573,
      "epoch": 0.7800714569273521,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947705197361872,
      "loss": 5.3639,
      "mean_token_accuracy": 0.15284403711557387,
      "num_tokens": 17208153.0,
      "step": 7860
    },
    {
      "entropy": 5.5301109790802006,
      "epoch": 0.7805676855895196,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947629224230401,
      "loss": 5.3434,
      "mean_token_accuracy": 0.16163252890110016,
      "num_tokens": 17218914.0,
      "step": 7865
    },
    {
      "entropy": 5.598738956451416,
      "epoch": 0.7810639142516872,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947553196602461,
      "loss": 5.3323,
      "mean_token_accuracy": 0.15915288925170898,
      "num_tokens": 17229241.0,
      "step": 7870
    },
    {
      "entropy": 5.565932559967041,
      "epoch": 0.7815601429138547,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004947477114479939,
      "loss": 5.4012,
      "mean_token_accuracy": 0.15394682288169861,
      "num_tokens": 17239869.0,
      "step": 7875
    },
    {
      "entropy": 5.579468011856079,
      "epoch": 0.7820563715760223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947400977864718,
      "loss": 5.33,
      "mean_token_accuracy": 0.16071168035268785,
      "num_tokens": 17251761.0,
      "step": 7880
    },
    {
      "entropy": 5.619021606445313,
      "epoch": 0.7825526002381897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947324786758688,
      "loss": 5.4659,
      "mean_token_accuracy": 0.15247756689786912,
      "num_tokens": 17262812.0,
      "step": 7885
    },
    {
      "entropy": 5.608502960205078,
      "epoch": 0.7830488289003573,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947248541163738,
      "loss": 5.3148,
      "mean_token_accuracy": 0.1518617108464241,
      "num_tokens": 17273349.0,
      "step": 7890
    },
    {
      "entropy": 5.528835391998291,
      "epoch": 0.7835450575625248,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004947172241081759,
      "loss": 5.3794,
      "mean_token_accuracy": 0.1544090747833252,
      "num_tokens": 17284187.0,
      "step": 7895
    },
    {
      "entropy": 5.5203284740448,
      "epoch": 0.7840412862246924,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004947095886514643,
      "loss": 5.3168,
      "mean_token_accuracy": 0.15941943675279618,
      "num_tokens": 17295087.0,
      "step": 7900
    },
    {
      "entropy": 5.645417547225952,
      "epoch": 0.7845375148868599,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004947019477464283,
      "loss": 5.3478,
      "mean_token_accuracy": 0.15335867255926133,
      "num_tokens": 17305375.0,
      "step": 7905
    },
    {
      "entropy": 5.599386548995971,
      "epoch": 0.7850337435490274,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946943013932574,
      "loss": 5.4148,
      "mean_token_accuracy": 0.15695241540670396,
      "num_tokens": 17315618.0,
      "step": 7910
    },
    {
      "entropy": 5.504259967803955,
      "epoch": 0.7855299722111949,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004946866495921413,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1577161356806755,
      "num_tokens": 17327063.0,
      "step": 7915
    },
    {
      "entropy": 5.527556991577148,
      "epoch": 0.7860262008733624,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004946789923432695,
      "loss": 5.2841,
      "mean_token_accuracy": 0.15886348038911818,
      "num_tokens": 17338229.0,
      "step": 7920
    },
    {
      "entropy": 5.529096937179565,
      "epoch": 0.78652242953553,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004946713296468322,
      "loss": 5.2904,
      "mean_token_accuracy": 0.15761329531669616,
      "num_tokens": 17348917.0,
      "step": 7925
    },
    {
      "entropy": 5.524925708770752,
      "epoch": 0.7870186581976975,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946636615030192,
      "loss": 5.2972,
      "mean_token_accuracy": 0.15885206013917924,
      "num_tokens": 17359534.0,
      "step": 7930
    },
    {
      "entropy": 5.555211925506592,
      "epoch": 0.787514886859865,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004946559879120208,
      "loss": 5.364,
      "mean_token_accuracy": 0.15812431126832963,
      "num_tokens": 17369758.0,
      "step": 7935
    },
    {
      "entropy": 5.622425889968872,
      "epoch": 0.7880111155220325,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004946483088740273,
      "loss": 5.3779,
      "mean_token_accuracy": 0.1623288005590439,
      "num_tokens": 17380541.0,
      "step": 7940
    },
    {
      "entropy": 5.6299113750457765,
      "epoch": 0.7885073441842001,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004946406243892289,
      "loss": 5.482,
      "mean_token_accuracy": 0.1532146453857422,
      "num_tokens": 17391586.0,
      "step": 7945
    },
    {
      "entropy": 5.534591817855835,
      "epoch": 0.7890035728463676,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004946329344578163,
      "loss": 5.3528,
      "mean_token_accuracy": 0.15972859859466554,
      "num_tokens": 17401324.0,
      "step": 7950
    },
    {
      "entropy": 5.558495378494262,
      "epoch": 0.7894998015085352,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004946252390799803,
      "loss": 5.3668,
      "mean_token_accuracy": 0.15208089351654053,
      "num_tokens": 17413785.0,
      "step": 7955
    },
    {
      "entropy": 5.563504457473755,
      "epoch": 0.7899960301707026,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946175382559116,
      "loss": 5.3978,
      "mean_token_accuracy": 0.15269475132226945,
      "num_tokens": 17424571.0,
      "step": 7960
    },
    {
      "entropy": 5.5074962139129635,
      "epoch": 0.7904922588328702,
      "grad_norm": 0.90625,
      "learning_rate": 0.000494609831985801,
      "loss": 5.2952,
      "mean_token_accuracy": 0.15911698639392852,
      "num_tokens": 17435636.0,
      "step": 7965
    },
    {
      "entropy": 5.5618809223175045,
      "epoch": 0.7909884874950377,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049460212026984,
      "loss": 5.3888,
      "mean_token_accuracy": 0.15821974873542785,
      "num_tokens": 17446827.0,
      "step": 7970
    },
    {
      "entropy": 5.615344858169555,
      "epoch": 0.7914847161572053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004945944031082195,
      "loss": 5.4653,
      "mean_token_accuracy": 0.14881453514099122,
      "num_tokens": 17458381.0,
      "step": 7975
    },
    {
      "entropy": 5.627829933166504,
      "epoch": 0.7919809448193728,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000494586680501131,
      "loss": 5.4052,
      "mean_token_accuracy": 0.15582360774278642,
      "num_tokens": 17469686.0,
      "step": 7980
    },
    {
      "entropy": 5.623086833953858,
      "epoch": 0.7924771734815403,
      "grad_norm": 0.84375,
      "learning_rate": 0.000494578952448766,
      "loss": 5.3549,
      "mean_token_accuracy": 0.15843169391155243,
      "num_tokens": 17481546.0,
      "step": 7985
    },
    {
      "entropy": 5.510445547103882,
      "epoch": 0.7929734021437078,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945712189513161,
      "loss": 5.3126,
      "mean_token_accuracy": 0.1647471785545349,
      "num_tokens": 17492245.0,
      "step": 7990
    },
    {
      "entropy": 5.498069524765015,
      "epoch": 0.7934696308058753,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945634800089733,
      "loss": 5.3418,
      "mean_token_accuracy": 0.1637618474662304,
      "num_tokens": 17502159.0,
      "step": 7995
    },
    {
      "entropy": 5.5637451171875,
      "epoch": 0.7939658594680429,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004945557356219291,
      "loss": 5.4008,
      "mean_token_accuracy": 0.15005240887403487,
      "num_tokens": 17514576.0,
      "step": 8000
    },
    {
      "entropy": 5.600296926498413,
      "epoch": 0.7944620881302104,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004945479857903758,
      "loss": 5.3014,
      "mean_token_accuracy": 0.16126843243837358,
      "num_tokens": 17525663.0,
      "step": 8005
    },
    {
      "entropy": 5.535210084915161,
      "epoch": 0.7949583167923779,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945402305145056,
      "loss": 5.4113,
      "mean_token_accuracy": 0.15712710320949555,
      "num_tokens": 17537955.0,
      "step": 8010
    },
    {
      "entropy": 5.5215741157531735,
      "epoch": 0.7954545454545454,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004945324697945107,
      "loss": 5.3657,
      "mean_token_accuracy": 0.15642546415328978,
      "num_tokens": 17548904.0,
      "step": 8015
    },
    {
      "entropy": 5.605803775787353,
      "epoch": 0.795950774116713,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945247036305836,
      "loss": 5.3668,
      "mean_token_accuracy": 0.15853585451841354,
      "num_tokens": 17560274.0,
      "step": 8020
    },
    {
      "entropy": 5.529565477371216,
      "epoch": 0.7964470027788805,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945169320229169,
      "loss": 5.4314,
      "mean_token_accuracy": 0.15147985965013505,
      "num_tokens": 17571308.0,
      "step": 8025
    },
    {
      "entropy": 5.548037099838257,
      "epoch": 0.7969432314410481,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004945091549717033,
      "loss": 5.3203,
      "mean_token_accuracy": 0.15559469759464264,
      "num_tokens": 17582446.0,
      "step": 8030
    },
    {
      "entropy": 5.612826728820801,
      "epoch": 0.7974394601032155,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945013724771356,
      "loss": 5.3532,
      "mean_token_accuracy": 0.1566250890493393,
      "num_tokens": 17594397.0,
      "step": 8035
    },
    {
      "entropy": 5.592671251296997,
      "epoch": 0.7979356887653831,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004944935845394069,
      "loss": 5.4091,
      "mean_token_accuracy": 0.16092928647994995,
      "num_tokens": 17605131.0,
      "step": 8040
    },
    {
      "entropy": 5.577481508255005,
      "epoch": 0.7984319174275506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944857911587103,
      "loss": 5.2879,
      "mean_token_accuracy": 0.15620463490486144,
      "num_tokens": 17615958.0,
      "step": 8045
    },
    {
      "entropy": 5.51493763923645,
      "epoch": 0.7989281460897182,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004944779923352389,
      "loss": 5.2046,
      "mean_token_accuracy": 0.16234661340713502,
      "num_tokens": 17626571.0,
      "step": 8050
    },
    {
      "entropy": 5.596336603164673,
      "epoch": 0.7994243747518857,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004944701880691864,
      "loss": 5.3513,
      "mean_token_accuracy": 0.1588498055934906,
      "num_tokens": 17638158.0,
      "step": 8055
    },
    {
      "entropy": 5.568134927749634,
      "epoch": 0.7999206034140532,
      "grad_norm": 0.96875,
      "learning_rate": 0.000494462378360746,
      "loss": 5.4172,
      "mean_token_accuracy": 0.1578426703810692,
      "num_tokens": 17649384.0,
      "step": 8060
    },
    {
      "entropy": 5.552943229675293,
      "epoch": 0.8004168320762207,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944545632101117,
      "loss": 5.3316,
      "mean_token_accuracy": 0.15939236730337142,
      "num_tokens": 17661311.0,
      "step": 8065
    },
    {
      "entropy": 5.604061269760132,
      "epoch": 0.8009130607383882,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004944467426174769,
      "loss": 5.3709,
      "mean_token_accuracy": 0.1562285378575325,
      "num_tokens": 17672710.0,
      "step": 8070
    },
    {
      "entropy": 5.629287910461426,
      "epoch": 0.8014092894005558,
      "grad_norm": 0.84375,
      "learning_rate": 0.000494438916583036,
      "loss": 5.4789,
      "mean_token_accuracy": 0.15015798434615135,
      "num_tokens": 17683521.0,
      "step": 8075
    },
    {
      "entropy": 5.554546928405761,
      "epoch": 0.8019055180627233,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944310851069826,
      "loss": 5.3068,
      "mean_token_accuracy": 0.15868283808231354,
      "num_tokens": 17693183.0,
      "step": 8080
    },
    {
      "entropy": 5.5492284297943115,
      "epoch": 0.8024017467248908,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004944232481895112,
      "loss": 5.295,
      "mean_token_accuracy": 0.16501085013151168,
      "num_tokens": 17704503.0,
      "step": 8085
    },
    {
      "entropy": 5.511980485916138,
      "epoch": 0.8028979753870583,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000494415405830816,
      "loss": 5.4216,
      "mean_token_accuracy": 0.1577194944024086,
      "num_tokens": 17715902.0,
      "step": 8090
    },
    {
      "entropy": 5.563834285736084,
      "epoch": 0.8033942040492259,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004944075580310916,
      "loss": 5.3619,
      "mean_token_accuracy": 0.15873897820711136,
      "num_tokens": 17727305.0,
      "step": 8095
    },
    {
      "entropy": 5.544140625,
      "epoch": 0.8038904327113934,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943997047905325,
      "loss": 5.3088,
      "mean_token_accuracy": 0.16212920248508453,
      "num_tokens": 17737906.0,
      "step": 8100
    },
    {
      "entropy": 5.624442720413208,
      "epoch": 0.804386661373561,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943918461093335,
      "loss": 5.3524,
      "mean_token_accuracy": 0.15311113744974136,
      "num_tokens": 17749503.0,
      "step": 8105
    },
    {
      "entropy": 5.49956579208374,
      "epoch": 0.8048828900357284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943839819876894,
      "loss": 5.3287,
      "mean_token_accuracy": 0.16051498353481292,
      "num_tokens": 17759774.0,
      "step": 8110
    },
    {
      "entropy": 5.517120981216431,
      "epoch": 0.805379118697896,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004943761124257954,
      "loss": 5.3993,
      "mean_token_accuracy": 0.15647923052310944,
      "num_tokens": 17772101.0,
      "step": 8115
    },
    {
      "entropy": 5.587121152877808,
      "epoch": 0.8058753473600635,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004943682374238464,
      "loss": 5.3496,
      "mean_token_accuracy": 0.15384727865457534,
      "num_tokens": 17781923.0,
      "step": 8120
    },
    {
      "entropy": 5.530916690826416,
      "epoch": 0.8063715760222311,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000494360356982038,
      "loss": 5.3851,
      "mean_token_accuracy": 0.15419470369815827,
      "num_tokens": 17794787.0,
      "step": 8125
    },
    {
      "entropy": 5.619162082672119,
      "epoch": 0.8068678046843986,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943524711005653,
      "loss": 5.4023,
      "mean_token_accuracy": 0.15798399299383165,
      "num_tokens": 17805370.0,
      "step": 8130
    },
    {
      "entropy": 5.586359739303589,
      "epoch": 0.8073640333465661,
      "grad_norm": 0.890625,
      "learning_rate": 0.000494344579779624,
      "loss": 5.4421,
      "mean_token_accuracy": 0.14996136873960494,
      "num_tokens": 17816910.0,
      "step": 8135
    },
    {
      "entropy": 5.600253677368164,
      "epoch": 0.8078602620087336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004943366830194098,
      "loss": 5.4462,
      "mean_token_accuracy": 0.15920065343379974,
      "num_tokens": 17826792.0,
      "step": 8140
    },
    {
      "entropy": 5.524454498291016,
      "epoch": 0.8083564906709012,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004943287808201183,
      "loss": 5.3342,
      "mean_token_accuracy": 0.15956548005342483,
      "num_tokens": 17836986.0,
      "step": 8145
    },
    {
      "entropy": 5.555715751647949,
      "epoch": 0.8088527193330687,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943208731819459,
      "loss": 5.4294,
      "mean_token_accuracy": 0.14959967285394668,
      "num_tokens": 17849765.0,
      "step": 8150
    },
    {
      "entropy": 5.545638513565064,
      "epoch": 0.8093489479952362,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943129601050883,
      "loss": 5.2671,
      "mean_token_accuracy": 0.15734123289585114,
      "num_tokens": 17859666.0,
      "step": 8155
    },
    {
      "entropy": 5.5894801139831545,
      "epoch": 0.8098451766574037,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494305041589742,
      "loss": 5.3126,
      "mean_token_accuracy": 0.16374527662992477,
      "num_tokens": 17869543.0,
      "step": 8160
    },
    {
      "entropy": 5.535119533538818,
      "epoch": 0.8103414053195712,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004942971176361031,
      "loss": 5.366,
      "mean_token_accuracy": 0.16160104274749756,
      "num_tokens": 17880347.0,
      "step": 8165
    },
    {
      "entropy": 5.56554536819458,
      "epoch": 0.8108376339817388,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004942891882443682,
      "loss": 5.2239,
      "mean_token_accuracy": 0.16553542762994766,
      "num_tokens": 17889743.0,
      "step": 8170
    },
    {
      "entropy": 5.530171060562134,
      "epoch": 0.8113338626439063,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494281253414734,
      "loss": 5.2244,
      "mean_token_accuracy": 0.17022799849510192,
      "num_tokens": 17899853.0,
      "step": 8175
    },
    {
      "entropy": 5.508146524429321,
      "epoch": 0.8118300913060739,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942733131473973,
      "loss": 5.3161,
      "mean_token_accuracy": 0.16692925691604615,
      "num_tokens": 17909866.0,
      "step": 8180
    },
    {
      "entropy": 5.524717712402344,
      "epoch": 0.8123263199682413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942653674425548,
      "loss": 5.373,
      "mean_token_accuracy": 0.15655714571475982,
      "num_tokens": 17920086.0,
      "step": 8185
    },
    {
      "entropy": 5.588309621810913,
      "epoch": 0.8128225486304089,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004942574163004038,
      "loss": 5.3204,
      "mean_token_accuracy": 0.15204589068889618,
      "num_tokens": 17930529.0,
      "step": 8190
    },
    {
      "entropy": 5.624948120117187,
      "epoch": 0.8133187772925764,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004942494597211412,
      "loss": 5.4352,
      "mean_token_accuracy": 0.154278514534235,
      "num_tokens": 17940818.0,
      "step": 8195
    },
    {
      "entropy": 5.544540643692017,
      "epoch": 0.813815005954744,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942414977049646,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1577209435403347,
      "num_tokens": 17951241.0,
      "step": 8200
    },
    {
      "entropy": 5.62036099433899,
      "epoch": 0.8143112346169115,
      "grad_norm": 0.953125,
      "learning_rate": 0.000494233530252071,
      "loss": 5.4453,
      "mean_token_accuracy": 0.15579812377691268,
      "num_tokens": 17962353.0,
      "step": 8205
    },
    {
      "entropy": 5.577807283401489,
      "epoch": 0.814807463279079,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004942255573626585,
      "loss": 5.3356,
      "mean_token_accuracy": 0.15496407598257064,
      "num_tokens": 17973672.0,
      "step": 8210
    },
    {
      "entropy": 5.50259428024292,
      "epoch": 0.8153036919412465,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004942175790369244,
      "loss": 5.3826,
      "mean_token_accuracy": 0.15946246087551116,
      "num_tokens": 17985668.0,
      "step": 8215
    },
    {
      "entropy": 5.488262939453125,
      "epoch": 0.8157999206034141,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004942095952750669,
      "loss": 5.2891,
      "mean_token_accuracy": 0.16085183173418044,
      "num_tokens": 17996326.0,
      "step": 8220
    },
    {
      "entropy": 5.580463457107544,
      "epoch": 0.8162961492655816,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004942016060772837,
      "loss": 5.3716,
      "mean_token_accuracy": 0.15862178504467012,
      "num_tokens": 18007991.0,
      "step": 8225
    },
    {
      "entropy": 5.553548526763916,
      "epoch": 0.8167923779277491,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004941936114437731,
      "loss": 5.3495,
      "mean_token_accuracy": 0.15924760550260544,
      "num_tokens": 18019356.0,
      "step": 8230
    },
    {
      "entropy": 5.514718675613404,
      "epoch": 0.8172886065899166,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941856113747331,
      "loss": 5.3152,
      "mean_token_accuracy": 0.15735982432961465,
      "num_tokens": 18030131.0,
      "step": 8235
    },
    {
      "entropy": 5.505092906951904,
      "epoch": 0.8177848352520841,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004941776058703623,
      "loss": 5.2609,
      "mean_token_accuracy": 0.15927048176527023,
      "num_tokens": 18040848.0,
      "step": 8240
    },
    {
      "entropy": 5.638324928283692,
      "epoch": 0.8182810639142517,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004941695949308593,
      "loss": 5.5763,
      "mean_token_accuracy": 0.1455054424703121,
      "num_tokens": 18051888.0,
      "step": 8245
    },
    {
      "entropy": 5.619413995742798,
      "epoch": 0.8187772925764192,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004941615785564224,
      "loss": 5.3865,
      "mean_token_accuracy": 0.15458545535802842,
      "num_tokens": 18064720.0,
      "step": 8250
    },
    {
      "entropy": 5.586575651168824,
      "epoch": 0.8192735212385868,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941535567472506,
      "loss": 5.3898,
      "mean_token_accuracy": 0.16333946883678435,
      "num_tokens": 18075208.0,
      "step": 8255
    },
    {
      "entropy": 5.51199254989624,
      "epoch": 0.8197697499007542,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494145529503543,
      "loss": 5.3108,
      "mean_token_accuracy": 0.15912991464138032,
      "num_tokens": 18086717.0,
      "step": 8260
    },
    {
      "entropy": 5.447174215316773,
      "epoch": 0.8202659785629218,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004941374968254985,
      "loss": 5.2628,
      "mean_token_accuracy": 0.15956565588712693,
      "num_tokens": 18097593.0,
      "step": 8265
    },
    {
      "entropy": 5.528260135650635,
      "epoch": 0.8207622072250893,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004941294587133161,
      "loss": 5.3072,
      "mean_token_accuracy": 0.1594291463494301,
      "num_tokens": 18108470.0,
      "step": 8270
    },
    {
      "entropy": 5.569295883178711,
      "epoch": 0.8212584358872569,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941214151671954,
      "loss": 5.2871,
      "mean_token_accuracy": 0.16323021426796913,
      "num_tokens": 18119854.0,
      "step": 8275
    },
    {
      "entropy": 5.490022373199463,
      "epoch": 0.8217546645494244,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004941133661873357,
      "loss": 5.2839,
      "mean_token_accuracy": 0.15674965232610702,
      "num_tokens": 18130684.0,
      "step": 8280
    },
    {
      "entropy": 5.572978496551514,
      "epoch": 0.822250893211592,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004941053117739367,
      "loss": 5.3337,
      "mean_token_accuracy": 0.159407277405262,
      "num_tokens": 18140555.0,
      "step": 8285
    },
    {
      "entropy": 5.491158390045166,
      "epoch": 0.8227471218737594,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940972519271982,
      "loss": 5.2516,
      "mean_token_accuracy": 0.16636134535074235,
      "num_tokens": 18150478.0,
      "step": 8290
    },
    {
      "entropy": 5.587179803848267,
      "epoch": 0.823243350535927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940891866473199,
      "loss": 5.4481,
      "mean_token_accuracy": 0.14665500074625015,
      "num_tokens": 18160889.0,
      "step": 8295
    },
    {
      "entropy": 5.549789047241211,
      "epoch": 0.8237395791980945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004940811159345017,
      "loss": 5.3656,
      "mean_token_accuracy": 0.15916227251291276,
      "num_tokens": 18171086.0,
      "step": 8300
    },
    {
      "entropy": 5.6116612434387205,
      "epoch": 0.824235807860262,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004940730397889441,
      "loss": 5.3285,
      "mean_token_accuracy": 0.15945313423871993,
      "num_tokens": 18182711.0,
      "step": 8305
    },
    {
      "entropy": 5.549581146240234,
      "epoch": 0.8247320365224295,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004940649582108472,
      "loss": 5.382,
      "mean_token_accuracy": 0.15296767354011537,
      "num_tokens": 18193655.0,
      "step": 8310
    },
    {
      "entropy": 5.564826202392578,
      "epoch": 0.825228265184597,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004940568712004113,
      "loss": 5.3469,
      "mean_token_accuracy": 0.16141577661037446,
      "num_tokens": 18203921.0,
      "step": 8315
    },
    {
      "entropy": 5.5363771438598635,
      "epoch": 0.8257244938467646,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004940487787578369,
      "loss": 5.4392,
      "mean_token_accuracy": 0.1583532616496086,
      "num_tokens": 18215581.0,
      "step": 8320
    },
    {
      "entropy": 5.6165392875671385,
      "epoch": 0.8262207225089321,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000494040680883325,
      "loss": 5.3674,
      "mean_token_accuracy": 0.1542344093322754,
      "num_tokens": 18226273.0,
      "step": 8325
    },
    {
      "entropy": 5.54599962234497,
      "epoch": 0.8267169511710997,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004940325775770762,
      "loss": 5.3191,
      "mean_token_accuracy": 0.15517500787973404,
      "num_tokens": 18236653.0,
      "step": 8330
    },
    {
      "entropy": 5.5635011196136475,
      "epoch": 0.8272131798332671,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004940244688392913,
      "loss": 5.3864,
      "mean_token_accuracy": 0.15477572232484818,
      "num_tokens": 18246421.0,
      "step": 8335
    },
    {
      "entropy": 5.5829856395721436,
      "epoch": 0.8277094084954347,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940163546701716,
      "loss": 5.3967,
      "mean_token_accuracy": 0.16117299646139144,
      "num_tokens": 18257698.0,
      "step": 8340
    },
    {
      "entropy": 5.628828811645508,
      "epoch": 0.8282056371576022,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940082350699183,
      "loss": 5.4403,
      "mean_token_accuracy": 0.15531728863716127,
      "num_tokens": 18267499.0,
      "step": 8345
    },
    {
      "entropy": 5.563839769363403,
      "epoch": 0.8287018658197698,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940001100387325,
      "loss": 5.3745,
      "mean_token_accuracy": 0.15990617126226425,
      "num_tokens": 18279070.0,
      "step": 8350
    },
    {
      "entropy": 5.573468017578125,
      "epoch": 0.8291980944819373,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004939919795768161,
      "loss": 5.3165,
      "mean_token_accuracy": 0.15125838965177535,
      "num_tokens": 18289843.0,
      "step": 8355
    },
    {
      "entropy": 5.59449462890625,
      "epoch": 0.8296943231441049,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939838436843704,
      "loss": 5.3627,
      "mean_token_accuracy": 0.1598544552922249,
      "num_tokens": 18302089.0,
      "step": 8360
    },
    {
      "entropy": 5.5382894515991214,
      "epoch": 0.8301905518062723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004939757023615973,
      "loss": 5.428,
      "mean_token_accuracy": 0.1563342936336994,
      "num_tokens": 18313173.0,
      "step": 8365
    },
    {
      "entropy": 5.513939046859742,
      "epoch": 0.8306867804684399,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939675556086986,
      "loss": 5.2988,
      "mean_token_accuracy": 0.16526466459035874,
      "num_tokens": 18324108.0,
      "step": 8370
    },
    {
      "entropy": 5.580731296539307,
      "epoch": 0.8311830091306074,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939594034258763,
      "loss": 5.2598,
      "mean_token_accuracy": 0.16394211649894713,
      "num_tokens": 18334146.0,
      "step": 8375
    },
    {
      "entropy": 5.532615900039673,
      "epoch": 0.8316792377927749,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004939512458133326,
      "loss": 5.2762,
      "mean_token_accuracy": 0.1579000785946846,
      "num_tokens": 18345543.0,
      "step": 8380
    },
    {
      "entropy": 5.572006607055664,
      "epoch": 0.8321754664549424,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939430827712698,
      "loss": 5.3685,
      "mean_token_accuracy": 0.15714070051908494,
      "num_tokens": 18357923.0,
      "step": 8385
    },
    {
      "entropy": 5.551369333267212,
      "epoch": 0.8326716951171099,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939349142998905,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1533169686794281,
      "num_tokens": 18369211.0,
      "step": 8390
    },
    {
      "entropy": 5.600291538238525,
      "epoch": 0.8331679237792775,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939267403993968,
      "loss": 5.2672,
      "mean_token_accuracy": 0.15729450136423112,
      "num_tokens": 18380307.0,
      "step": 8395
    },
    {
      "entropy": 5.527000379562378,
      "epoch": 0.833664152441445,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493918561069992,
      "loss": 5.2871,
      "mean_token_accuracy": 0.159009350836277,
      "num_tokens": 18391099.0,
      "step": 8400
    },
    {
      "entropy": 5.502962112426758,
      "epoch": 0.8341603811036126,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004939103763118784,
      "loss": 5.3226,
      "mean_token_accuracy": 0.15802555084228515,
      "num_tokens": 18402882.0,
      "step": 8405
    },
    {
      "entropy": 5.601411819458008,
      "epoch": 0.83465660976578,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004939021861252593,
      "loss": 5.3794,
      "mean_token_accuracy": 0.154118911921978,
      "num_tokens": 18414815.0,
      "step": 8410
    },
    {
      "entropy": 5.586397886276245,
      "epoch": 0.8351528384279476,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004938939905103376,
      "loss": 5.3561,
      "mean_token_accuracy": 0.15807112753391267,
      "num_tokens": 18426394.0,
      "step": 8415
    },
    {
      "entropy": 5.574497938156128,
      "epoch": 0.8356490670901151,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004938857894673166,
      "loss": 5.2978,
      "mean_token_accuracy": 0.16356425285339354,
      "num_tokens": 18436147.0,
      "step": 8420
    },
    {
      "entropy": 5.501009845733643,
      "epoch": 0.8361452957522827,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938775829963997,
      "loss": 5.3098,
      "mean_token_accuracy": 0.16611545383930207,
      "num_tokens": 18447536.0,
      "step": 8425
    },
    {
      "entropy": 5.57528338432312,
      "epoch": 0.8366415244144502,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004938693710977904,
      "loss": 5.3387,
      "mean_token_accuracy": 0.1548042804002762,
      "num_tokens": 18458956.0,
      "step": 8430
    },
    {
      "entropy": 5.538770437240601,
      "epoch": 0.8371377530766178,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938611537716922,
      "loss": 5.3371,
      "mean_token_accuracy": 0.15586729645729064,
      "num_tokens": 18470490.0,
      "step": 8435
    },
    {
      "entropy": 5.522247743606568,
      "epoch": 0.8376339817387852,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004938529310183091,
      "loss": 5.3419,
      "mean_token_accuracy": 0.16169845461845397,
      "num_tokens": 18480512.0,
      "step": 8440
    },
    {
      "entropy": 5.554907417297363,
      "epoch": 0.8381302104009528,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493844702837845,
      "loss": 5.3264,
      "mean_token_accuracy": 0.15487321168184282,
      "num_tokens": 18492657.0,
      "step": 8445
    },
    {
      "entropy": 5.574155282974243,
      "epoch": 0.8386264390631203,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004938364692305036,
      "loss": 5.2844,
      "mean_token_accuracy": 0.16307801455259324,
      "num_tokens": 18504108.0,
      "step": 8450
    },
    {
      "entropy": 5.600583267211914,
      "epoch": 0.8391226677252878,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938282301964896,
      "loss": 5.4122,
      "mean_token_accuracy": 0.15657274425029755,
      "num_tokens": 18514760.0,
      "step": 8455
    },
    {
      "entropy": 5.612451696395874,
      "epoch": 0.8396188963874553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004938199857360068,
      "loss": 5.4077,
      "mean_token_accuracy": 0.15967583358287812,
      "num_tokens": 18526039.0,
      "step": 8460
    },
    {
      "entropy": 5.568667936325073,
      "epoch": 0.8401151250496228,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00049381173584926,
      "loss": 5.4513,
      "mean_token_accuracy": 0.1483660653233528,
      "num_tokens": 18537818.0,
      "step": 8465
    },
    {
      "entropy": 5.620733785629272,
      "epoch": 0.8406113537117904,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004938034805364535,
      "loss": 5.3966,
      "mean_token_accuracy": 0.1547456845641136,
      "num_tokens": 18549028.0,
      "step": 8470
    },
    {
      "entropy": 5.561071157455444,
      "epoch": 0.8411075823739579,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004937952197977923,
      "loss": 5.2677,
      "mean_token_accuracy": 0.1600969135761261,
      "num_tokens": 18559231.0,
      "step": 8475
    },
    {
      "entropy": 5.578726720809937,
      "epoch": 0.8416038110361255,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493786953633481,
      "loss": 5.3809,
      "mean_token_accuracy": 0.1585715189576149,
      "num_tokens": 18570232.0,
      "step": 8480
    },
    {
      "entropy": 5.6052491664886475,
      "epoch": 0.842100039698293,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004937786820437248,
      "loss": 5.3505,
      "mean_token_accuracy": 0.15627303272485732,
      "num_tokens": 18581856.0,
      "step": 8485
    },
    {
      "entropy": 5.601402282714844,
      "epoch": 0.8425962683604605,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004937704050287286,
      "loss": 5.3736,
      "mean_token_accuracy": 0.165021675825119,
      "num_tokens": 18592412.0,
      "step": 8490
    },
    {
      "entropy": 5.574635076522827,
      "epoch": 0.843092497022628,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937621225886978,
      "loss": 5.2658,
      "mean_token_accuracy": 0.16478906720876693,
      "num_tokens": 18602182.0,
      "step": 8495
    },
    {
      "entropy": 5.562773180007935,
      "epoch": 0.8435887256847956,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004937538347238378,
      "loss": 5.3855,
      "mean_token_accuracy": 0.15963244140148164,
      "num_tokens": 18612701.0,
      "step": 8500
    },
    {
      "entropy": 5.539256286621094,
      "epoch": 0.8440849543469631,
      "grad_norm": 0.84375,
      "learning_rate": 0.000493745541434354,
      "loss": 5.3358,
      "mean_token_accuracy": 0.15210745483636856,
      "num_tokens": 18624231.0,
      "step": 8505
    },
    {
      "entropy": 5.601503515243531,
      "epoch": 0.8445811830091307,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937372427204522,
      "loss": 5.32,
      "mean_token_accuracy": 0.1641618236899376,
      "num_tokens": 18634424.0,
      "step": 8510
    },
    {
      "entropy": 5.529595804214478,
      "epoch": 0.8450774116712981,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004937289385823382,
      "loss": 5.3274,
      "mean_token_accuracy": 0.16318677365779877,
      "num_tokens": 18646357.0,
      "step": 8515
    },
    {
      "entropy": 5.504410362243652,
      "epoch": 0.8455736403334657,
      "grad_norm": 1.375,
      "learning_rate": 0.0004937206290202176,
      "loss": 5.2407,
      "mean_token_accuracy": 0.15793229937553405,
      "num_tokens": 18657836.0,
      "step": 8520
    },
    {
      "entropy": 5.524774694442749,
      "epoch": 0.8460698689956332,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004937123140342969,
      "loss": 5.323,
      "mean_token_accuracy": 0.15458544045686723,
      "num_tokens": 18668426.0,
      "step": 8525
    },
    {
      "entropy": 5.563634967803955,
      "epoch": 0.8465660976578007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493703993624782,
      "loss": 5.2939,
      "mean_token_accuracy": 0.15594468265771866,
      "num_tokens": 18677975.0,
      "step": 8530
    },
    {
      "entropy": 5.60366268157959,
      "epoch": 0.8470623263199683,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004936956677918794,
      "loss": 5.393,
      "mean_token_accuracy": 0.15691497921943665,
      "num_tokens": 18689944.0,
      "step": 8535
    },
    {
      "entropy": 5.511834621429443,
      "epoch": 0.8475585549821357,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004936873365357955,
      "loss": 5.2934,
      "mean_token_accuracy": 0.15993214249610901,
      "num_tokens": 18700524.0,
      "step": 8540
    },
    {
      "entropy": 5.558034944534302,
      "epoch": 0.8480547836443033,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004936789998567368,
      "loss": 5.3182,
      "mean_token_accuracy": 0.15960853695869445,
      "num_tokens": 18711184.0,
      "step": 8545
    },
    {
      "entropy": 5.513194751739502,
      "epoch": 0.8485510123064708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004936706577549101,
      "loss": 5.309,
      "mean_token_accuracy": 0.15579880625009537,
      "num_tokens": 18721393.0,
      "step": 8550
    },
    {
      "entropy": 5.607612228393554,
      "epoch": 0.8490472409686384,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004936623102305223,
      "loss": 5.4469,
      "mean_token_accuracy": 0.15231576412916184,
      "num_tokens": 18732690.0,
      "step": 8555
    },
    {
      "entropy": 5.547397899627685,
      "epoch": 0.8495434696308058,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004936539572837804,
      "loss": 5.2531,
      "mean_token_accuracy": 0.1642311543226242,
      "num_tokens": 18744637.0,
      "step": 8560
    },
    {
      "entropy": 5.454610013961792,
      "epoch": 0.8500396982929734,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004936455989148916,
      "loss": 5.2184,
      "mean_token_accuracy": 0.16149047762155533,
      "num_tokens": 18754751.0,
      "step": 8565
    },
    {
      "entropy": 5.589911890029907,
      "epoch": 0.8505359269551409,
      "grad_norm": 0.875,
      "learning_rate": 0.0004936372351240629,
      "loss": 5.3843,
      "mean_token_accuracy": 0.158094185590744,
      "num_tokens": 18767412.0,
      "step": 8570
    },
    {
      "entropy": 5.537066173553467,
      "epoch": 0.8510321556173085,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000493628865911502,
      "loss": 5.3149,
      "mean_token_accuracy": 0.15923092216253282,
      "num_tokens": 18779287.0,
      "step": 8575
    },
    {
      "entropy": 5.541104936599732,
      "epoch": 0.851528384279476,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004936204912774163,
      "loss": 5.2961,
      "mean_token_accuracy": 0.15669662356376649,
      "num_tokens": 18790461.0,
      "step": 8580
    },
    {
      "entropy": 5.56708025932312,
      "epoch": 0.8520246129416436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004936121112220135,
      "loss": 5.2997,
      "mean_token_accuracy": 0.15767405852675437,
      "num_tokens": 18801310.0,
      "step": 8585
    },
    {
      "entropy": 5.544286823272705,
      "epoch": 0.852520841603811,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936037257455015,
      "loss": 5.34,
      "mean_token_accuracy": 0.16227839142084122,
      "num_tokens": 18811002.0,
      "step": 8590
    },
    {
      "entropy": 5.5590098857879635,
      "epoch": 0.8530170702659786,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493595334848088,
      "loss": 5.3206,
      "mean_token_accuracy": 0.16257548332214355,
      "num_tokens": 18821333.0,
      "step": 8595
    },
    {
      "entropy": 5.590433502197266,
      "epoch": 0.8535132989281461,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004935869385299813,
      "loss": 5.3826,
      "mean_token_accuracy": 0.15827727019786836,
      "num_tokens": 18832173.0,
      "step": 8600
    },
    {
      "entropy": 5.587846088409424,
      "epoch": 0.8540095275903136,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935785367913896,
      "loss": 5.3181,
      "mean_token_accuracy": 0.16261859536170958,
      "num_tokens": 18843298.0,
      "step": 8605
    },
    {
      "entropy": 5.600197887420654,
      "epoch": 0.8545057562524812,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004935701296325212,
      "loss": 5.3909,
      "mean_token_accuracy": 0.14815575927495955,
      "num_tokens": 18854720.0,
      "step": 8610
    },
    {
      "entropy": 5.531178092956543,
      "epoch": 0.8550019849146486,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935617170535845,
      "loss": 5.3357,
      "mean_token_accuracy": 0.16121273189783097,
      "num_tokens": 18866060.0,
      "step": 8615
    },
    {
      "entropy": 5.5127527713775635,
      "epoch": 0.8554982135768162,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004935532990547882,
      "loss": 5.3447,
      "mean_token_accuracy": 0.1572854921221733,
      "num_tokens": 18877646.0,
      "step": 8620
    },
    {
      "entropy": 5.54137806892395,
      "epoch": 0.8559944422389837,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004935448756363409,
      "loss": 5.3093,
      "mean_token_accuracy": 0.15891131907701492,
      "num_tokens": 18888426.0,
      "step": 8625
    },
    {
      "entropy": 5.564518165588379,
      "epoch": 0.8564906709011513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935364467984518,
      "loss": 5.3302,
      "mean_token_accuracy": 0.1565878227353096,
      "num_tokens": 18897985.0,
      "step": 8630
    },
    {
      "entropy": 5.573563051223755,
      "epoch": 0.8569868995633187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004935280125413295,
      "loss": 5.3766,
      "mean_token_accuracy": 0.15929335057735444,
      "num_tokens": 18908804.0,
      "step": 8635
    },
    {
      "entropy": 5.520459365844727,
      "epoch": 0.8574831282254863,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935195728651836,
      "loss": 5.2588,
      "mean_token_accuracy": 0.16084133535623552,
      "num_tokens": 18918347.0,
      "step": 8640
    },
    {
      "entropy": 5.509659910202027,
      "epoch": 0.8579793568876538,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493511127770223,
      "loss": 5.2039,
      "mean_token_accuracy": 0.1710530176758766,
      "num_tokens": 18927996.0,
      "step": 8645
    },
    {
      "entropy": 5.503261423110962,
      "epoch": 0.8584755855498214,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935026772566574,
      "loss": 5.261,
      "mean_token_accuracy": 0.16746046394109726,
      "num_tokens": 18939603.0,
      "step": 8650
    },
    {
      "entropy": 5.582817840576172,
      "epoch": 0.8589718142119889,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004934942213246961,
      "loss": 5.376,
      "mean_token_accuracy": 0.1575402185320854,
      "num_tokens": 18950721.0,
      "step": 8655
    },
    {
      "entropy": 5.477072143554688,
      "epoch": 0.8594680428741565,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493485759974549,
      "loss": 5.2478,
      "mean_token_accuracy": 0.16302145421504974,
      "num_tokens": 18961666.0,
      "step": 8660
    },
    {
      "entropy": 5.49755449295044,
      "epoch": 0.8599642715363239,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934772932064259,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1639523386955261,
      "num_tokens": 18972223.0,
      "step": 8665
    },
    {
      "entropy": 5.534092044830322,
      "epoch": 0.8604605001984915,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004934688210205367,
      "loss": 5.3116,
      "mean_token_accuracy": 0.1588273122906685,
      "num_tokens": 18982983.0,
      "step": 8670
    },
    {
      "entropy": 5.551003551483154,
      "epoch": 0.860956728860659,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004934603434170913,
      "loss": 5.3187,
      "mean_token_accuracy": 0.1562140092253685,
      "num_tokens": 18993521.0,
      "step": 8675
    },
    {
      "entropy": 5.591272783279419,
      "epoch": 0.8614529575228265,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004934518603963003,
      "loss": 5.4258,
      "mean_token_accuracy": 0.15756160765886307,
      "num_tokens": 19004901.0,
      "step": 8680
    },
    {
      "entropy": 5.465965509414673,
      "epoch": 0.861949186184994,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004934433719583739,
      "loss": 5.2361,
      "mean_token_accuracy": 0.1667258471250534,
      "num_tokens": 19016123.0,
      "step": 8685
    },
    {
      "entropy": 5.610170507431031,
      "epoch": 0.8624454148471615,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004934348781035226,
      "loss": 5.3787,
      "mean_token_accuracy": 0.15522417202591896,
      "num_tokens": 19026760.0,
      "step": 8690
    },
    {
      "entropy": 5.5450029373168945,
      "epoch": 0.8629416435093291,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493426378831957,
      "loss": 5.2536,
      "mean_token_accuracy": 0.16278370171785356,
      "num_tokens": 19036665.0,
      "step": 8695
    },
    {
      "entropy": 5.495017719268799,
      "epoch": 0.8634378721714966,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004934178741438877,
      "loss": 5.3256,
      "mean_token_accuracy": 0.15975018590688705,
      "num_tokens": 19047779.0,
      "step": 8700
    },
    {
      "entropy": 5.536338472366333,
      "epoch": 0.8639341008336642,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493409364039526,
      "loss": 5.3778,
      "mean_token_accuracy": 0.1555515468120575,
      "num_tokens": 19058292.0,
      "step": 8705
    },
    {
      "entropy": 5.616223526000977,
      "epoch": 0.8644303294958317,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004934008485190825,
      "loss": 5.3351,
      "mean_token_accuracy": 0.15970531478524208,
      "num_tokens": 19069938.0,
      "step": 8710
    },
    {
      "entropy": 5.615974855422974,
      "epoch": 0.8649265581579992,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933923275827687,
      "loss": 5.3552,
      "mean_token_accuracy": 0.15216721892356871,
      "num_tokens": 19080492.0,
      "step": 8715
    },
    {
      "entropy": 5.555372810363769,
      "epoch": 0.8654227868201667,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004933838012307957,
      "loss": 5.4375,
      "mean_token_accuracy": 0.15620280653238297,
      "num_tokens": 19091723.0,
      "step": 8720
    },
    {
      "entropy": 5.577514171600342,
      "epoch": 0.8659190154823343,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493375269463375,
      "loss": 5.3194,
      "mean_token_accuracy": 0.15924772024154663,
      "num_tokens": 19102652.0,
      "step": 8725
    },
    {
      "entropy": 5.515443992614746,
      "epoch": 0.8664152441445018,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004933667322807182,
      "loss": 5.2214,
      "mean_token_accuracy": 0.1621010720729828,
      "num_tokens": 19114139.0,
      "step": 8730
    },
    {
      "entropy": 5.529138040542603,
      "epoch": 0.8669114728066694,
      "grad_norm": 0.875,
      "learning_rate": 0.000493358189683037,
      "loss": 5.3648,
      "mean_token_accuracy": 0.15938409119844438,
      "num_tokens": 19126293.0,
      "step": 8735
    },
    {
      "entropy": 5.512266826629639,
      "epoch": 0.8674077014688368,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933496416705432,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1681327149271965,
      "num_tokens": 19136947.0,
      "step": 8740
    },
    {
      "entropy": 5.563701915740967,
      "epoch": 0.8679039301310044,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004933410882434488,
      "loss": 5.2978,
      "mean_token_accuracy": 0.16255763173103333,
      "num_tokens": 19146987.0,
      "step": 8745
    },
    {
      "entropy": 5.525736665725708,
      "epoch": 0.8684001587931719,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933325294019659,
      "loss": 5.3537,
      "mean_token_accuracy": 0.15458540469408036,
      "num_tokens": 19157972.0,
      "step": 8750
    },
    {
      "entropy": 5.550621604919433,
      "epoch": 0.8688963874553394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933239651463068,
      "loss": 5.3722,
      "mean_token_accuracy": 0.151882254332304,
      "num_tokens": 19170313.0,
      "step": 8755
    },
    {
      "entropy": 5.632871627807617,
      "epoch": 0.869392616117507,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004933153954766837,
      "loss": 5.3722,
      "mean_token_accuracy": 0.15413001775741578,
      "num_tokens": 19183134.0,
      "step": 8760
    },
    {
      "entropy": 5.589765787124634,
      "epoch": 0.8698888447796744,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004933068203933092,
      "loss": 5.3011,
      "mean_token_accuracy": 0.16230310052633284,
      "num_tokens": 19194900.0,
      "step": 8765
    },
    {
      "entropy": 5.504223489761353,
      "epoch": 0.870385073441842,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493298239896396,
      "loss": 5.3527,
      "mean_token_accuracy": 0.15844140350818633,
      "num_tokens": 19204980.0,
      "step": 8770
    },
    {
      "entropy": 5.5504638195037845,
      "epoch": 0.8708813021040095,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004932896539861568,
      "loss": 5.3908,
      "mean_token_accuracy": 0.15713052749633788,
      "num_tokens": 19217581.0,
      "step": 8775
    },
    {
      "entropy": 5.552046585083008,
      "epoch": 0.8713775307661771,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932810626628047,
      "loss": 5.2394,
      "mean_token_accuracy": 0.16367227733135223,
      "num_tokens": 19227694.0,
      "step": 8780
    },
    {
      "entropy": 5.566386556625366,
      "epoch": 0.8718737594283446,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004932724659265526,
      "loss": 5.3709,
      "mean_token_accuracy": 0.15730719715356828,
      "num_tokens": 19238579.0,
      "step": 8785
    },
    {
      "entropy": 5.520002651214599,
      "epoch": 0.8723699880905121,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932638637776134,
      "loss": 5.3082,
      "mean_token_accuracy": 0.16142989099025726,
      "num_tokens": 19249658.0,
      "step": 8790
    },
    {
      "entropy": 5.497355270385742,
      "epoch": 0.8728662167526796,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000493255256216201,
      "loss": 5.289,
      "mean_token_accuracy": 0.1629670575261116,
      "num_tokens": 19260273.0,
      "step": 8795
    },
    {
      "entropy": 5.599217939376831,
      "epoch": 0.8733624454148472,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004932466432425284,
      "loss": 5.3519,
      "mean_token_accuracy": 0.15759664326906203,
      "num_tokens": 19270344.0,
      "step": 8800
    },
    {
      "entropy": 5.551671886444092,
      "epoch": 0.8738586740770147,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004932380248568094,
      "loss": 5.2333,
      "mean_token_accuracy": 0.16226085126399994,
      "num_tokens": 19280934.0,
      "step": 8805
    },
    {
      "entropy": 5.565516471862793,
      "epoch": 0.8743549027391823,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004932294010592575,
      "loss": 5.293,
      "mean_token_accuracy": 0.15469006597995758,
      "num_tokens": 19291028.0,
      "step": 8810
    },
    {
      "entropy": 5.511228561401367,
      "epoch": 0.8748511314013497,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004932207718500868,
      "loss": 5.2727,
      "mean_token_accuracy": 0.16160213500261306,
      "num_tokens": 19302436.0,
      "step": 8815
    },
    {
      "entropy": 5.582405948638916,
      "epoch": 0.8753473600635173,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932121372295112,
      "loss": 5.3116,
      "mean_token_accuracy": 0.1588225394487381,
      "num_tokens": 19312650.0,
      "step": 8820
    },
    {
      "entropy": 5.4919774532318115,
      "epoch": 0.8758435887256848,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004932034971977448,
      "loss": 5.2375,
      "mean_token_accuracy": 0.1643138885498047,
      "num_tokens": 19323727.0,
      "step": 8825
    },
    {
      "entropy": 5.513283205032349,
      "epoch": 0.8763398173878523,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931948517550019,
      "loss": 5.2623,
      "mean_token_accuracy": 0.15896197855472566,
      "num_tokens": 19333286.0,
      "step": 8830
    },
    {
      "entropy": 5.556285333633423,
      "epoch": 0.8768360460500199,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004931862009014968,
      "loss": 5.2972,
      "mean_token_accuracy": 0.16313159018754958,
      "num_tokens": 19343766.0,
      "step": 8835
    },
    {
      "entropy": 5.5845636367797855,
      "epoch": 0.8773322747121873,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000493177544637444,
      "loss": 5.2833,
      "mean_token_accuracy": 0.15809592604637146,
      "num_tokens": 19354944.0,
      "step": 8840
    },
    {
      "entropy": 5.58414397239685,
      "epoch": 0.8778285033743549,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004931688829630582,
      "loss": 5.3346,
      "mean_token_accuracy": 0.15988069474697114,
      "num_tokens": 19366832.0,
      "step": 8845
    },
    {
      "entropy": 5.593709993362427,
      "epoch": 0.8783247320365224,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004931602158785544,
      "loss": 5.4129,
      "mean_token_accuracy": 0.15583591908216476,
      "num_tokens": 19379067.0,
      "step": 8850
    },
    {
      "entropy": 5.5948878765106205,
      "epoch": 0.87882096069869,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931515433841472,
      "loss": 5.3164,
      "mean_token_accuracy": 0.160463547706604,
      "num_tokens": 19390419.0,
      "step": 8855
    },
    {
      "entropy": 5.57219123840332,
      "epoch": 0.8793171893608575,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931428654800518,
      "loss": 5.3547,
      "mean_token_accuracy": 0.14935329109430312,
      "num_tokens": 19401482.0,
      "step": 8860
    },
    {
      "entropy": 5.530307388305664,
      "epoch": 0.879813418023025,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004931341821664834,
      "loss": 5.2709,
      "mean_token_accuracy": 0.16195659786462785,
      "num_tokens": 19411620.0,
      "step": 8865
    },
    {
      "entropy": 5.545097923278808,
      "epoch": 0.8803096466851925,
      "grad_norm": 1.0,
      "learning_rate": 0.0004931254934436573,
      "loss": 5.3504,
      "mean_token_accuracy": 0.15893469452857972,
      "num_tokens": 19422247.0,
      "step": 8870
    },
    {
      "entropy": 5.555587673187256,
      "epoch": 0.8808058753473601,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931167993117891,
      "loss": 5.3262,
      "mean_token_accuracy": 0.15557404607534409,
      "num_tokens": 19433687.0,
      "step": 8875
    },
    {
      "entropy": 5.574374008178711,
      "epoch": 0.8813021040095276,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004931080997710942,
      "loss": 5.2308,
      "mean_token_accuracy": 0.1704422876238823,
      "num_tokens": 19443884.0,
      "step": 8880
    },
    {
      "entropy": 5.597978925704956,
      "epoch": 0.8817983326716952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004930993948217884,
      "loss": 5.4021,
      "mean_token_accuracy": 0.15130480825901033,
      "num_tokens": 19456425.0,
      "step": 8885
    },
    {
      "entropy": 5.643082952499389,
      "epoch": 0.8822945613338626,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004930906844640876,
      "loss": 5.3843,
      "mean_token_accuracy": 0.15100670605897903,
      "num_tokens": 19468838.0,
      "step": 8890
    },
    {
      "entropy": 5.542707681655884,
      "epoch": 0.8827907899960302,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004930819686982077,
      "loss": 5.2687,
      "mean_token_accuracy": 0.15994668304920195,
      "num_tokens": 19479603.0,
      "step": 8895
    },
    {
      "entropy": 5.4817603588104244,
      "epoch": 0.8832870186581977,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493073247524365,
      "loss": 5.2755,
      "mean_token_accuracy": 0.1608395367860794,
      "num_tokens": 19491793.0,
      "step": 8900
    },
    {
      "entropy": 5.549676322937012,
      "epoch": 0.8837832473203652,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004930645209427756,
      "loss": 5.3019,
      "mean_token_accuracy": 0.15942948311567307,
      "num_tokens": 19502869.0,
      "step": 8905
    },
    {
      "entropy": 5.574150133132934,
      "epoch": 0.8842794759825328,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493055788953656,
      "loss": 5.2684,
      "mean_token_accuracy": 0.1655476212501526,
      "num_tokens": 19513766.0,
      "step": 8910
    },
    {
      "entropy": 5.485655355453491,
      "epoch": 0.8847757046447002,
      "grad_norm": 1.875,
      "learning_rate": 0.0004930470515572227,
      "loss": 5.2868,
      "mean_token_accuracy": 0.1672094777226448,
      "num_tokens": 19524561.0,
      "step": 8915
    },
    {
      "entropy": 5.5656825542449955,
      "epoch": 0.8852719333068678,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930383087536923,
      "loss": 5.2315,
      "mean_token_accuracy": 0.16222631484270095,
      "num_tokens": 19535253.0,
      "step": 8920
    },
    {
      "entropy": 5.51240873336792,
      "epoch": 0.8857681619690353,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004930295605432818,
      "loss": 5.2742,
      "mean_token_accuracy": 0.1626346930861473,
      "num_tokens": 19546650.0,
      "step": 8925
    },
    {
      "entropy": 5.558320569992065,
      "epoch": 0.8862643906312029,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004930208069262079,
      "loss": 5.3107,
      "mean_token_accuracy": 0.16672074943780898,
      "num_tokens": 19558012.0,
      "step": 8930
    },
    {
      "entropy": 5.527715873718262,
      "epoch": 0.8867606192933704,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004930120479026878,
      "loss": 5.3234,
      "mean_token_accuracy": 0.16105586439371108,
      "num_tokens": 19569491.0,
      "step": 8935
    },
    {
      "entropy": 5.501845407485962,
      "epoch": 0.8872568479555379,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004930032834729388,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1616614803671837,
      "num_tokens": 19581116.0,
      "step": 8940
    },
    {
      "entropy": 5.571308851242065,
      "epoch": 0.8877530766177054,
      "grad_norm": 1.0,
      "learning_rate": 0.000492994513637178,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1554987534880638,
      "num_tokens": 19591351.0,
      "step": 8945
    },
    {
      "entropy": 5.548324012756348,
      "epoch": 0.888249305279873,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929857383956229,
      "loss": 5.3143,
      "mean_token_accuracy": 0.16284678280353546,
      "num_tokens": 19602071.0,
      "step": 8950
    },
    {
      "entropy": 5.590748453140259,
      "epoch": 0.8887455339420405,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004929769577484914,
      "loss": 5.3385,
      "mean_token_accuracy": 0.16362452059984206,
      "num_tokens": 19612799.0,
      "step": 8955
    },
    {
      "entropy": 5.573906517028808,
      "epoch": 0.8892417626042081,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000492968171696001,
      "loss": 5.2898,
      "mean_token_accuracy": 0.16145564019680023,
      "num_tokens": 19623901.0,
      "step": 8960
    },
    {
      "entropy": 5.558178043365478,
      "epoch": 0.8897379912663755,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004929593802383696,
      "loss": 5.2239,
      "mean_token_accuracy": 0.16059719473123552,
      "num_tokens": 19633963.0,
      "step": 8965
    },
    {
      "entropy": 5.476822328567505,
      "epoch": 0.8902342199285431,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004929505833758152,
      "loss": 5.2375,
      "mean_token_accuracy": 0.16544606238603593,
      "num_tokens": 19644009.0,
      "step": 8970
    },
    {
      "entropy": 5.496410846710205,
      "epoch": 0.8907304485907106,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004929417811085561,
      "loss": 5.1847,
      "mean_token_accuracy": 0.17268571853637696,
      "num_tokens": 19654051.0,
      "step": 8975
    },
    {
      "entropy": 5.537240362167358,
      "epoch": 0.8912266772528781,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004929329734368103,
      "loss": 5.3408,
      "mean_token_accuracy": 0.16192645728588104,
      "num_tokens": 19664118.0,
      "step": 8980
    },
    {
      "entropy": 5.52949709892273,
      "epoch": 0.8917229059150457,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929241603607964,
      "loss": 5.3102,
      "mean_token_accuracy": 0.15772383213043212,
      "num_tokens": 19674598.0,
      "step": 8985
    },
    {
      "entropy": 5.463305282592773,
      "epoch": 0.8922191345772131,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000492915341880733,
      "loss": 5.3153,
      "mean_token_accuracy": 0.15742891877889634,
      "num_tokens": 19686311.0,
      "step": 8990
    },
    {
      "entropy": 5.555131101608277,
      "epoch": 0.8927153632393807,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929065179968387,
      "loss": 5.3511,
      "mean_token_accuracy": 0.15476148948073387,
      "num_tokens": 19697518.0,
      "step": 8995
    },
    {
      "entropy": 5.601098918914795,
      "epoch": 0.8932115919015482,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004928976887093321,
      "loss": 5.3463,
      "mean_token_accuracy": 0.16341836452484132,
      "num_tokens": 19709768.0,
      "step": 9000
    },
    {
      "epoch": 0.8932115919015482,
      "eval_entropy": 5.427808056480388,
      "eval_loss": 5.369091033935547,
      "eval_mean_token_accuracy": 0.1621848368168897,
      "eval_num_tokens": 19709768.0,
      "eval_runtime": 20.5816,
      "eval_samples_per_second": 1635.875,
      "eval_steps_per_second": 204.503,
      "step": 9000
    },
    {
      "entropy": 5.5485231399536135,
      "epoch": 0.8937078205637158,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928888540184325,
      "loss": 5.2369,
      "mean_token_accuracy": 0.16167478263378143,
      "num_tokens": 19721154.0,
      "step": 9005
    },
    {
      "entropy": 5.5139179706573485,
      "epoch": 0.8942040492258833,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004928800139243587,
      "loss": 5.2177,
      "mean_token_accuracy": 0.16153546422719955,
      "num_tokens": 19731965.0,
      "step": 9010
    },
    {
      "entropy": 5.5536058902740475,
      "epoch": 0.8947002778880508,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004928711684273302,
      "loss": 5.3528,
      "mean_token_accuracy": 0.15001175850629805,
      "num_tokens": 19742385.0,
      "step": 9015
    },
    {
      "entropy": 5.604755449295044,
      "epoch": 0.8951965065502183,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004928623175275661,
      "loss": 5.3552,
      "mean_token_accuracy": 0.15330888330936432,
      "num_tokens": 19753027.0,
      "step": 9020
    },
    {
      "entropy": 5.5971746921539305,
      "epoch": 0.8956927352123859,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004928534612252862,
      "loss": 5.3244,
      "mean_token_accuracy": 0.15479455143213272,
      "num_tokens": 19763625.0,
      "step": 9025
    },
    {
      "entropy": 5.587306070327759,
      "epoch": 0.8961889638745534,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004928445995207095,
      "loss": 5.311,
      "mean_token_accuracy": 0.15927476137876512,
      "num_tokens": 19775425.0,
      "step": 9030
    },
    {
      "entropy": 5.550523567199707,
      "epoch": 0.896685192536721,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004928357324140564,
      "loss": 5.3355,
      "mean_token_accuracy": 0.15736784636974335,
      "num_tokens": 19786343.0,
      "step": 9035
    },
    {
      "entropy": 5.503166532516479,
      "epoch": 0.8971814211988884,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928268599055465,
      "loss": 5.2283,
      "mean_token_accuracy": 0.16160772293806075,
      "num_tokens": 19796903.0,
      "step": 9040
    },
    {
      "entropy": 5.60651044845581,
      "epoch": 0.897677649861056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004928179819953998,
      "loss": 5.3501,
      "mean_token_accuracy": 0.1579892262816429,
      "num_tokens": 19808026.0,
      "step": 9045
    },
    {
      "entropy": 5.574759387969971,
      "epoch": 0.8981738785232235,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928090986838366,
      "loss": 5.289,
      "mean_token_accuracy": 0.15758469551801682,
      "num_tokens": 19819343.0,
      "step": 9050
    },
    {
      "entropy": 5.610906744003296,
      "epoch": 0.898670107185391,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492800209971077,
      "loss": 5.3787,
      "mean_token_accuracy": 0.1635238617658615,
      "num_tokens": 19831264.0,
      "step": 9055
    },
    {
      "entropy": 5.629992961883545,
      "epoch": 0.8991663358475586,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004927913158573416,
      "loss": 5.3809,
      "mean_token_accuracy": 0.15479213893413543,
      "num_tokens": 19842101.0,
      "step": 9060
    },
    {
      "entropy": 5.535437154769897,
      "epoch": 0.899662564509726,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927824163428508,
      "loss": 5.3121,
      "mean_token_accuracy": 0.15996735841035842,
      "num_tokens": 19853789.0,
      "step": 9065
    },
    {
      "entropy": 5.520318031311035,
      "epoch": 0.9001587931718936,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927735114278254,
      "loss": 5.2752,
      "mean_token_accuracy": 0.15823260098695754,
      "num_tokens": 19865855.0,
      "step": 9070
    },
    {
      "entropy": 5.5822607517242435,
      "epoch": 0.9006550218340611,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927646011124861,
      "loss": 5.414,
      "mean_token_accuracy": 0.15900951027870178,
      "num_tokens": 19877443.0,
      "step": 9075
    },
    {
      "entropy": 5.548061513900757,
      "epoch": 0.9011512504962287,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004927556853970539,
      "loss": 5.2603,
      "mean_token_accuracy": 0.1633693531155586,
      "num_tokens": 19889503.0,
      "step": 9080
    },
    {
      "entropy": 5.627478122711182,
      "epoch": 0.9016474791583962,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927467642817499,
      "loss": 5.3652,
      "mean_token_accuracy": 0.15618871301412582,
      "num_tokens": 19899969.0,
      "step": 9085
    },
    {
      "entropy": 5.570292568206787,
      "epoch": 0.9021437078205637,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004927378377667955,
      "loss": 5.3334,
      "mean_token_accuracy": 0.15900463610887527,
      "num_tokens": 19911326.0,
      "step": 9090
    },
    {
      "entropy": 5.612566995620727,
      "epoch": 0.9026399364827312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492728905852412,
      "loss": 5.3307,
      "mean_token_accuracy": 0.15698612183332444,
      "num_tokens": 19922297.0,
      "step": 9095
    },
    {
      "entropy": 5.607271909713745,
      "epoch": 0.9031361651448988,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927199685388205,
      "loss": 5.3121,
      "mean_token_accuracy": 0.15922553539276124,
      "num_tokens": 19934727.0,
      "step": 9100
    },
    {
      "entropy": 5.5068422794342045,
      "epoch": 0.9036323938070663,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000492711025826243,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16628592163324357,
      "num_tokens": 19945485.0,
      "step": 9105
    },
    {
      "entropy": 5.532880020141602,
      "epoch": 0.9041286224692339,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004927020777149014,
      "loss": 5.2887,
      "mean_token_accuracy": 0.16440802812576294,
      "num_tokens": 19956561.0,
      "step": 9110
    },
    {
      "entropy": 5.563728666305542,
      "epoch": 0.9046248511314013,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926931242050171,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1543963611125946,
      "num_tokens": 19967775.0,
      "step": 9115
    },
    {
      "entropy": 5.5880828380584715,
      "epoch": 0.9051210797935689,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926841652968125,
      "loss": 5.3567,
      "mean_token_accuracy": 0.15127435475587844,
      "num_tokens": 19978935.0,
      "step": 9120
    },
    {
      "entropy": 5.499078464508057,
      "epoch": 0.9056173084557364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926752009905099,
      "loss": 5.2092,
      "mean_token_accuracy": 0.16924717873334885,
      "num_tokens": 19989137.0,
      "step": 9125
    },
    {
      "entropy": 5.444391250610352,
      "epoch": 0.9061135371179039,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004926662312863311,
      "loss": 5.206,
      "mean_token_accuracy": 0.1635851413011551,
      "num_tokens": 20000507.0,
      "step": 9130
    },
    {
      "entropy": 5.572239875793457,
      "epoch": 0.9066097657800715,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004926572561844988,
      "loss": 5.305,
      "mean_token_accuracy": 0.1622836172580719,
      "num_tokens": 20010907.0,
      "step": 9135
    },
    {
      "entropy": 5.584706401824951,
      "epoch": 0.9071059944422389,
      "grad_norm": 0.875,
      "learning_rate": 0.0004926482756852356,
      "loss": 5.2854,
      "mean_token_accuracy": 0.16619354337453843,
      "num_tokens": 20021403.0,
      "step": 9140
    },
    {
      "entropy": 5.545332384109497,
      "epoch": 0.9076022231044065,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926392897887642,
      "loss": 5.2932,
      "mean_token_accuracy": 0.16017937362194062,
      "num_tokens": 20032409.0,
      "step": 9145
    },
    {
      "entropy": 5.542720603942871,
      "epoch": 0.908098451766574,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926302984953073,
      "loss": 5.2247,
      "mean_token_accuracy": 0.16138189882040024,
      "num_tokens": 20043197.0,
      "step": 9150
    },
    {
      "entropy": 5.557456922531128,
      "epoch": 0.9085946804287416,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004926213018050881,
      "loss": 5.3423,
      "mean_token_accuracy": 0.1589226469397545,
      "num_tokens": 20054496.0,
      "step": 9155
    },
    {
      "entropy": 5.578097820281982,
      "epoch": 0.9090909090909091,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926122997183295,
      "loss": 5.386,
      "mean_token_accuracy": 0.15607479214668274,
      "num_tokens": 20065639.0,
      "step": 9160
    },
    {
      "entropy": 5.6064369678497314,
      "epoch": 0.9095871377530766,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004926032922352546,
      "loss": 5.2826,
      "mean_token_accuracy": 0.15926680564880372,
      "num_tokens": 20077354.0,
      "step": 9165
    },
    {
      "entropy": 5.577102565765381,
      "epoch": 0.9100833664152441,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004925942793560872,
      "loss": 5.3162,
      "mean_token_accuracy": 0.16004526615142822,
      "num_tokens": 20089088.0,
      "step": 9170
    },
    {
      "entropy": 5.668724918365479,
      "epoch": 0.9105795950774117,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925852610810504,
      "loss": 5.4843,
      "mean_token_accuracy": 0.15233470648527145,
      "num_tokens": 20101496.0,
      "step": 9175
    },
    {
      "entropy": 5.6101014614105225,
      "epoch": 0.9110758237395792,
      "grad_norm": 0.84375,
      "learning_rate": 0.000492576237410368,
      "loss": 5.2895,
      "mean_token_accuracy": 0.15828832983970642,
      "num_tokens": 20113682.0,
      "step": 9180
    },
    {
      "entropy": 5.405985546112061,
      "epoch": 0.9115720524017468,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004925672083442638,
      "loss": 5.155,
      "mean_token_accuracy": 0.17822301536798477,
      "num_tokens": 20123831.0,
      "step": 9185
    },
    {
      "entropy": 5.516451358795166,
      "epoch": 0.9120682810639142,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004925581738829616,
      "loss": 5.2897,
      "mean_token_accuracy": 0.16344057470560075,
      "num_tokens": 20133414.0,
      "step": 9190
    },
    {
      "entropy": 5.5019396305084225,
      "epoch": 0.9125645097260818,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004925491340266855,
      "loss": 5.2409,
      "mean_token_accuracy": 0.16906660348176955,
      "num_tokens": 20145120.0,
      "step": 9195
    },
    {
      "entropy": 5.540193891525268,
      "epoch": 0.9130607383882493,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004925400887756596,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1590275824069977,
      "num_tokens": 20156807.0,
      "step": 9200
    },
    {
      "entropy": 5.586658000946045,
      "epoch": 0.9135569670504168,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925310381301083,
      "loss": 5.3054,
      "mean_token_accuracy": 0.15902692973613738,
      "num_tokens": 20166775.0,
      "step": 9205
    },
    {
      "entropy": 5.5219961643219,
      "epoch": 0.9140531957125844,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000492521982090256,
      "loss": 5.237,
      "mean_token_accuracy": 0.1685601606965065,
      "num_tokens": 20178695.0,
      "step": 9210
    },
    {
      "entropy": 5.620602703094482,
      "epoch": 0.9145494243747518,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004925129206563273,
      "loss": 5.3412,
      "mean_token_accuracy": 0.16078450083732604,
      "num_tokens": 20190222.0,
      "step": 9215
    },
    {
      "entropy": 5.556102275848389,
      "epoch": 0.9150456530369194,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004925038538285467,
      "loss": 5.2951,
      "mean_token_accuracy": 0.15934197455644608,
      "num_tokens": 20202519.0,
      "step": 9220
    },
    {
      "entropy": 5.551261854171753,
      "epoch": 0.9155418816990869,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004924947816071394,
      "loss": 5.4085,
      "mean_token_accuracy": 0.15333180576562883,
      "num_tokens": 20214685.0,
      "step": 9225
    },
    {
      "entropy": 5.668244314193726,
      "epoch": 0.9160381103612545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049248570399233,
      "loss": 5.3504,
      "mean_token_accuracy": 0.15317084193229674,
      "num_tokens": 20225597.0,
      "step": 9230
    },
    {
      "entropy": 5.579620599746704,
      "epoch": 0.916534339023422,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492476620984344,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1628316804766655,
      "num_tokens": 20237067.0,
      "step": 9235
    },
    {
      "entropy": 5.626915979385376,
      "epoch": 0.9170305676855895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924675325834063,
      "loss": 5.368,
      "mean_token_accuracy": 0.15557292774319648,
      "num_tokens": 20247587.0,
      "step": 9240
    },
    {
      "entropy": 5.576964426040649,
      "epoch": 0.917526796347757,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924584387897424,
      "loss": 5.3514,
      "mean_token_accuracy": 0.16091438978910447,
      "num_tokens": 20259094.0,
      "step": 9245
    },
    {
      "entropy": 5.575494718551636,
      "epoch": 0.9180230250099246,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004924493396035777,
      "loss": 5.2696,
      "mean_token_accuracy": 0.1690424159169197,
      "num_tokens": 20269776.0,
      "step": 9250
    },
    {
      "entropy": 5.5706860542297365,
      "epoch": 0.9185192536720921,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000492440235025138,
      "loss": 5.2895,
      "mean_token_accuracy": 0.15555474907159805,
      "num_tokens": 20281814.0,
      "step": 9255
    },
    {
      "entropy": 5.538059329986572,
      "epoch": 0.9190154823342597,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004924311250546491,
      "loss": 5.341,
      "mean_token_accuracy": 0.15312860012054444,
      "num_tokens": 20293185.0,
      "step": 9260
    },
    {
      "entropy": 5.5473240375518795,
      "epoch": 0.9195117109964271,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004924220096923367,
      "loss": 5.2321,
      "mean_token_accuracy": 0.16510707437992095,
      "num_tokens": 20305097.0,
      "step": 9265
    },
    {
      "entropy": 5.538658380508423,
      "epoch": 0.9200079396585947,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004924128889384271,
      "loss": 5.2871,
      "mean_token_accuracy": 0.16214432269334794,
      "num_tokens": 20316257.0,
      "step": 9270
    },
    {
      "entropy": 5.528434991836548,
      "epoch": 0.9205041683207622,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924037627931463,
      "loss": 5.2481,
      "mean_token_accuracy": 0.1619844377040863,
      "num_tokens": 20326209.0,
      "step": 9275
    },
    {
      "entropy": 5.585993003845215,
      "epoch": 0.9210003969829297,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004923946312567206,
      "loss": 5.2805,
      "mean_token_accuracy": 0.16394528448581697,
      "num_tokens": 20337665.0,
      "step": 9280
    },
    {
      "entropy": 5.556754875183105,
      "epoch": 0.9214966256450973,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004923854943293765,
      "loss": 5.3203,
      "mean_token_accuracy": 0.1548920065164566,
      "num_tokens": 20350074.0,
      "step": 9285
    },
    {
      "entropy": 5.518982172012329,
      "epoch": 0.9219928543072647,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004923763520113406,
      "loss": 5.2761,
      "mean_token_accuracy": 0.159541492164135,
      "num_tokens": 20361058.0,
      "step": 9290
    },
    {
      "entropy": 5.448275804519653,
      "epoch": 0.9224890829694323,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004923672043028395,
      "loss": 5.2258,
      "mean_token_accuracy": 0.1650228515267372,
      "num_tokens": 20372605.0,
      "step": 9295
    },
    {
      "entropy": 5.597142934799194,
      "epoch": 0.9229853116315998,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004923580512041001,
      "loss": 5.353,
      "mean_token_accuracy": 0.1553454339504242,
      "num_tokens": 20383512.0,
      "step": 9300
    },
    {
      "entropy": 5.617527818679809,
      "epoch": 0.9234815402937674,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004923488927153495,
      "loss": 5.3276,
      "mean_token_accuracy": 0.1534099519252777,
      "num_tokens": 20395862.0,
      "step": 9305
    },
    {
      "entropy": 5.463476848602295,
      "epoch": 0.9239777689559349,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923397288368146,
      "loss": 5.1582,
      "mean_token_accuracy": 0.17000781893730163,
      "num_tokens": 20407349.0,
      "step": 9310
    },
    {
      "entropy": 5.551022577285766,
      "epoch": 0.9244739976181024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004923305595687229,
      "loss": 5.3761,
      "mean_token_accuracy": 0.16101160645484924,
      "num_tokens": 20419468.0,
      "step": 9315
    },
    {
      "entropy": 5.529065895080566,
      "epoch": 0.9249702262802699,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923213849113015,
      "loss": 5.1895,
      "mean_token_accuracy": 0.16460828930139543,
      "num_tokens": 20429689.0,
      "step": 9320
    },
    {
      "entropy": 5.525329160690307,
      "epoch": 0.9254664549424375,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004923122048647781,
      "loss": 5.2871,
      "mean_token_accuracy": 0.15861476212739944,
      "num_tokens": 20440232.0,
      "step": 9325
    },
    {
      "entropy": 5.50602126121521,
      "epoch": 0.925962683604605,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923030194293802,
      "loss": 5.2811,
      "mean_token_accuracy": 0.15639358162879943,
      "num_tokens": 20451021.0,
      "step": 9330
    },
    {
      "entropy": 5.556261205673218,
      "epoch": 0.9264589122667726,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004922938286053358,
      "loss": 5.2426,
      "mean_token_accuracy": 0.16136108040809632,
      "num_tokens": 20461301.0,
      "step": 9335
    },
    {
      "entropy": 5.53191032409668,
      "epoch": 0.92695514092894,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922846323928726,
      "loss": 5.3043,
      "mean_token_accuracy": 0.16135042309761047,
      "num_tokens": 20472232.0,
      "step": 9340
    },
    {
      "entropy": 5.560789585113525,
      "epoch": 0.9274513695911076,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004922754307922188,
      "loss": 5.3666,
      "mean_token_accuracy": 0.15169524848461152,
      "num_tokens": 20483063.0,
      "step": 9345
    },
    {
      "entropy": 5.599881267547607,
      "epoch": 0.9279475982532751,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922662238036024,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1592492312192917,
      "num_tokens": 20493631.0,
      "step": 9350
    },
    {
      "entropy": 5.5158891677856445,
      "epoch": 0.9284438269154426,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922570114272518,
      "loss": 5.2887,
      "mean_token_accuracy": 0.1648649290204048,
      "num_tokens": 20505173.0,
      "step": 9355
    },
    {
      "entropy": 5.538170242309571,
      "epoch": 0.9289400555776102,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922477936633955,
      "loss": 5.2435,
      "mean_token_accuracy": 0.16174097061157228,
      "num_tokens": 20516354.0,
      "step": 9360
    },
    {
      "entropy": 5.467098379135132,
      "epoch": 0.9294362842397776,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004922385705122621,
      "loss": 5.1589,
      "mean_token_accuracy": 0.17511531710624695,
      "num_tokens": 20528531.0,
      "step": 9365
    },
    {
      "entropy": 5.513939094543457,
      "epoch": 0.9299325129019452,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922293419740802,
      "loss": 5.2347,
      "mean_token_accuracy": 0.17164021134376525,
      "num_tokens": 20539957.0,
      "step": 9370
    },
    {
      "entropy": 5.530021286010742,
      "epoch": 0.9304287415641127,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922201080490787,
      "loss": 5.2576,
      "mean_token_accuracy": 0.1641483798623085,
      "num_tokens": 20550656.0,
      "step": 9375
    },
    {
      "entropy": 5.534614562988281,
      "epoch": 0.9309249702262803,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922108687374867,
      "loss": 5.2675,
      "mean_token_accuracy": 0.16341327726840973,
      "num_tokens": 20561025.0,
      "step": 9380
    },
    {
      "entropy": 5.518497133255005,
      "epoch": 0.9314211988884478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492201624039533,
      "loss": 5.197,
      "mean_token_accuracy": 0.16047702133655548,
      "num_tokens": 20570839.0,
      "step": 9385
    },
    {
      "entropy": 5.493470621109009,
      "epoch": 0.9319174275506154,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004921923739554472,
      "loss": 5.2294,
      "mean_token_accuracy": 0.1624532476067543,
      "num_tokens": 20581748.0,
      "step": 9390
    },
    {
      "entropy": 5.514760541915893,
      "epoch": 0.9324136562127828,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921831184854585,
      "loss": 5.2201,
      "mean_token_accuracy": 0.16752604097127916,
      "num_tokens": 20592639.0,
      "step": 9395
    },
    {
      "entropy": 5.5155205726623535,
      "epoch": 0.9329098848749504,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004921738576297965,
      "loss": 5.2423,
      "mean_token_accuracy": 0.16669758409261703,
      "num_tokens": 20603660.0,
      "step": 9400
    },
    {
      "entropy": 5.544377326965332,
      "epoch": 0.9334061135371179,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921645913886907,
      "loss": 5.3372,
      "mean_token_accuracy": 0.1593969851732254,
      "num_tokens": 20614541.0,
      "step": 9405
    },
    {
      "entropy": 5.427964210510254,
      "epoch": 0.9339023421992855,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004921553197623711,
      "loss": 5.2119,
      "mean_token_accuracy": 0.17062114775180817,
      "num_tokens": 20625274.0,
      "step": 9410
    },
    {
      "entropy": 5.562678623199463,
      "epoch": 0.934398570861453,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004921460427510674,
      "loss": 5.3105,
      "mean_token_accuracy": 0.16440874934196473,
      "num_tokens": 20636282.0,
      "step": 9415
    },
    {
      "entropy": 5.559782266616821,
      "epoch": 0.9348947995236205,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004921367603550098,
      "loss": 5.2259,
      "mean_token_accuracy": 0.16438380926847457,
      "num_tokens": 20647651.0,
      "step": 9420
    },
    {
      "entropy": 5.53304500579834,
      "epoch": 0.935391028185788,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004921274725744283,
      "loss": 5.2599,
      "mean_token_accuracy": 0.16550061404705046,
      "num_tokens": 20657587.0,
      "step": 9425
    },
    {
      "entropy": 5.521426391601563,
      "epoch": 0.9358872568479555,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004921181794095536,
      "loss": 5.3263,
      "mean_token_accuracy": 0.16232970505952835,
      "num_tokens": 20668672.0,
      "step": 9430
    },
    {
      "entropy": 5.636467361450196,
      "epoch": 0.9363834855101231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921088808606157,
      "loss": 5.3453,
      "mean_token_accuracy": 0.15954353362321855,
      "num_tokens": 20680020.0,
      "step": 9435
    },
    {
      "entropy": 5.528496408462525,
      "epoch": 0.9368797141722905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004920995769278454,
      "loss": 5.3063,
      "mean_token_accuracy": 0.15940629243850707,
      "num_tokens": 20690693.0,
      "step": 9440
    },
    {
      "entropy": 5.498102521896362,
      "epoch": 0.9373759428344581,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920902676114734,
      "loss": 5.2638,
      "mean_token_accuracy": 0.16349878460168837,
      "num_tokens": 20701943.0,
      "step": 9445
    },
    {
      "entropy": 5.5208837509155275,
      "epoch": 0.9378721714966256,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004920809529117306,
      "loss": 5.2491,
      "mean_token_accuracy": 0.16090769767761232,
      "num_tokens": 20711680.0,
      "step": 9450
    },
    {
      "entropy": 5.611314821243286,
      "epoch": 0.9383684001587932,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004920716328288479,
      "loss": 5.3524,
      "mean_token_accuracy": 0.15770208686590195,
      "num_tokens": 20722273.0,
      "step": 9455
    },
    {
      "entropy": 5.522660112380981,
      "epoch": 0.9388646288209607,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004920623073630566,
      "loss": 5.1768,
      "mean_token_accuracy": 0.17186968475580217,
      "num_tokens": 20732914.0,
      "step": 9460
    },
    {
      "entropy": 5.510355806350708,
      "epoch": 0.9393608574831283,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920529765145876,
      "loss": 5.2919,
      "mean_token_accuracy": 0.15436756908893584,
      "num_tokens": 20743485.0,
      "step": 9465
    },
    {
      "entropy": 5.58152289390564,
      "epoch": 0.9398570861452957,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920436402836726,
      "loss": 5.3295,
      "mean_token_accuracy": 0.16058432906866074,
      "num_tokens": 20754130.0,
      "step": 9470
    },
    {
      "entropy": 5.5492674827575685,
      "epoch": 0.9403533148074633,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000492034298670543,
      "loss": 5.2979,
      "mean_token_accuracy": 0.16547994315624237,
      "num_tokens": 20766181.0,
      "step": 9475
    },
    {
      "entropy": 5.540728569030762,
      "epoch": 0.9408495434696308,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920249516754304,
      "loss": 5.2294,
      "mean_token_accuracy": 0.16097873449325562,
      "num_tokens": 20777321.0,
      "step": 9480
    },
    {
      "entropy": 5.518664026260376,
      "epoch": 0.9413457721317984,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920155992985667,
      "loss": 5.1977,
      "mean_token_accuracy": 0.16674454957246782,
      "num_tokens": 20788133.0,
      "step": 9485
    },
    {
      "entropy": 5.4627128601074215,
      "epoch": 0.9418420007939659,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920062415401837,
      "loss": 5.2353,
      "mean_token_accuracy": 0.15938979387283325,
      "num_tokens": 20798019.0,
      "step": 9490
    },
    {
      "entropy": 5.545669221878052,
      "epoch": 0.9423382294561334,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004919968784005136,
      "loss": 5.2421,
      "mean_token_accuracy": 0.15849845856428146,
      "num_tokens": 20808449.0,
      "step": 9495
    },
    {
      "entropy": 5.4911627769470215,
      "epoch": 0.9428344581183009,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919875098797886,
      "loss": 5.207,
      "mean_token_accuracy": 0.16638871431350707,
      "num_tokens": 20819365.0,
      "step": 9500
    },
    {
      "entropy": 5.495926570892334,
      "epoch": 0.9433306867804684,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004919781359782407,
      "loss": 5.2465,
      "mean_token_accuracy": 0.16293539702892304,
      "num_tokens": 20829988.0,
      "step": 9505
    },
    {
      "entropy": 5.466719293594361,
      "epoch": 0.943826915442636,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004919687566961027,
      "loss": 5.2598,
      "mean_token_accuracy": 0.160459241271019,
      "num_tokens": 20841047.0,
      "step": 9510
    },
    {
      "entropy": 5.496150875091553,
      "epoch": 0.9443231441048034,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491959372033607,
      "loss": 5.2535,
      "mean_token_accuracy": 0.17023346573114395,
      "num_tokens": 20852652.0,
      "step": 9515
    },
    {
      "entropy": 5.6562072277069095,
      "epoch": 0.944819372766971,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004919499819909865,
      "loss": 5.3543,
      "mean_token_accuracy": 0.16004971116781236,
      "num_tokens": 20863558.0,
      "step": 9520
    },
    {
      "entropy": 5.58142580986023,
      "epoch": 0.9453156014291385,
      "grad_norm": 0.875,
      "learning_rate": 0.0004919405865684738,
      "loss": 5.3055,
      "mean_token_accuracy": 0.15883271992206574,
      "num_tokens": 20873906.0,
      "step": 9525
    },
    {
      "entropy": 5.549357748031616,
      "epoch": 0.9458118300913061,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491931185766302,
      "loss": 5.3699,
      "mean_token_accuracy": 0.163249172270298,
      "num_tokens": 20885205.0,
      "step": 9530
    },
    {
      "entropy": 5.589345264434814,
      "epoch": 0.9463080587534736,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919217795847044,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1664511114358902,
      "num_tokens": 20895172.0,
      "step": 9535
    },
    {
      "entropy": 5.575771999359131,
      "epoch": 0.9468042874156412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919123680239139,
      "loss": 5.4533,
      "mean_token_accuracy": 0.15265266895294188,
      "num_tokens": 20907454.0,
      "step": 9540
    },
    {
      "entropy": 5.570123147964478,
      "epoch": 0.9473005160778086,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919029510841641,
      "loss": 5.3317,
      "mean_token_accuracy": 0.1641636848449707,
      "num_tokens": 20919080.0,
      "step": 9545
    },
    {
      "entropy": 5.6188253402709964,
      "epoch": 0.9477967447399762,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918935287656887,
      "loss": 5.2659,
      "mean_token_accuracy": 0.16217341125011445,
      "num_tokens": 20930807.0,
      "step": 9550
    },
    {
      "entropy": 5.550796365737915,
      "epoch": 0.9482929734021437,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491884101068721,
      "loss": 5.2409,
      "mean_token_accuracy": 0.1619820326566696,
      "num_tokens": 20941255.0,
      "step": 9555
    },
    {
      "entropy": 5.466588640213013,
      "epoch": 0.9487892020643113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004918746679934949,
      "loss": 5.243,
      "mean_token_accuracy": 0.16703216433525087,
      "num_tokens": 20951008.0,
      "step": 9560
    },
    {
      "entropy": 5.582643318176269,
      "epoch": 0.9492854307264788,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004918652295402444,
      "loss": 5.3841,
      "mean_token_accuracy": 0.15777651965618134,
      "num_tokens": 20963296.0,
      "step": 9565
    },
    {
      "entropy": 5.653816843032837,
      "epoch": 0.9497816593886463,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004918557857092036,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1512746311724186,
      "num_tokens": 20974995.0,
      "step": 9570
    },
    {
      "entropy": 5.614622402191162,
      "epoch": 0.9502778880508138,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918463365006065,
      "loss": 5.3353,
      "mean_token_accuracy": 0.15771914422512054,
      "num_tokens": 20986297.0,
      "step": 9575
    },
    {
      "entropy": 5.522190856933594,
      "epoch": 0.9507741167129813,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918368819146875,
      "loss": 5.2799,
      "mean_token_accuracy": 0.1641617864370346,
      "num_tokens": 20997871.0,
      "step": 9580
    },
    {
      "entropy": 5.5050122261047365,
      "epoch": 0.9512703453751489,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004918274219516811,
      "loss": 5.281,
      "mean_token_accuracy": 0.16019633561372756,
      "num_tokens": 21010295.0,
      "step": 9585
    },
    {
      "entropy": 5.481592893600464,
      "epoch": 0.9517665740373163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918179566118218,
      "loss": 5.2291,
      "mean_token_accuracy": 0.16115749925374984,
      "num_tokens": 21021307.0,
      "step": 9590
    },
    {
      "entropy": 5.5202374935150145,
      "epoch": 0.9522628026994839,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004918084858953445,
      "loss": 5.2217,
      "mean_token_accuracy": 0.16853663176298142,
      "num_tokens": 21032325.0,
      "step": 9595
    },
    {
      "entropy": 5.574051189422607,
      "epoch": 0.9527590313616514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004917990098024839,
      "loss": 5.2639,
      "mean_token_accuracy": 0.16447817534208298,
      "num_tokens": 21043881.0,
      "step": 9600
    },
    {
      "entropy": 5.490035724639893,
      "epoch": 0.953255260023819,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917895283334749,
      "loss": 5.2395,
      "mean_token_accuracy": 0.1668274939060211,
      "num_tokens": 21054042.0,
      "step": 9605
    },
    {
      "entropy": 5.5206738948822025,
      "epoch": 0.9537514886859865,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491780041488553,
      "loss": 5.3325,
      "mean_token_accuracy": 0.16412782520055771,
      "num_tokens": 21064374.0,
      "step": 9610
    },
    {
      "entropy": 5.565544986724854,
      "epoch": 0.9542477173481541,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917705492679529,
      "loss": 5.3144,
      "mean_token_accuracy": 0.1583273231983185,
      "num_tokens": 21075525.0,
      "step": 9615
    },
    {
      "entropy": 5.567374610900879,
      "epoch": 0.9547439460103215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917610516719104,
      "loss": 5.2252,
      "mean_token_accuracy": 0.16568062603473663,
      "num_tokens": 21086259.0,
      "step": 9620
    },
    {
      "entropy": 5.5197385311126705,
      "epoch": 0.9552401746724891,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004917515487006609,
      "loss": 5.1877,
      "mean_token_accuracy": 0.17206861972808837,
      "num_tokens": 21096987.0,
      "step": 9625
    },
    {
      "entropy": 5.6087610721588135,
      "epoch": 0.9557364033346566,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049174204035444,
      "loss": 5.3341,
      "mean_token_accuracy": 0.15802735835313797,
      "num_tokens": 21108220.0,
      "step": 9630
    },
    {
      "entropy": 5.56252818107605,
      "epoch": 0.9562326319968242,
      "grad_norm": 0.875,
      "learning_rate": 0.0004917325266334836,
      "loss": 5.2995,
      "mean_token_accuracy": 0.15769239962100984,
      "num_tokens": 21118679.0,
      "step": 9635
    },
    {
      "entropy": 5.565669393539428,
      "epoch": 0.9567288606589917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004917230075380276,
      "loss": 5.2839,
      "mean_token_accuracy": 0.15459821969270707,
      "num_tokens": 21129945.0,
      "step": 9640
    },
    {
      "entropy": 5.561633920669555,
      "epoch": 0.9572250893211592,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917134830683079,
      "loss": 5.3328,
      "mean_token_accuracy": 0.16172449439764022,
      "num_tokens": 21140577.0,
      "step": 9645
    },
    {
      "entropy": 5.533219575881958,
      "epoch": 0.9577213179833267,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917039532245608,
      "loss": 5.2057,
      "mean_token_accuracy": 0.16022719591856002,
      "num_tokens": 21151041.0,
      "step": 9650
    },
    {
      "entropy": 5.500655746459961,
      "epoch": 0.9582175466454942,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004916944180070226,
      "loss": 5.2927,
      "mean_token_accuracy": 0.16059710830450058,
      "num_tokens": 21161401.0,
      "step": 9655
    },
    {
      "entropy": 5.547111129760742,
      "epoch": 0.9587137753076618,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004916848774159299,
      "loss": 5.3315,
      "mean_token_accuracy": 0.1570285901427269,
      "num_tokens": 21170897.0,
      "step": 9660
    },
    {
      "entropy": 5.60529842376709,
      "epoch": 0.9592100039698293,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491675331451519,
      "loss": 5.3121,
      "mean_token_accuracy": 0.1545352578163147,
      "num_tokens": 21183313.0,
      "step": 9665
    },
    {
      "entropy": 5.588545513153076,
      "epoch": 0.9597062326319968,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004916657801140269,
      "loss": 5.2359,
      "mean_token_accuracy": 0.15909406989812852,
      "num_tokens": 21192820.0,
      "step": 9670
    },
    {
      "entropy": 5.522636938095093,
      "epoch": 0.9602024612941643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004916562234036903,
      "loss": 5.2107,
      "mean_token_accuracy": 0.1702083468437195,
      "num_tokens": 21203493.0,
      "step": 9675
    },
    {
      "entropy": 5.5380126953125,
      "epoch": 0.9606986899563319,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916466613207462,
      "loss": 5.2426,
      "mean_token_accuracy": 0.15627574622631074,
      "num_tokens": 21214345.0,
      "step": 9680
    },
    {
      "entropy": 5.5811567306518555,
      "epoch": 0.9611949186184994,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004916370938654317,
      "loss": 5.2638,
      "mean_token_accuracy": 0.16205609887838363,
      "num_tokens": 21224345.0,
      "step": 9685
    },
    {
      "entropy": 5.5387749671936035,
      "epoch": 0.961691147280667,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004916275210379841,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1599640429019928,
      "num_tokens": 21235938.0,
      "step": 9690
    },
    {
      "entropy": 5.599554777145386,
      "epoch": 0.9621873759428344,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916179428386408,
      "loss": 5.3546,
      "mean_token_accuracy": 0.16278104186058046,
      "num_tokens": 21247670.0,
      "step": 9695
    },
    {
      "entropy": 5.619308519363403,
      "epoch": 0.962683604605002,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916083592676392,
      "loss": 5.2433,
      "mean_token_accuracy": 0.16622216403484344,
      "num_tokens": 21257982.0,
      "step": 9700
    },
    {
      "entropy": 5.480779981613159,
      "epoch": 0.9631798332671695,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004915987703252171,
      "loss": 5.1972,
      "mean_token_accuracy": 0.16707677245140076,
      "num_tokens": 21268044.0,
      "step": 9705
    },
    {
      "entropy": 5.540554666519165,
      "epoch": 0.9636760619293371,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004915891760116123,
      "loss": 5.2603,
      "mean_token_accuracy": 0.16277814656496048,
      "num_tokens": 21278331.0,
      "step": 9710
    },
    {
      "entropy": 5.560805082321167,
      "epoch": 0.9641722905915046,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004915795763270626,
      "loss": 5.3066,
      "mean_token_accuracy": 0.15798047333955764,
      "num_tokens": 21287225.0,
      "step": 9715
    },
    {
      "entropy": 5.593976593017578,
      "epoch": 0.9646685192536721,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000491569971271806,
      "loss": 5.3705,
      "mean_token_accuracy": 0.1539733186364174,
      "num_tokens": 21298607.0,
      "step": 9720
    },
    {
      "entropy": 5.5799847602844235,
      "epoch": 0.9651647479158396,
      "grad_norm": 0.875,
      "learning_rate": 0.0004915603608460809,
      "loss": 5.29,
      "mean_token_accuracy": 0.15675615072250365,
      "num_tokens": 21311308.0,
      "step": 9725
    },
    {
      "entropy": 5.578527307510376,
      "epoch": 0.9656609765780071,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004915507450501254,
      "loss": 5.3548,
      "mean_token_accuracy": 0.156254905462265,
      "num_tokens": 21322370.0,
      "step": 9730
    },
    {
      "entropy": 5.541017627716064,
      "epoch": 0.9661572052401747,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000491541123884178,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16704410165548325,
      "num_tokens": 21333488.0,
      "step": 9735
    },
    {
      "entropy": 5.570231914520264,
      "epoch": 0.9666534339023422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915314973484775,
      "loss": 5.3935,
      "mean_token_accuracy": 0.15849739164114,
      "num_tokens": 21344907.0,
      "step": 9740
    },
    {
      "entropy": 5.564306116104126,
      "epoch": 0.9671496625645097,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004915218654432624,
      "loss": 5.3767,
      "mean_token_accuracy": 0.15869806408882142,
      "num_tokens": 21355983.0,
      "step": 9745
    },
    {
      "entropy": 5.650417137145996,
      "epoch": 0.9676458912266772,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004915122281687716,
      "loss": 5.3289,
      "mean_token_accuracy": 0.16182188093662261,
      "num_tokens": 21366313.0,
      "step": 9750
    },
    {
      "entropy": 5.638453149795533,
      "epoch": 0.9681421198888448,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004915025855252441,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1546548068523407,
      "num_tokens": 21377239.0,
      "step": 9755
    },
    {
      "entropy": 5.551052761077881,
      "epoch": 0.9686383485510123,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004914929375129188,
      "loss": 5.3521,
      "mean_token_accuracy": 0.1624363034963608,
      "num_tokens": 21387622.0,
      "step": 9760
    },
    {
      "entropy": 5.512719583511353,
      "epoch": 0.9691345772131799,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004914832841320353,
      "loss": 5.2123,
      "mean_token_accuracy": 0.16549569964408875,
      "num_tokens": 21397298.0,
      "step": 9765
    },
    {
      "entropy": 5.513585424423217,
      "epoch": 0.9696308058753473,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491473625382833,
      "loss": 5.2182,
      "mean_token_accuracy": 0.15830965489149093,
      "num_tokens": 21408444.0,
      "step": 9770
    },
    {
      "entropy": 5.6223774433135985,
      "epoch": 0.9701270345375149,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000491463961265551,
      "loss": 5.3745,
      "mean_token_accuracy": 0.16331460773944856,
      "num_tokens": 21420313.0,
      "step": 9775
    },
    {
      "entropy": 5.558877754211426,
      "epoch": 0.9706232631996824,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004914542917804292,
      "loss": 5.2845,
      "mean_token_accuracy": 0.1638398140668869,
      "num_tokens": 21432068.0,
      "step": 9780
    },
    {
      "entropy": 5.614261150360107,
      "epoch": 0.97111949186185,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004914446169277075,
      "loss": 5.3663,
      "mean_token_accuracy": 0.15839477479457856,
      "num_tokens": 21442824.0,
      "step": 9785
    },
    {
      "entropy": 5.6283040046691895,
      "epoch": 0.9716157205240175,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004914349367076258,
      "loss": 5.3539,
      "mean_token_accuracy": 0.15483997762203217,
      "num_tokens": 21452991.0,
      "step": 9790
    },
    {
      "entropy": 5.5307928085327145,
      "epoch": 0.972111949186185,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004914252511204238,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1677130788564682,
      "num_tokens": 21465610.0,
      "step": 9795
    },
    {
      "entropy": 5.483946228027344,
      "epoch": 0.9726081778483525,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004914155601663422,
      "loss": 5.1966,
      "mean_token_accuracy": 0.17021872848272324,
      "num_tokens": 21475951.0,
      "step": 9800
    },
    {
      "entropy": 5.546132755279541,
      "epoch": 0.97310440651052,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914058638456208,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1595159277319908,
      "num_tokens": 21486101.0,
      "step": 9805
    },
    {
      "entropy": 5.556393766403199,
      "epoch": 0.9736006351726876,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004913961621585003,
      "loss": 5.2434,
      "mean_token_accuracy": 0.16645517349243164,
      "num_tokens": 21497494.0,
      "step": 9810
    },
    {
      "entropy": 5.457012462615967,
      "epoch": 0.974096863834855,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913864551052214,
      "loss": 5.2069,
      "mean_token_accuracy": 0.16951558738946915,
      "num_tokens": 21507840.0,
      "step": 9815
    },
    {
      "entropy": 5.52414026260376,
      "epoch": 0.9745930924970226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004913767426860246,
      "loss": 5.2632,
      "mean_token_accuracy": 0.16290182173252105,
      "num_tokens": 21518099.0,
      "step": 9820
    },
    {
      "entropy": 5.520859718322754,
      "epoch": 0.9750893211591901,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913670249011508,
      "loss": 5.2538,
      "mean_token_accuracy": 0.1606919065117836,
      "num_tokens": 21529465.0,
      "step": 9825
    },
    {
      "entropy": 5.492925691604614,
      "epoch": 0.9755855498213577,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913573017508411,
      "loss": 5.3015,
      "mean_token_accuracy": 0.15913283079862595,
      "num_tokens": 21540206.0,
      "step": 9830
    },
    {
      "entropy": 5.591370916366577,
      "epoch": 0.9760817784835252,
      "grad_norm": 0.875,
      "learning_rate": 0.0004913475732353365,
      "loss": 5.2881,
      "mean_token_accuracy": 0.16092725545167924,
      "num_tokens": 21551055.0,
      "step": 9835
    },
    {
      "entropy": 5.504221820831299,
      "epoch": 0.9765780071456928,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004913378393548784,
      "loss": 5.22,
      "mean_token_accuracy": 0.16311650723218918,
      "num_tokens": 21562001.0,
      "step": 9840
    },
    {
      "entropy": 5.507087993621826,
      "epoch": 0.9770742358078602,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913281001097078,
      "loss": 5.2617,
      "mean_token_accuracy": 0.15980887562036514,
      "num_tokens": 21571781.0,
      "step": 9845
    },
    {
      "entropy": 5.537432622909546,
      "epoch": 0.9775704644700278,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913183555000667,
      "loss": 5.2465,
      "mean_token_accuracy": 0.16694073081016542,
      "num_tokens": 21581300.0,
      "step": 9850
    },
    {
      "entropy": 5.5360527515411375,
      "epoch": 0.9780666931321953,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913086055261963,
      "loss": 5.3552,
      "mean_token_accuracy": 0.1563612550497055,
      "num_tokens": 21593540.0,
      "step": 9855
    },
    {
      "entropy": 5.5462486267089846,
      "epoch": 0.9785629217943629,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004912988501883387,
      "loss": 5.211,
      "mean_token_accuracy": 0.16728620529174804,
      "num_tokens": 21603395.0,
      "step": 9860
    },
    {
      "entropy": 5.6481705665588375,
      "epoch": 0.9790591504565304,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004912890894867357,
      "loss": 5.4157,
      "mean_token_accuracy": 0.15332029461860658,
      "num_tokens": 21615508.0,
      "step": 9865
    },
    {
      "entropy": 5.450258207321167,
      "epoch": 0.9795553791186979,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004912793234216293,
      "loss": 5.1957,
      "mean_token_accuracy": 0.1716061219573021,
      "num_tokens": 21625832.0,
      "step": 9870
    },
    {
      "entropy": 5.509284734725952,
      "epoch": 0.9800516077808654,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004912695519932618,
      "loss": 5.2667,
      "mean_token_accuracy": 0.16307174265384675,
      "num_tokens": 21637256.0,
      "step": 9875
    },
    {
      "entropy": 5.5890600204467775,
      "epoch": 0.9805478364430329,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004912597752018753,
      "loss": 5.3211,
      "mean_token_accuracy": 0.15741544365882873,
      "num_tokens": 21647944.0,
      "step": 9880
    },
    {
      "entropy": 5.52183952331543,
      "epoch": 0.9810440651052005,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004912499930477126,
      "loss": 5.1851,
      "mean_token_accuracy": 0.16473323553800584,
      "num_tokens": 21658572.0,
      "step": 9885
    },
    {
      "entropy": 5.5505866527557375,
      "epoch": 0.981540293767368,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004912402055310158,
      "loss": 5.2752,
      "mean_token_accuracy": 0.16552853882312774,
      "num_tokens": 21668782.0,
      "step": 9890
    },
    {
      "entropy": 5.55018720626831,
      "epoch": 0.9820365224295355,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004912304126520279,
      "loss": 5.3107,
      "mean_token_accuracy": 0.15749423652887345,
      "num_tokens": 21679820.0,
      "step": 9895
    },
    {
      "entropy": 5.5020262718200685,
      "epoch": 0.982532751091703,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004912206144109917,
      "loss": 5.2817,
      "mean_token_accuracy": 0.16745246350765228,
      "num_tokens": 21690912.0,
      "step": 9900
    },
    {
      "entropy": 5.483227252960205,
      "epoch": 0.9830289797538706,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004912108108081503,
      "loss": 5.2019,
      "mean_token_accuracy": 0.1591826781630516,
      "num_tokens": 21701653.0,
      "step": 9905
    },
    {
      "entropy": 5.537767744064331,
      "epoch": 0.9835252084160381,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912010018437465,
      "loss": 5.3328,
      "mean_token_accuracy": 0.16012009382247924,
      "num_tokens": 21713841.0,
      "step": 9910
    },
    {
      "entropy": 5.581033563613891,
      "epoch": 0.9840214370782057,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491191187518024,
      "loss": 5.2873,
      "mean_token_accuracy": 0.16304097324609756,
      "num_tokens": 21724343.0,
      "step": 9915
    },
    {
      "entropy": 5.5859270095825195,
      "epoch": 0.9845176657403731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004911813678312257,
      "loss": 5.2935,
      "mean_token_accuracy": 0.15534278601408005,
      "num_tokens": 21734424.0,
      "step": 9920
    },
    {
      "entropy": 5.480744981765747,
      "epoch": 0.9850138944025407,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911715427835953,
      "loss": 5.2612,
      "mean_token_accuracy": 0.16374149173498154,
      "num_tokens": 21745001.0,
      "step": 9925
    },
    {
      "entropy": 5.522067642211914,
      "epoch": 0.9855101230647082,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004911617123753765,
      "loss": 5.242,
      "mean_token_accuracy": 0.16677838563919067,
      "num_tokens": 21755838.0,
      "step": 9930
    },
    {
      "entropy": 5.531601095199585,
      "epoch": 0.9860063517268758,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004911518766068131,
      "loss": 5.2169,
      "mean_token_accuracy": 0.16688804626464843,
      "num_tokens": 21768057.0,
      "step": 9935
    },
    {
      "entropy": 5.557320880889892,
      "epoch": 0.9865025803890433,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911420354781487,
      "loss": 5.3268,
      "mean_token_accuracy": 0.15836716294288636,
      "num_tokens": 21779136.0,
      "step": 9940
    },
    {
      "entropy": 5.54874587059021,
      "epoch": 0.9869988090512108,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004911321889896278,
      "loss": 5.29,
      "mean_token_accuracy": 0.15742596983909607,
      "num_tokens": 21790556.0,
      "step": 9945
    },
    {
      "entropy": 5.595358276367188,
      "epoch": 0.9874950377133783,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004911223371414943,
      "loss": 5.2919,
      "mean_token_accuracy": 0.16788019984960556,
      "num_tokens": 21801931.0,
      "step": 9950
    },
    {
      "entropy": 5.505606031417846,
      "epoch": 0.9879912663755459,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004911124799339924,
      "loss": 5.3287,
      "mean_token_accuracy": 0.1579960435628891,
      "num_tokens": 21813923.0,
      "step": 9955
    },
    {
      "entropy": 5.49721097946167,
      "epoch": 0.9884874950377134,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911026173673667,
      "loss": 5.1755,
      "mean_token_accuracy": 0.17224640846252443,
      "num_tokens": 21825230.0,
      "step": 9960
    },
    {
      "entropy": 5.570822525024414,
      "epoch": 0.9889837236998809,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004910927494418619,
      "loss": 5.298,
      "mean_token_accuracy": 0.1619612157344818,
      "num_tokens": 21835047.0,
      "step": 9965
    },
    {
      "entropy": 5.536602783203125,
      "epoch": 0.9894799523620484,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004910828761577224,
      "loss": 5.2506,
      "mean_token_accuracy": 0.15741328597068788,
      "num_tokens": 21845063.0,
      "step": 9970
    },
    {
      "entropy": 5.531167364120483,
      "epoch": 0.9899761810242159,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004910729975151933,
      "loss": 5.2115,
      "mean_token_accuracy": 0.16591019183397293,
      "num_tokens": 21856283.0,
      "step": 9975
    },
    {
      "entropy": 5.54023585319519,
      "epoch": 0.9904724096863835,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004910631135145194,
      "loss": 5.2577,
      "mean_token_accuracy": 0.1648746907711029,
      "num_tokens": 21867159.0,
      "step": 9980
    },
    {
      "entropy": 5.554834365844727,
      "epoch": 0.990968638348551,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910532241559458,
      "loss": 5.2458,
      "mean_token_accuracy": 0.1617730513215065,
      "num_tokens": 21877373.0,
      "step": 9985
    },
    {
      "entropy": 5.529282522201538,
      "epoch": 0.9914648670107186,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910433294397179,
      "loss": 5.2819,
      "mean_token_accuracy": 0.1645874187350273,
      "num_tokens": 21889855.0,
      "step": 9990
    },
    {
      "entropy": 5.450818967819214,
      "epoch": 0.991961095672886,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000491033429366081,
      "loss": 5.1959,
      "mean_token_accuracy": 0.16901558637619019,
      "num_tokens": 21899522.0,
      "step": 9995
    },
    {
      "entropy": 5.467387390136719,
      "epoch": 0.9924573243350536,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004910235239352804,
      "loss": 5.2247,
      "mean_token_accuracy": 0.17409602403640748,
      "num_tokens": 21910578.0,
      "step": 10000
    },
    {
      "entropy": 5.511497879028321,
      "epoch": 0.9929535529972211,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491013613147562,
      "loss": 5.2572,
      "mean_token_accuracy": 0.1583801418542862,
      "num_tokens": 21921048.0,
      "step": 10005
    },
    {
      "entropy": 5.481568384170532,
      "epoch": 0.9934497816593887,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910036970031715,
      "loss": 5.2511,
      "mean_token_accuracy": 0.1595163196325302,
      "num_tokens": 21930950.0,
      "step": 10010
    },
    {
      "entropy": 5.672370386123657,
      "epoch": 0.9939460103215562,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004909937755023547,
      "loss": 5.3918,
      "mean_token_accuracy": 0.15413139760494232,
      "num_tokens": 21941333.0,
      "step": 10015
    },
    {
      "entropy": 5.53797721862793,
      "epoch": 0.9944422389837237,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004909838486453577,
      "loss": 5.1866,
      "mean_token_accuracy": 0.16462625861167907,
      "num_tokens": 21951895.0,
      "step": 10020
    },
    {
      "entropy": 5.553476715087891,
      "epoch": 0.9949384676458912,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004909739164324267,
      "loss": 5.2973,
      "mean_token_accuracy": 0.15845824182033538,
      "num_tokens": 21963915.0,
      "step": 10025
    },
    {
      "entropy": 5.6089620113372805,
      "epoch": 0.9954346963080588,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004909639788638079,
      "loss": 5.3275,
      "mean_token_accuracy": 0.1611268237233162,
      "num_tokens": 21973653.0,
      "step": 10030
    },
    {
      "entropy": 5.603273963928222,
      "epoch": 0.9959309249702263,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004909540359397477,
      "loss": 5.2633,
      "mean_token_accuracy": 0.1604244366288185,
      "num_tokens": 21985042.0,
      "step": 10035
    },
    {
      "entropy": 5.543943643569946,
      "epoch": 0.9964271536323938,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909440876604929,
      "loss": 5.3212,
      "mean_token_accuracy": 0.1594669207930565,
      "num_tokens": 21996276.0,
      "step": 10040
    },
    {
      "entropy": 5.520408582687378,
      "epoch": 0.9969233822945613,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049093413402629,
      "loss": 5.2865,
      "mean_token_accuracy": 0.16114149391651153,
      "num_tokens": 22006774.0,
      "step": 10045
    },
    {
      "entropy": 5.50577712059021,
      "epoch": 0.9974196109567288,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004909241750373858,
      "loss": 5.2991,
      "mean_token_accuracy": 0.1550804540514946,
      "num_tokens": 22017708.0,
      "step": 10050
    },
    {
      "entropy": 5.518759059906006,
      "epoch": 0.9979158396188964,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004909142106940275,
      "loss": 5.2192,
      "mean_token_accuracy": 0.16255709528923035,
      "num_tokens": 22030393.0,
      "step": 10055
    },
    {
      "entropy": 5.575761222839356,
      "epoch": 0.9984120682810639,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000490904240996462,
      "loss": 5.2808,
      "mean_token_accuracy": 0.16375949531793593,
      "num_tokens": 22041124.0,
      "step": 10060
    },
    {
      "entropy": 5.569854354858398,
      "epoch": 0.9989082969432315,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004908942659449364,
      "loss": 5.2752,
      "mean_token_accuracy": 0.16698591262102128,
      "num_tokens": 22052512.0,
      "step": 10065
    },
    {
      "entropy": 5.524609375,
      "epoch": 0.9994045256053989,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004908842855396982,
      "loss": 5.2677,
      "mean_token_accuracy": 0.16285080164670945,
      "num_tokens": 22063689.0,
      "step": 10070
    },
    {
      "entropy": 5.4688667297363285,
      "epoch": 0.9999007542675665,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004908742997809951,
      "loss": 5.1789,
      "mean_token_accuracy": 0.1693073406815529,
      "num_tokens": 22074375.0,
      "step": 10075
    },
    {
      "entropy": 5.58414249420166,
      "epoch": 1.0003969829297341,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004908643086690744,
      "loss": 5.2739,
      "mean_token_accuracy": 0.15737587660551072,
      "num_tokens": 22084040.0,
      "step": 10080
    },
    {
      "entropy": 5.544978380203247,
      "epoch": 1.0008932115919016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490854312204184,
      "loss": 5.1243,
      "mean_token_accuracy": 0.1701789453625679,
      "num_tokens": 22095964.0,
      "step": 10085
    },
    {
      "entropy": 5.488552665710449,
      "epoch": 1.001389440254069,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004908443103865719,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16692287027835845,
      "num_tokens": 22106395.0,
      "step": 10090
    },
    {
      "entropy": 5.451639413833618,
      "epoch": 1.0018856689162365,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004908343032164859,
      "loss": 5.039,
      "mean_token_accuracy": 0.17925864905118943,
      "num_tokens": 22115903.0,
      "step": 10095
    },
    {
      "entropy": 5.570309114456177,
      "epoch": 1.0023818975784042,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004908242906941743,
      "loss": 5.2511,
      "mean_token_accuracy": 0.1675417900085449,
      "num_tokens": 22127302.0,
      "step": 10100
    },
    {
      "entropy": 5.627327632904053,
      "epoch": 1.0028781262405717,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004908142728198853,
      "loss": 5.2671,
      "mean_token_accuracy": 0.1644710287451744,
      "num_tokens": 22136916.0,
      "step": 10105
    },
    {
      "entropy": 5.5305054664611815,
      "epoch": 1.0033743549027392,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004908042495938673,
      "loss": 5.1724,
      "mean_token_accuracy": 0.17025380283594133,
      "num_tokens": 22146857.0,
      "step": 10110
    },
    {
      "entropy": 5.512251710891723,
      "epoch": 1.0038705835649067,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004907942210163692,
      "loss": 5.2021,
      "mean_token_accuracy": 0.17099592983722686,
      "num_tokens": 22158039.0,
      "step": 10115
    },
    {
      "entropy": 5.59221510887146,
      "epoch": 1.0043668122270741,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004907841870876392,
      "loss": 5.2036,
      "mean_token_accuracy": 0.16703417897224426,
      "num_tokens": 22167779.0,
      "step": 10120
    },
    {
      "entropy": 5.527970600128174,
      "epoch": 1.0048630408892418,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004907741478079263,
      "loss": 5.2042,
      "mean_token_accuracy": 0.16038398891687394,
      "num_tokens": 22179405.0,
      "step": 10125
    },
    {
      "entropy": 5.474751329421997,
      "epoch": 1.0053592695514093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004907641031774796,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1645661011338234,
      "num_tokens": 22189716.0,
      "step": 10130
    },
    {
      "entropy": 5.543405437469483,
      "epoch": 1.0058554982135768,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000490754053196548,
      "loss": 5.1322,
      "mean_token_accuracy": 0.16880860775709153,
      "num_tokens": 22199610.0,
      "step": 10135
    },
    {
      "entropy": 5.573388671875,
      "epoch": 1.0063517268757443,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004907439978653809,
      "loss": 5.1563,
      "mean_token_accuracy": 0.16745618879795074,
      "num_tokens": 22210900.0,
      "step": 10140
    },
    {
      "entropy": 5.4497514247894285,
      "epoch": 1.006847955537912,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004907339371842275,
      "loss": 5.1432,
      "mean_token_accuracy": 0.18161418288946152,
      "num_tokens": 22222156.0,
      "step": 10145
    },
    {
      "entropy": 5.4877588748931885,
      "epoch": 1.0073441842000794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004907238711533373,
      "loss": 5.1399,
      "mean_token_accuracy": 0.17118117660284043,
      "num_tokens": 22232744.0,
      "step": 10150
    },
    {
      "entropy": 5.519137048721314,
      "epoch": 1.007840412862247,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004907137997729599,
      "loss": 5.1693,
      "mean_token_accuracy": 0.17034501135349273,
      "num_tokens": 22242291.0,
      "step": 10155
    },
    {
      "entropy": 5.552015924453736,
      "epoch": 1.0083366415244144,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000490703723043345,
      "loss": 5.1771,
      "mean_token_accuracy": 0.16895595490932463,
      "num_tokens": 22253517.0,
      "step": 10160
    },
    {
      "entropy": 5.593055295944214,
      "epoch": 1.008832870186582,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004906936409647428,
      "loss": 5.2794,
      "mean_token_accuracy": 0.1588181659579277,
      "num_tokens": 22265364.0,
      "step": 10165
    },
    {
      "entropy": 5.519245147705078,
      "epoch": 1.0093290988487496,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000490683553537403,
      "loss": 5.2514,
      "mean_token_accuracy": 0.16883167773485183,
      "num_tokens": 22276169.0,
      "step": 10170
    },
    {
      "entropy": 5.541768503189087,
      "epoch": 1.009825327510917,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004906734607615758,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1684512749314308,
      "num_tokens": 22287451.0,
      "step": 10175
    },
    {
      "entropy": 5.534524440765381,
      "epoch": 1.0103215561730845,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004906633626375114,
      "loss": 5.1841,
      "mean_token_accuracy": 0.16539886444807053,
      "num_tokens": 22298206.0,
      "step": 10180
    },
    {
      "entropy": 5.537748241424561,
      "epoch": 1.010817784835252,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004906532591654604,
      "loss": 5.1136,
      "mean_token_accuracy": 0.17467305511236192,
      "num_tokens": 22308950.0,
      "step": 10185
    },
    {
      "entropy": 5.4958549499511715,
      "epoch": 1.0113140134974197,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004906431503456733,
      "loss": 5.1582,
      "mean_token_accuracy": 0.16435473561286926,
      "num_tokens": 22319875.0,
      "step": 10190
    },
    {
      "entropy": 5.481654787063599,
      "epoch": 1.0118102421595871,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004906330361784007,
      "loss": 5.1204,
      "mean_token_accuracy": 0.1727033317089081,
      "num_tokens": 22330781.0,
      "step": 10195
    },
    {
      "entropy": 5.52365403175354,
      "epoch": 1.0123064708217546,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004906229166638936,
      "loss": 5.1664,
      "mean_token_accuracy": 0.17334433495998383,
      "num_tokens": 22341459.0,
      "step": 10200
    },
    {
      "entropy": 5.493383502960205,
      "epoch": 1.012802699483922,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004906127918024026,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17117146849632264,
      "num_tokens": 22352360.0,
      "step": 10205
    },
    {
      "entropy": 5.505085134506226,
      "epoch": 1.0132989281460898,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004906026615941791,
      "loss": 5.1737,
      "mean_token_accuracy": 0.1652582108974457,
      "num_tokens": 22363027.0,
      "step": 10210
    },
    {
      "entropy": 5.566107559204101,
      "epoch": 1.0137951568082573,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004905925260394741,
      "loss": 5.2497,
      "mean_token_accuracy": 0.15793152898550034,
      "num_tokens": 22374816.0,
      "step": 10215
    },
    {
      "entropy": 5.581737470626831,
      "epoch": 1.0142913854704247,
      "grad_norm": 0.921875,
      "learning_rate": 0.000490582385138539,
      "loss": 5.2473,
      "mean_token_accuracy": 0.1624793216586113,
      "num_tokens": 22385294.0,
      "step": 10220
    },
    {
      "entropy": 5.611543369293213,
      "epoch": 1.0147876141325922,
      "grad_norm": 0.875,
      "learning_rate": 0.0004905722388916254,
      "loss": 5.2141,
      "mean_token_accuracy": 0.16168177723884583,
      "num_tokens": 22396435.0,
      "step": 10225
    },
    {
      "entropy": 5.477035427093506,
      "epoch": 1.01528384279476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905620872989846,
      "loss": 5.1686,
      "mean_token_accuracy": 0.17243548035621642,
      "num_tokens": 22407160.0,
      "step": 10230
    },
    {
      "entropy": 5.540544414520264,
      "epoch": 1.0157800714569274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004905519303608687,
      "loss": 5.2317,
      "mean_token_accuracy": 0.16321061551570892,
      "num_tokens": 22419186.0,
      "step": 10235
    },
    {
      "entropy": 5.490870571136474,
      "epoch": 1.0162763001190949,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004905417680775294,
      "loss": 5.1421,
      "mean_token_accuracy": 0.16054446548223494,
      "num_tokens": 22430884.0,
      "step": 10240
    },
    {
      "entropy": 5.481886339187622,
      "epoch": 1.0167725287812623,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004905316004492186,
      "loss": 5.1493,
      "mean_token_accuracy": 0.16458605527877807,
      "num_tokens": 22441494.0,
      "step": 10245
    },
    {
      "entropy": 5.540880155563355,
      "epoch": 1.01726875744343,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004905214274761885,
      "loss": 5.2455,
      "mean_token_accuracy": 0.16141593605279922,
      "num_tokens": 22453127.0,
      "step": 10250
    },
    {
      "entropy": 5.5083168029785154,
      "epoch": 1.0177649861055975,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004905112491586915,
      "loss": 5.1694,
      "mean_token_accuracy": 0.17111731469631195,
      "num_tokens": 22465230.0,
      "step": 10255
    },
    {
      "entropy": 5.572668170928955,
      "epoch": 1.018261214767765,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049050106549698,
      "loss": 5.148,
      "mean_token_accuracy": 0.17091835737228395,
      "num_tokens": 22476450.0,
      "step": 10260
    },
    {
      "entropy": 5.557070779800415,
      "epoch": 1.0187574434299325,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004904908764913063,
      "loss": 5.2138,
      "mean_token_accuracy": 0.16103540360927582,
      "num_tokens": 22487557.0,
      "step": 10265
    },
    {
      "entropy": 5.494057559967041,
      "epoch": 1.0192536720921,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004904806821419233,
      "loss": 5.2289,
      "mean_token_accuracy": 0.17572347819805145,
      "num_tokens": 22498893.0,
      "step": 10270
    },
    {
      "entropy": 5.506337594985962,
      "epoch": 1.0197499007542676,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904704824490835,
      "loss": 5.2023,
      "mean_token_accuracy": 0.16585147231817246,
      "num_tokens": 22509873.0,
      "step": 10275
    },
    {
      "entropy": 5.577800941467285,
      "epoch": 1.020246129416435,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004904602774130402,
      "loss": 5.196,
      "mean_token_accuracy": 0.170413938164711,
      "num_tokens": 22521501.0,
      "step": 10280
    },
    {
      "entropy": 5.509530639648437,
      "epoch": 1.0207423580786026,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004904500670340463,
      "loss": 5.1973,
      "mean_token_accuracy": 0.16456578075885772,
      "num_tokens": 22532350.0,
      "step": 10285
    },
    {
      "entropy": 5.503167009353637,
      "epoch": 1.02123858674077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490439851312355,
      "loss": 5.138,
      "mean_token_accuracy": 0.17081402838230134,
      "num_tokens": 22542632.0,
      "step": 10290
    },
    {
      "entropy": 5.501250410079956,
      "epoch": 1.0217348154029378,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004904296302482196,
      "loss": 5.2109,
      "mean_token_accuracy": 0.15983759313821794,
      "num_tokens": 22552894.0,
      "step": 10295
    },
    {
      "entropy": 5.54918794631958,
      "epoch": 1.0222310440651052,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904194038418935,
      "loss": 5.2753,
      "mean_token_accuracy": 0.1689208820462227,
      "num_tokens": 22563279.0,
      "step": 10300
    },
    {
      "entropy": 5.520370626449585,
      "epoch": 1.0227272727272727,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004904091720936305,
      "loss": 5.1989,
      "mean_token_accuracy": 0.1598470240831375,
      "num_tokens": 22574217.0,
      "step": 10305
    },
    {
      "entropy": 5.571960067749023,
      "epoch": 1.0232235013894402,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004903989350036841,
      "loss": 5.2074,
      "mean_token_accuracy": 0.16496137380599976,
      "num_tokens": 22585001.0,
      "step": 10310
    },
    {
      "entropy": 5.592140150070191,
      "epoch": 1.0237197300516079,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903886925723083,
      "loss": 5.2524,
      "mean_token_accuracy": 0.1585180029273033,
      "num_tokens": 22596068.0,
      "step": 10315
    },
    {
      "entropy": 5.597181367874145,
      "epoch": 1.0242159587137754,
      "grad_norm": 0.90625,
      "learning_rate": 0.000490378444799757,
      "loss": 5.2069,
      "mean_token_accuracy": 0.15527265816926955,
      "num_tokens": 22606818.0,
      "step": 10320
    },
    {
      "entropy": 5.506962442398072,
      "epoch": 1.0247121873759428,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004903681916862844,
      "loss": 5.1835,
      "mean_token_accuracy": 0.16809259802103044,
      "num_tokens": 22617429.0,
      "step": 10325
    },
    {
      "entropy": 5.453313827514648,
      "epoch": 1.0252084160381103,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004903579332321447,
      "loss": 5.1383,
      "mean_token_accuracy": 0.17030074149370195,
      "num_tokens": 22628619.0,
      "step": 10330
    },
    {
      "entropy": 5.5100743770599365,
      "epoch": 1.0257046447002778,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903476694375924,
      "loss": 5.1996,
      "mean_token_accuracy": 0.17016830295324326,
      "num_tokens": 22640753.0,
      "step": 10335
    },
    {
      "entropy": 5.5347325801849365,
      "epoch": 1.0262008733624455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004903374003028819,
      "loss": 5.2896,
      "mean_token_accuracy": 0.1573953852057457,
      "num_tokens": 22652814.0,
      "step": 10340
    },
    {
      "entropy": 5.5958513736724855,
      "epoch": 1.026697102024613,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004903271258282678,
      "loss": 5.2042,
      "mean_token_accuracy": 0.1654478281736374,
      "num_tokens": 22663848.0,
      "step": 10345
    },
    {
      "entropy": 5.501072216033935,
      "epoch": 1.0271933306867804,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903168460140049,
      "loss": 5.2188,
      "mean_token_accuracy": 0.16320096850395202,
      "num_tokens": 22676859.0,
      "step": 10350
    },
    {
      "entropy": 5.546438884735108,
      "epoch": 1.027689559348948,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903065608603484,
      "loss": 5.1609,
      "mean_token_accuracy": 0.16275554448366164,
      "num_tokens": 22687000.0,
      "step": 10355
    },
    {
      "entropy": 5.555092906951904,
      "epoch": 1.0281857880111156,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000490296270367553,
      "loss": 5.1508,
      "mean_token_accuracy": 0.16401474922895432,
      "num_tokens": 22697440.0,
      "step": 10360
    },
    {
      "entropy": 5.5434661388397215,
      "epoch": 1.028682016673283,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004902859745358741,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17007107585668563,
      "num_tokens": 22709031.0,
      "step": 10365
    },
    {
      "entropy": 5.570174026489258,
      "epoch": 1.0291782453354505,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004902756733655668,
      "loss": 5.3315,
      "mean_token_accuracy": 0.1571981817483902,
      "num_tokens": 22721247.0,
      "step": 10370
    },
    {
      "entropy": 5.543043756484986,
      "epoch": 1.029674473997618,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004902653668568868,
      "loss": 5.2173,
      "mean_token_accuracy": 0.1675634041428566,
      "num_tokens": 22731784.0,
      "step": 10375
    },
    {
      "entropy": 5.613442611694336,
      "epoch": 1.0301707026597857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004902550550100896,
      "loss": 5.2476,
      "mean_token_accuracy": 0.15973951518535615,
      "num_tokens": 22742715.0,
      "step": 10380
    },
    {
      "entropy": 5.593408679962158,
      "epoch": 1.0306669313219532,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004902447378254307,
      "loss": 5.3128,
      "mean_token_accuracy": 0.1622981071472168,
      "num_tokens": 22752086.0,
      "step": 10385
    },
    {
      "entropy": 5.530457639694214,
      "epoch": 1.0311631599841207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004902344153031663,
      "loss": 5.2017,
      "mean_token_accuracy": 0.16949792206287384,
      "num_tokens": 22764407.0,
      "step": 10390
    },
    {
      "entropy": 5.5086853981018065,
      "epoch": 1.0316593886462881,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490224087443552,
      "loss": 5.1839,
      "mean_token_accuracy": 0.16677739173173906,
      "num_tokens": 22774457.0,
      "step": 10395
    },
    {
      "entropy": 5.526150798797607,
      "epoch": 1.0321556173084558,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004902137542468442,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16944143325090408,
      "num_tokens": 22786726.0,
      "step": 10400
    },
    {
      "entropy": 5.538340377807617,
      "epoch": 1.0326518459706233,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902034157132991,
      "loss": 5.2107,
      "mean_token_accuracy": 0.16216875612735748,
      "num_tokens": 22797677.0,
      "step": 10405
    },
    {
      "entropy": 5.487375974655151,
      "epoch": 1.0331480746327908,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004901930718431729,
      "loss": 5.1816,
      "mean_token_accuracy": 0.1696225330233574,
      "num_tokens": 22808230.0,
      "step": 10410
    },
    {
      "entropy": 5.5550378322601315,
      "epoch": 1.0336443032949583,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004901827226367223,
      "loss": 5.2329,
      "mean_token_accuracy": 0.16538107246160508,
      "num_tokens": 22821111.0,
      "step": 10415
    },
    {
      "entropy": 5.488890838623047,
      "epoch": 1.0341405319571257,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901723680942038,
      "loss": 5.14,
      "mean_token_accuracy": 0.1700485333800316,
      "num_tokens": 22832531.0,
      "step": 10420
    },
    {
      "entropy": 5.578724575042725,
      "epoch": 1.0346367606192934,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004901620082158742,
      "loss": 5.2248,
      "mean_token_accuracy": 0.1695777431130409,
      "num_tokens": 22845065.0,
      "step": 10425
    },
    {
      "entropy": 5.521230411529541,
      "epoch": 1.035132989281461,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004901516430019905,
      "loss": 5.2009,
      "mean_token_accuracy": 0.1583025112748146,
      "num_tokens": 22857652.0,
      "step": 10430
    },
    {
      "entropy": 5.4886627197265625,
      "epoch": 1.0356292179436284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004901412724528097,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1656071200966835,
      "num_tokens": 22866827.0,
      "step": 10435
    },
    {
      "entropy": 5.52748670578003,
      "epoch": 1.0361254466057959,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004901308965685889,
      "loss": 5.1771,
      "mean_token_accuracy": 0.16696747541427612,
      "num_tokens": 22877669.0,
      "step": 10440
    },
    {
      "entropy": 5.523586130142212,
      "epoch": 1.0366216752679636,
      "grad_norm": 1.0,
      "learning_rate": 0.0004901205153495853,
      "loss": 5.1742,
      "mean_token_accuracy": 0.1659868061542511,
      "num_tokens": 22888122.0,
      "step": 10445
    },
    {
      "entropy": 5.562162208557129,
      "epoch": 1.037117903930131,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004901101287960566,
      "loss": 5.2404,
      "mean_token_accuracy": 0.1635718375444412,
      "num_tokens": 22898977.0,
      "step": 10450
    },
    {
      "entropy": 5.55811619758606,
      "epoch": 1.0376141325922985,
      "grad_norm": 1.125,
      "learning_rate": 0.0004900997369082602,
      "loss": 5.1885,
      "mean_token_accuracy": 0.1671081006526947,
      "num_tokens": 22909666.0,
      "step": 10455
    },
    {
      "entropy": 5.484081220626831,
      "epoch": 1.038110361254466,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004900893396864539,
      "loss": 5.155,
      "mean_token_accuracy": 0.16652456670999527,
      "num_tokens": 22920258.0,
      "step": 10460
    },
    {
      "entropy": 5.511610698699951,
      "epoch": 1.0386065899166337,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004900789371308953,
      "loss": 5.297,
      "mean_token_accuracy": 0.1627779111266136,
      "num_tokens": 22931694.0,
      "step": 10465
    },
    {
      "entropy": 5.576668405532837,
      "epoch": 1.0391028185788012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004900685292418425,
      "loss": 5.2412,
      "mean_token_accuracy": 0.16220398098230362,
      "num_tokens": 22942462.0,
      "step": 10470
    },
    {
      "entropy": 5.539755249023438,
      "epoch": 1.0395990472409686,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004900581160195537,
      "loss": 5.1617,
      "mean_token_accuracy": 0.1728810116648674,
      "num_tokens": 22954752.0,
      "step": 10475
    },
    {
      "entropy": 5.560678339004516,
      "epoch": 1.040095275903136,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000490047697464287,
      "loss": 5.2686,
      "mean_token_accuracy": 0.16289661452174187,
      "num_tokens": 22965630.0,
      "step": 10480
    },
    {
      "entropy": 5.50338978767395,
      "epoch": 1.0405915045653038,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004900372735763009,
      "loss": 5.1752,
      "mean_token_accuracy": 0.1680406168103218,
      "num_tokens": 22976403.0,
      "step": 10485
    },
    {
      "entropy": 5.4619372367858885,
      "epoch": 1.0410877332274713,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004900268443558536,
      "loss": 5.1527,
      "mean_token_accuracy": 0.16580265015363693,
      "num_tokens": 22987558.0,
      "step": 10490
    },
    {
      "entropy": 5.541800451278687,
      "epoch": 1.0415839618896388,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000490016409803204,
      "loss": 5.2052,
      "mean_token_accuracy": 0.16486751586198806,
      "num_tokens": 22998340.0,
      "step": 10495
    },
    {
      "entropy": 5.5467846393585205,
      "epoch": 1.0420801905518062,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004900059699186108,
      "loss": 5.1623,
      "mean_token_accuracy": 0.16431513726711272,
      "num_tokens": 23010058.0,
      "step": 10500
    },
    {
      "entropy": 5.516400671005249,
      "epoch": 1.0425764192139737,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004899955247023328,
      "loss": 5.1454,
      "mean_token_accuracy": 0.17321737408638,
      "num_tokens": 23020777.0,
      "step": 10505
    },
    {
      "entropy": 5.566682243347168,
      "epoch": 1.0430726478761414,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000489985074154629,
      "loss": 5.3337,
      "mean_token_accuracy": 0.15668004155158996,
      "num_tokens": 23033510.0,
      "step": 10510
    },
    {
      "entropy": 5.5463779926300045,
      "epoch": 1.0435688765383089,
      "grad_norm": 0.875,
      "learning_rate": 0.0004899746182757587,
      "loss": 5.1543,
      "mean_token_accuracy": 0.17041831612586975,
      "num_tokens": 23044385.0,
      "step": 10515
    },
    {
      "entropy": 5.496448373794555,
      "epoch": 1.0440651052004764,
      "grad_norm": 0.875,
      "learning_rate": 0.0004899641570659811,
      "loss": 5.1713,
      "mean_token_accuracy": 0.17011463344097139,
      "num_tokens": 23054907.0,
      "step": 10520
    },
    {
      "entropy": 5.52000241279602,
      "epoch": 1.0445613338626438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899536905255555,
      "loss": 5.2862,
      "mean_token_accuracy": 0.16456290632486342,
      "num_tokens": 23067035.0,
      "step": 10525
    },
    {
      "entropy": 5.526064825057984,
      "epoch": 1.0450575625248115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899432186547416,
      "loss": 5.2047,
      "mean_token_accuracy": 0.1659158304333687,
      "num_tokens": 23077084.0,
      "step": 10530
    },
    {
      "entropy": 5.560393190383911,
      "epoch": 1.045553791186979,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004899327414537991,
      "loss": 5.2187,
      "mean_token_accuracy": 0.17221810072660446,
      "num_tokens": 23088287.0,
      "step": 10535
    },
    {
      "entropy": 5.472218132019043,
      "epoch": 1.0460500198491465,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004899222589229877,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1685749739408493,
      "num_tokens": 23099252.0,
      "step": 10540
    },
    {
      "entropy": 5.455918645858764,
      "epoch": 1.046546248511314,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004899117710625674,
      "loss": 5.1594,
      "mean_token_accuracy": 0.16976117193698884,
      "num_tokens": 23109752.0,
      "step": 10545
    },
    {
      "entropy": 5.520242977142334,
      "epoch": 1.0470424771734816,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004899012778727983,
      "loss": 5.2078,
      "mean_token_accuracy": 0.15944095700979233,
      "num_tokens": 23120218.0,
      "step": 10550
    },
    {
      "entropy": 5.569714117050171,
      "epoch": 1.0475387058356491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004898907793539406,
      "loss": 5.1776,
      "mean_token_accuracy": 0.17072055786848067,
      "num_tokens": 23131156.0,
      "step": 10555
    },
    {
      "entropy": 5.533555030822754,
      "epoch": 1.0480349344978166,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004898802755062546,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1638398624956608,
      "num_tokens": 23142248.0,
      "step": 10560
    },
    {
      "entropy": 5.502655649185181,
      "epoch": 1.048531163159984,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898697663300007,
      "loss": 5.233,
      "mean_token_accuracy": 0.1585530400276184,
      "num_tokens": 23153888.0,
      "step": 10565
    },
    {
      "entropy": 5.6199414253234865,
      "epoch": 1.0490273918221515,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004898592518254398,
      "loss": 5.271,
      "mean_token_accuracy": 0.16794027090072633,
      "num_tokens": 23164411.0,
      "step": 10570
    },
    {
      "entropy": 5.4982226371765135,
      "epoch": 1.0495236204843192,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004898487319928322,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1680179789662361,
      "num_tokens": 23176008.0,
      "step": 10575
    },
    {
      "entropy": 5.610879850387573,
      "epoch": 1.0500198491464867,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898382068324392,
      "loss": 5.3907,
      "mean_token_accuracy": 0.1491311475634575,
      "num_tokens": 23187163.0,
      "step": 10580
    },
    {
      "entropy": 5.6217811584472654,
      "epoch": 1.0505160778086542,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898276763445215,
      "loss": 5.2693,
      "mean_token_accuracy": 0.16425759643316268,
      "num_tokens": 23197567.0,
      "step": 10585
    },
    {
      "entropy": 5.626307964324951,
      "epoch": 1.0510123064708217,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004898171405293404,
      "loss": 5.2366,
      "mean_token_accuracy": 0.16573391407728194,
      "num_tokens": 23208900.0,
      "step": 10590
    },
    {
      "entropy": 5.562443637847901,
      "epoch": 1.0515085351329894,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004898065993871571,
      "loss": 5.2479,
      "mean_token_accuracy": 0.1618798017501831,
      "num_tokens": 23220874.0,
      "step": 10595
    },
    {
      "entropy": 5.504487419128418,
      "epoch": 1.0520047637951568,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000489796052918233,
      "loss": 5.1411,
      "mean_token_accuracy": 0.1672385111451149,
      "num_tokens": 23232051.0,
      "step": 10600
    },
    {
      "entropy": 5.5772566318511965,
      "epoch": 1.0525009924573243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897855011228295,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16635403782129288,
      "num_tokens": 23243159.0,
      "step": 10605
    },
    {
      "entropy": 5.514183282852173,
      "epoch": 1.0529972211194918,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897749440012085,
      "loss": 5.2076,
      "mean_token_accuracy": 0.1627987340092659,
      "num_tokens": 23255911.0,
      "step": 10610
    },
    {
      "entropy": 5.6060010433197025,
      "epoch": 1.0534934497816595,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004897643815536316,
      "loss": 5.3012,
      "mean_token_accuracy": 0.15436629354953765,
      "num_tokens": 23268773.0,
      "step": 10615
    },
    {
      "entropy": 5.497804498672485,
      "epoch": 1.053989678443827,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000489753813780361,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17430178821086884,
      "num_tokens": 23278980.0,
      "step": 10620
    },
    {
      "entropy": 5.4951714992523195,
      "epoch": 1.0544859071059944,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897432406816585,
      "loss": 5.2379,
      "mean_token_accuracy": 0.15753610730171203,
      "num_tokens": 23288941.0,
      "step": 10625
    },
    {
      "entropy": 5.509159183502197,
      "epoch": 1.054982135768162,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004897326622577863,
      "loss": 5.158,
      "mean_token_accuracy": 0.17109932452440263,
      "num_tokens": 23298941.0,
      "step": 10630
    },
    {
      "entropy": 5.515296888351441,
      "epoch": 1.0554783644303294,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004897220785090067,
      "loss": 5.1641,
      "mean_token_accuracy": 0.16556718796491623,
      "num_tokens": 23310979.0,
      "step": 10635
    },
    {
      "entropy": 5.529790353775025,
      "epoch": 1.055974593092497,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004897114894355824,
      "loss": 5.1251,
      "mean_token_accuracy": 0.1709260806441307,
      "num_tokens": 23321755.0,
      "step": 10640
    },
    {
      "entropy": 5.440692567825318,
      "epoch": 1.0564708217546646,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004897008950377758,
      "loss": 5.1211,
      "mean_token_accuracy": 0.1651822343468666,
      "num_tokens": 23331966.0,
      "step": 10645
    },
    {
      "entropy": 5.521901416778564,
      "epoch": 1.056967050416832,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004896902953158497,
      "loss": 5.1586,
      "mean_token_accuracy": 0.16464222967624664,
      "num_tokens": 23343866.0,
      "step": 10650
    },
    {
      "entropy": 5.502094745635986,
      "epoch": 1.0574632790789995,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004896796902700667,
      "loss": 5.0967,
      "mean_token_accuracy": 0.16428715139627456,
      "num_tokens": 23354835.0,
      "step": 10655
    },
    {
      "entropy": 5.540447092056274,
      "epoch": 1.0579595077411672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896690799006901,
      "loss": 5.1848,
      "mean_token_accuracy": 0.16812039464712142,
      "num_tokens": 23366446.0,
      "step": 10660
    },
    {
      "entropy": 5.50201644897461,
      "epoch": 1.0584557364033347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004896584642079828,
      "loss": 5.2278,
      "mean_token_accuracy": 0.16272713541984557,
      "num_tokens": 23377971.0,
      "step": 10665
    },
    {
      "entropy": 5.497213554382324,
      "epoch": 1.0589519650655022,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896478431922083,
      "loss": 5.0979,
      "mean_token_accuracy": 0.17634347081184387,
      "num_tokens": 23388238.0,
      "step": 10670
    },
    {
      "entropy": 5.523705768585205,
      "epoch": 1.0594481937276696,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004896372168536297,
      "loss": 5.1711,
      "mean_token_accuracy": 0.16594589203596116,
      "num_tokens": 23399319.0,
      "step": 10675
    },
    {
      "entropy": 5.498851156234741,
      "epoch": 1.0599444223898373,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004896265851925107,
      "loss": 5.2431,
      "mean_token_accuracy": 0.16213784813880922,
      "num_tokens": 23410882.0,
      "step": 10680
    },
    {
      "entropy": 5.519414710998535,
      "epoch": 1.0604406510520048,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004896159482091148,
      "loss": 5.181,
      "mean_token_accuracy": 0.16952611207962037,
      "num_tokens": 23421775.0,
      "step": 10685
    },
    {
      "entropy": 5.544865465164184,
      "epoch": 1.0609368797141723,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004896053059037058,
      "loss": 5.1889,
      "mean_token_accuracy": 0.16588531881570817,
      "num_tokens": 23432760.0,
      "step": 10690
    },
    {
      "entropy": 5.482811450958252,
      "epoch": 1.0614331083763398,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895946582765477,
      "loss": 5.1428,
      "mean_token_accuracy": 0.16119630634784698,
      "num_tokens": 23442934.0,
      "step": 10695
    },
    {
      "entropy": 5.488082933425903,
      "epoch": 1.0619293370385074,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895840053279045,
      "loss": 5.1622,
      "mean_token_accuracy": 0.1674580842256546,
      "num_tokens": 23453365.0,
      "step": 10700
    },
    {
      "entropy": 5.543407869338989,
      "epoch": 1.062425565700675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004895733470580404,
      "loss": 5.1387,
      "mean_token_accuracy": 0.1701241284608841,
      "num_tokens": 23464500.0,
      "step": 10705
    },
    {
      "entropy": 5.510054349899292,
      "epoch": 1.0629217943628424,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004895626834672196,
      "loss": 5.1958,
      "mean_token_accuracy": 0.16455799341201782,
      "num_tokens": 23475702.0,
      "step": 10710
    },
    {
      "entropy": 5.517324018478393,
      "epoch": 1.0634180230250099,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004895520145557066,
      "loss": 5.1745,
      "mean_token_accuracy": 0.16711115837097168,
      "num_tokens": 23487271.0,
      "step": 10715
    },
    {
      "entropy": 5.469934177398682,
      "epoch": 1.0639142516871773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895413403237659,
      "loss": 5.1914,
      "mean_token_accuracy": 0.16403663009405137,
      "num_tokens": 23497092.0,
      "step": 10720
    },
    {
      "entropy": 5.45293960571289,
      "epoch": 1.064410480349345,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004895306607716624,
      "loss": 5.1012,
      "mean_token_accuracy": 0.17000951915979384,
      "num_tokens": 23507487.0,
      "step": 10725
    },
    {
      "entropy": 5.518583297729492,
      "epoch": 1.0649067090115125,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004895199758996607,
      "loss": 5.1479,
      "mean_token_accuracy": 0.16758019030094146,
      "num_tokens": 23517571.0,
      "step": 10730
    },
    {
      "entropy": 5.590923738479614,
      "epoch": 1.06540293767368,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895092857080258,
      "loss": 5.2216,
      "mean_token_accuracy": 0.1604047805070877,
      "num_tokens": 23527723.0,
      "step": 10735
    },
    {
      "entropy": 5.550404071807861,
      "epoch": 1.0658991663358475,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004894985901970229,
      "loss": 5.2318,
      "mean_token_accuracy": 0.16199800223112107,
      "num_tokens": 23539382.0,
      "step": 10740
    },
    {
      "entropy": 5.508830165863037,
      "epoch": 1.0663953949980152,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004894878893669171,
      "loss": 5.1503,
      "mean_token_accuracy": 0.16686597913503648,
      "num_tokens": 23550011.0,
      "step": 10745
    },
    {
      "entropy": 5.482184648513794,
      "epoch": 1.0668916236601826,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894771832179739,
      "loss": 5.1763,
      "mean_token_accuracy": 0.1755485787987709,
      "num_tokens": 23560837.0,
      "step": 10750
    },
    {
      "entropy": 5.574608373641968,
      "epoch": 1.0673878523223501,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004894664717504587,
      "loss": 5.2473,
      "mean_token_accuracy": 0.16698541045188903,
      "num_tokens": 23572083.0,
      "step": 10755
    },
    {
      "entropy": 5.599781179428101,
      "epoch": 1.0678840809845176,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894557549646371,
      "loss": 5.2253,
      "mean_token_accuracy": 0.16507660299539567,
      "num_tokens": 23581799.0,
      "step": 10760
    },
    {
      "entropy": 5.526648235321045,
      "epoch": 1.0683803096466853,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004894450328607749,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1661207377910614,
      "num_tokens": 23592662.0,
      "step": 10765
    },
    {
      "entropy": 5.546693325042725,
      "epoch": 1.0688765383088528,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004894343054391379,
      "loss": 5.1997,
      "mean_token_accuracy": 0.16091420799493789,
      "num_tokens": 23604040.0,
      "step": 10770
    },
    {
      "entropy": 5.487579679489135,
      "epoch": 1.0693727669710202,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004894235726999924,
      "loss": 5.1095,
      "mean_token_accuracy": 0.16986099928617476,
      "num_tokens": 23616182.0,
      "step": 10775
    },
    {
      "entropy": 5.609894227981568,
      "epoch": 1.0698689956331877,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894128346436042,
      "loss": 5.3406,
      "mean_token_accuracy": 0.15028574615716933,
      "num_tokens": 23628184.0,
      "step": 10780
    },
    {
      "entropy": 5.519876384735108,
      "epoch": 1.0703652242953554,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004894020912702397,
      "loss": 5.1609,
      "mean_token_accuracy": 0.1701563999056816,
      "num_tokens": 23638701.0,
      "step": 10785
    },
    {
      "entropy": 5.578057479858399,
      "epoch": 1.0708614529575229,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004893913425801653,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1672061949968338,
      "num_tokens": 23649020.0,
      "step": 10790
    },
    {
      "entropy": 5.5062150955200195,
      "epoch": 1.0713576816196904,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004893805885736477,
      "loss": 5.1488,
      "mean_token_accuracy": 0.17218293845653534,
      "num_tokens": 23660610.0,
      "step": 10795
    },
    {
      "entropy": 5.493595981597901,
      "epoch": 1.0718539102818578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004893698292509534,
      "loss": 5.1482,
      "mean_token_accuracy": 0.17492779046297074,
      "num_tokens": 23671684.0,
      "step": 10800
    },
    {
      "entropy": 5.47934627532959,
      "epoch": 1.0723501389440253,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004893590646123492,
      "loss": 5.1119,
      "mean_token_accuracy": 0.17221522480249404,
      "num_tokens": 23683577.0,
      "step": 10805
    },
    {
      "entropy": 5.542874050140381,
      "epoch": 1.072846367606193,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004893482946581022,
      "loss": 5.2138,
      "mean_token_accuracy": 0.16746090054512025,
      "num_tokens": 23695335.0,
      "step": 10810
    },
    {
      "entropy": 5.612048721313476,
      "epoch": 1.0733425962683605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004893375193884793,
      "loss": 5.3213,
      "mean_token_accuracy": 0.15415417402982712,
      "num_tokens": 23708017.0,
      "step": 10815
    },
    {
      "entropy": 5.501496982574463,
      "epoch": 1.073838824930528,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004893267388037478,
      "loss": 5.1696,
      "mean_token_accuracy": 0.16515096127986909,
      "num_tokens": 23719735.0,
      "step": 10820
    },
    {
      "entropy": 5.466482162475586,
      "epoch": 1.0743350535926954,
      "grad_norm": 0.96875,
      "learning_rate": 0.000489315952904175,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17650029957294464,
      "num_tokens": 23730731.0,
      "step": 10825
    },
    {
      "entropy": 5.564831495285034,
      "epoch": 1.0748312822548631,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004893051616900283,
      "loss": 5.3251,
      "mean_token_accuracy": 0.16294406056404115,
      "num_tokens": 23742588.0,
      "step": 10830
    },
    {
      "entropy": 5.48547477722168,
      "epoch": 1.0753275109170306,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004892943651615756,
      "loss": 5.1683,
      "mean_token_accuracy": 0.16865325719118118,
      "num_tokens": 23752965.0,
      "step": 10835
    },
    {
      "entropy": 5.482129621505737,
      "epoch": 1.075823739579198,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004892835633190843,
      "loss": 5.1758,
      "mean_token_accuracy": 0.16761947125196458,
      "num_tokens": 23764756.0,
      "step": 10840
    },
    {
      "entropy": 5.439739561080932,
      "epoch": 1.0763199682413656,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004892727561628223,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18737726658582687,
      "num_tokens": 23775595.0,
      "step": 10845
    },
    {
      "entropy": 5.514721870422363,
      "epoch": 1.0768161969035333,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892619436930577,
      "loss": 5.1916,
      "mean_token_accuracy": 0.1641121283173561,
      "num_tokens": 23786721.0,
      "step": 10850
    },
    {
      "entropy": 5.48105001449585,
      "epoch": 1.0773124255657007,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892511259100585,
      "loss": 5.1279,
      "mean_token_accuracy": 0.16122917234897613,
      "num_tokens": 23797440.0,
      "step": 10855
    },
    {
      "entropy": 5.491940879821778,
      "epoch": 1.0778086542278682,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004892403028140932,
      "loss": 5.1842,
      "mean_token_accuracy": 0.16325613260269164,
      "num_tokens": 23807807.0,
      "step": 10860
    },
    {
      "entropy": 5.586642646789551,
      "epoch": 1.0783048828900357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048922947440543,
      "loss": 5.2769,
      "mean_token_accuracy": 0.15880824774503707,
      "num_tokens": 23818558.0,
      "step": 10865
    },
    {
      "entropy": 5.5508890628814695,
      "epoch": 1.0788011115522032,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004892186406843374,
      "loss": 5.2505,
      "mean_token_accuracy": 0.1610873818397522,
      "num_tokens": 23829228.0,
      "step": 10870
    },
    {
      "entropy": 5.503239297866822,
      "epoch": 1.0792973402143708,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892078016510841,
      "loss": 5.1152,
      "mean_token_accuracy": 0.17354386299848557,
      "num_tokens": 23840161.0,
      "step": 10875
    },
    {
      "entropy": 5.414162254333496,
      "epoch": 1.0797935688765383,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000489196957305939,
      "loss": 5.0593,
      "mean_token_accuracy": 0.17279131412506105,
      "num_tokens": 23850516.0,
      "step": 10880
    },
    {
      "entropy": 5.447337245941162,
      "epoch": 1.0802897975387058,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891861076491709,
      "loss": 5.1086,
      "mean_token_accuracy": 0.17305748015642167,
      "num_tokens": 23860589.0,
      "step": 10885
    },
    {
      "entropy": 5.456099367141723,
      "epoch": 1.0807860262008733,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004891752526810488,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16641634106636047,
      "num_tokens": 23871363.0,
      "step": 10890
    },
    {
      "entropy": 5.494942998886108,
      "epoch": 1.081282254863041,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489164392401842,
      "loss": 5.2039,
      "mean_token_accuracy": 0.17014809399843217,
      "num_tokens": 23882297.0,
      "step": 10895
    },
    {
      "entropy": 5.507961416244507,
      "epoch": 1.0817784835252084,
      "grad_norm": 1.0,
      "learning_rate": 0.0004891535268118196,
      "loss": 5.1313,
      "mean_token_accuracy": 0.16809219419956206,
      "num_tokens": 23892966.0,
      "step": 10900
    },
    {
      "entropy": 5.52854208946228,
      "epoch": 1.082274712187376,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004891426559112514,
      "loss": 5.1411,
      "mean_token_accuracy": 0.1709982633590698,
      "num_tokens": 23902918.0,
      "step": 10905
    },
    {
      "entropy": 5.600217580795288,
      "epoch": 1.0827709408495434,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004891317797004066,
      "loss": 5.3109,
      "mean_token_accuracy": 0.1550906404852867,
      "num_tokens": 23913692.0,
      "step": 10910
    },
    {
      "entropy": 5.545831823348999,
      "epoch": 1.083267169511711,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004891208981795552,
      "loss": 5.2543,
      "mean_token_accuracy": 0.163834585249424,
      "num_tokens": 23923714.0,
      "step": 10915
    },
    {
      "entropy": 5.447899246215821,
      "epoch": 1.0837633981738786,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004891100113489669,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17301519066095353,
      "num_tokens": 23934398.0,
      "step": 10920
    },
    {
      "entropy": 5.506496953964233,
      "epoch": 1.084259626836046,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004890991192089116,
      "loss": 5.2441,
      "mean_token_accuracy": 0.16503501832485198,
      "num_tokens": 23946181.0,
      "step": 10925
    },
    {
      "entropy": 5.5120765209198,
      "epoch": 1.0847558554982135,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004890882217596595,
      "loss": 5.204,
      "mean_token_accuracy": 0.1634614497423172,
      "num_tokens": 23959248.0,
      "step": 10930
    },
    {
      "entropy": 5.487981224060059,
      "epoch": 1.085252084160381,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004890773190014809,
      "loss": 5.1694,
      "mean_token_accuracy": 0.16803600788116455,
      "num_tokens": 23972676.0,
      "step": 10935
    },
    {
      "entropy": 5.5662150382995605,
      "epoch": 1.0857483128225487,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000489066410934646,
      "loss": 5.2639,
      "mean_token_accuracy": 0.16441008746623992,
      "num_tokens": 23983720.0,
      "step": 10940
    },
    {
      "entropy": 5.540129089355469,
      "epoch": 1.0862445414847162,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004890554975594255,
      "loss": 5.1821,
      "mean_token_accuracy": 0.175816547870636,
      "num_tokens": 23994730.0,
      "step": 10945
    },
    {
      "entropy": 5.566362190246582,
      "epoch": 1.0867407701468836,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004890445788760898,
      "loss": 5.2473,
      "mean_token_accuracy": 0.16224570870399474,
      "num_tokens": 24006022.0,
      "step": 10950
    },
    {
      "entropy": 5.568926000595093,
      "epoch": 1.0872369988090511,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004890336548849098,
      "loss": 5.274,
      "mean_token_accuracy": 0.16772584468126298,
      "num_tokens": 24016670.0,
      "step": 10955
    },
    {
      "entropy": 5.548723030090332,
      "epoch": 1.0877332274712188,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004890227255861564,
      "loss": 5.2007,
      "mean_token_accuracy": 0.16560838222503663,
      "num_tokens": 24027677.0,
      "step": 10960
    },
    {
      "entropy": 5.519962024688721,
      "epoch": 1.0882294561333863,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004890117909801005,
      "loss": 5.2083,
      "mean_token_accuracy": 0.16975989639759065,
      "num_tokens": 24039191.0,
      "step": 10965
    },
    {
      "entropy": 5.466756057739258,
      "epoch": 1.0887256847955538,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004890008510670134,
      "loss": 5.0876,
      "mean_token_accuracy": 0.17397142946720123,
      "num_tokens": 24049282.0,
      "step": 10970
    },
    {
      "entropy": 5.443380975723267,
      "epoch": 1.0892219134577212,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889899058471663,
      "loss": 5.1758,
      "mean_token_accuracy": 0.16746412366628646,
      "num_tokens": 24059506.0,
      "step": 10975
    },
    {
      "entropy": 5.458825588226318,
      "epoch": 1.089718142119889,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889789553208306,
      "loss": 5.1737,
      "mean_token_accuracy": 0.17015302628278733,
      "num_tokens": 24068973.0,
      "step": 10980
    },
    {
      "entropy": 5.60464129447937,
      "epoch": 1.0902143707820564,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000488967999488278,
      "loss": 5.2695,
      "mean_token_accuracy": 0.16808143854141236,
      "num_tokens": 24081012.0,
      "step": 10985
    },
    {
      "entropy": 5.553222513198852,
      "epoch": 1.0907105994442239,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048895703834978,
      "loss": 5.1503,
      "mean_token_accuracy": 0.16987540125846862,
      "num_tokens": 24090540.0,
      "step": 10990
    },
    {
      "entropy": 5.510127401351928,
      "epoch": 1.0912068281063914,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004889460719056085,
      "loss": 5.1868,
      "mean_token_accuracy": 0.1706799402832985,
      "num_tokens": 24101704.0,
      "step": 10995
    },
    {
      "entropy": 5.509523916244507,
      "epoch": 1.091703056768559,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004889351001560354,
      "loss": 5.2373,
      "mean_token_accuracy": 0.16208492815494538,
      "num_tokens": 24110685.0,
      "step": 11000
    },
    {
      "entropy": 5.529576635360717,
      "epoch": 1.0921992854307265,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889241231013329,
      "loss": 5.1516,
      "mean_token_accuracy": 0.1659996673464775,
      "num_tokens": 24121962.0,
      "step": 11005
    },
    {
      "entropy": 5.481543827056885,
      "epoch": 1.092695514092894,
      "grad_norm": 0.96875,
      "learning_rate": 0.000488913140741773,
      "loss": 5.1157,
      "mean_token_accuracy": 0.1703410878777504,
      "num_tokens": 24132279.0,
      "step": 11010
    },
    {
      "entropy": 5.456655597686767,
      "epoch": 1.0931917427550615,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004889021530776282,
      "loss": 5.192,
      "mean_token_accuracy": 0.17254822552204133,
      "num_tokens": 24144149.0,
      "step": 11015
    },
    {
      "entropy": 5.5698198795318605,
      "epoch": 1.093687971417229,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004888911601091709,
      "loss": 5.2111,
      "mean_token_accuracy": 0.1655818909406662,
      "num_tokens": 24155917.0,
      "step": 11020
    },
    {
      "entropy": 5.521628141403198,
      "epoch": 1.0941842000793967,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004888801618366737,
      "loss": 5.1727,
      "mean_token_accuracy": 0.1654084637761116,
      "num_tokens": 24166214.0,
      "step": 11025
    },
    {
      "entropy": 5.525131893157959,
      "epoch": 1.0946804287415641,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004888691582604095,
      "loss": 5.2312,
      "mean_token_accuracy": 0.16885906606912612,
      "num_tokens": 24177040.0,
      "step": 11030
    },
    {
      "entropy": 5.506588697433472,
      "epoch": 1.0951766574037316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000488858149380651,
      "loss": 5.1636,
      "mean_token_accuracy": 0.16811756938695907,
      "num_tokens": 24188130.0,
      "step": 11035
    },
    {
      "entropy": 5.576376485824585,
      "epoch": 1.095672886065899,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488847135197671,
      "loss": 5.2591,
      "mean_token_accuracy": 0.16895211189985276,
      "num_tokens": 24197393.0,
      "step": 11040
    },
    {
      "entropy": 5.506489276885986,
      "epoch": 1.0961691147280668,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488836115711743,
      "loss": 5.1955,
      "mean_token_accuracy": 0.17008643001317977,
      "num_tokens": 24208078.0,
      "step": 11045
    },
    {
      "entropy": 5.590869283676147,
      "epoch": 1.0966653433902342,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004888250909231401,
      "loss": 5.2475,
      "mean_token_accuracy": 0.1647603318095207,
      "num_tokens": 24219188.0,
      "step": 11050
    },
    {
      "entropy": 5.550053071975708,
      "epoch": 1.0971615720524017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004888140608321357,
      "loss": 5.2247,
      "mean_token_accuracy": 0.16255518943071365,
      "num_tokens": 24229315.0,
      "step": 11055
    },
    {
      "entropy": 5.550967121124268,
      "epoch": 1.0976578007145692,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004888030254390033,
      "loss": 5.2686,
      "mean_token_accuracy": 0.16403433084487914,
      "num_tokens": 24240580.0,
      "step": 11060
    },
    {
      "entropy": 5.582862377166748,
      "epoch": 1.098154029376737,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004887919847440166,
      "loss": 5.2216,
      "mean_token_accuracy": 0.1693173184990883,
      "num_tokens": 24252048.0,
      "step": 11065
    },
    {
      "entropy": 5.503943204879761,
      "epoch": 1.0986502580389044,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004887809387474494,
      "loss": 5.1577,
      "mean_token_accuracy": 0.16421120315790178,
      "num_tokens": 24262479.0,
      "step": 11070
    },
    {
      "entropy": 5.471037769317627,
      "epoch": 1.0991464867010718,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004887698874495755,
      "loss": 5.1471,
      "mean_token_accuracy": 0.1684648111462593,
      "num_tokens": 24272790.0,
      "step": 11075
    },
    {
      "entropy": 5.539792919158936,
      "epoch": 1.0996427153632393,
      "grad_norm": 0.921875,
      "learning_rate": 0.000488758830850669,
      "loss": 5.1976,
      "mean_token_accuracy": 0.1702119678258896,
      "num_tokens": 24284150.0,
      "step": 11080
    },
    {
      "entropy": 5.542705106735229,
      "epoch": 1.100138944025407,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004887477689510042,
      "loss": 5.1954,
      "mean_token_accuracy": 0.16757844984531403,
      "num_tokens": 24295683.0,
      "step": 11085
    },
    {
      "entropy": 5.491123104095459,
      "epoch": 1.1006351726875745,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004887367017508553,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17057539522647858,
      "num_tokens": 24306087.0,
      "step": 11090
    },
    {
      "entropy": 5.50247974395752,
      "epoch": 1.101131401349742,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004887256292504967,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1621381476521492,
      "num_tokens": 24316677.0,
      "step": 11095
    },
    {
      "entropy": 5.519765138626099,
      "epoch": 1.1016276300119094,
      "grad_norm": 1.0,
      "learning_rate": 0.000488714551450203,
      "loss": 5.116,
      "mean_token_accuracy": 0.17258864343166352,
      "num_tokens": 24327445.0,
      "step": 11100
    },
    {
      "entropy": 5.563187217712402,
      "epoch": 1.102123858674077,
      "grad_norm": 0.828125,
      "learning_rate": 0.000488703468350249,
      "loss": 5.2354,
      "mean_token_accuracy": 0.16506665498018264,
      "num_tokens": 24340231.0,
      "step": 11105
    },
    {
      "entropy": 5.592190504074097,
      "epoch": 1.1026200873362446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004886923799509095,
      "loss": 5.2553,
      "mean_token_accuracy": 0.15924686193466187,
      "num_tokens": 24352443.0,
      "step": 11110
    },
    {
      "entropy": 5.491620445251465,
      "epoch": 1.103116315998412,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004886812862524595,
      "loss": 5.1564,
      "mean_token_accuracy": 0.16286266595125198,
      "num_tokens": 24364485.0,
      "step": 11115
    },
    {
      "entropy": 5.537065267562866,
      "epoch": 1.1036125446605796,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004886701872551739,
      "loss": 5.2629,
      "mean_token_accuracy": 0.16509271711111068,
      "num_tokens": 24375943.0,
      "step": 11120
    },
    {
      "entropy": 5.579772520065307,
      "epoch": 1.104108773322747,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004886590829593281,
      "loss": 5.3016,
      "mean_token_accuracy": 0.15478023886680603,
      "num_tokens": 24387703.0,
      "step": 11125
    },
    {
      "entropy": 5.549969434738159,
      "epoch": 1.1046050019849147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886479733651974,
      "loss": 5.1479,
      "mean_token_accuracy": 0.16522146612405778,
      "num_tokens": 24398888.0,
      "step": 11130
    },
    {
      "entropy": 5.54872612953186,
      "epoch": 1.1051012306470822,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004886368584730575,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1706989586353302,
      "num_tokens": 24409643.0,
      "step": 11135
    },
    {
      "entropy": 5.463132095336914,
      "epoch": 1.1055974593092497,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004886257382831837,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1691657781600952,
      "num_tokens": 24421433.0,
      "step": 11140
    },
    {
      "entropy": 5.511273813247681,
      "epoch": 1.1060936879714172,
      "grad_norm": 0.859375,
      "learning_rate": 0.000488614612795852,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16050995439291,
      "num_tokens": 24433649.0,
      "step": 11145
    },
    {
      "entropy": 5.531495857238769,
      "epoch": 1.1065899166335849,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004886034820113382,
      "loss": 5.1544,
      "mean_token_accuracy": 0.17138154208660125,
      "num_tokens": 24445216.0,
      "step": 11150
    },
    {
      "entropy": 5.49173412322998,
      "epoch": 1.1070861452957523,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004885923459299182,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1640295907855034,
      "num_tokens": 24456026.0,
      "step": 11155
    },
    {
      "entropy": 5.458842372894287,
      "epoch": 1.1075823739579198,
      "grad_norm": 0.875,
      "learning_rate": 0.0004885812045518685,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1684365451335907,
      "num_tokens": 24466623.0,
      "step": 11160
    },
    {
      "entropy": 5.515522718429565,
      "epoch": 1.1080786026200873,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000488570057877465,
      "loss": 5.1319,
      "mean_token_accuracy": 0.16621588319540023,
      "num_tokens": 24476711.0,
      "step": 11165
    },
    {
      "entropy": 5.489848709106445,
      "epoch": 1.108574831282255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885589059069842,
      "loss": 5.1237,
      "mean_token_accuracy": 0.16855642646551133,
      "num_tokens": 24487987.0,
      "step": 11170
    },
    {
      "entropy": 5.463328313827515,
      "epoch": 1.1090710599444225,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004885477486407028,
      "loss": 5.1044,
      "mean_token_accuracy": 0.1716208502650261,
      "num_tokens": 24498482.0,
      "step": 11175
    },
    {
      "entropy": 5.490412950515747,
      "epoch": 1.10956728860659,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004885365860788974,
      "loss": 5.2781,
      "mean_token_accuracy": 0.16297549158334732,
      "num_tokens": 24509206.0,
      "step": 11180
    },
    {
      "entropy": 5.498608589172363,
      "epoch": 1.1100635172687574,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004885254182218448,
      "loss": 5.156,
      "mean_token_accuracy": 0.1660614401102066,
      "num_tokens": 24519534.0,
      "step": 11185
    },
    {
      "entropy": 5.546081066131592,
      "epoch": 1.1105597459309249,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885142450698219,
      "loss": 5.2105,
      "mean_token_accuracy": 0.1636567920446396,
      "num_tokens": 24531013.0,
      "step": 11190
    },
    {
      "entropy": 5.495534133911133,
      "epoch": 1.1110559745930926,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004885030666231057,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1732105940580368,
      "num_tokens": 24540865.0,
      "step": 11195
    },
    {
      "entropy": 5.532433557510376,
      "epoch": 1.11155220325526,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004884918828819738,
      "loss": 5.1606,
      "mean_token_accuracy": 0.17059827893972396,
      "num_tokens": 24550262.0,
      "step": 11200
    },
    {
      "entropy": 5.574435138702393,
      "epoch": 1.1120484319174275,
      "grad_norm": 0.90625,
      "learning_rate": 0.000488480693846703,
      "loss": 5.2198,
      "mean_token_accuracy": 0.15835732370615005,
      "num_tokens": 24561933.0,
      "step": 11205
    },
    {
      "entropy": 5.509217977523804,
      "epoch": 1.112544660579595,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004884694995175711,
      "loss": 5.1856,
      "mean_token_accuracy": 0.1694389432668686,
      "num_tokens": 24572698.0,
      "step": 11210
    },
    {
      "entropy": 5.454828691482544,
      "epoch": 1.1130408892417627,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004884582998948555,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17339539527893066,
      "num_tokens": 24581900.0,
      "step": 11215
    },
    {
      "entropy": 5.544699811935425,
      "epoch": 1.1135371179039302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004884470949788342,
      "loss": 5.1954,
      "mean_token_accuracy": 0.16751116067171096,
      "num_tokens": 24592163.0,
      "step": 11220
    },
    {
      "entropy": 5.528286695480347,
      "epoch": 1.1140333465660976,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004884358847697847,
      "loss": 5.084,
      "mean_token_accuracy": 0.17182679027318953,
      "num_tokens": 24602544.0,
      "step": 11225
    },
    {
      "entropy": 5.517291927337647,
      "epoch": 1.1145295752282651,
      "grad_norm": 1.0,
      "learning_rate": 0.0004884246692679852,
      "loss": 5.1796,
      "mean_token_accuracy": 0.1684107393026352,
      "num_tokens": 24613204.0,
      "step": 11230
    },
    {
      "entropy": 5.529194068908692,
      "epoch": 1.1150258038904326,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004884134484737139,
      "loss": 5.2245,
      "mean_token_accuracy": 0.1614619866013527,
      "num_tokens": 24624045.0,
      "step": 11235
    },
    {
      "entropy": 5.598123264312744,
      "epoch": 1.1155220325526003,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004884022223872489,
      "loss": 5.2628,
      "mean_token_accuracy": 0.15983810871839524,
      "num_tokens": 24635629.0,
      "step": 11240
    },
    {
      "entropy": 5.58865180015564,
      "epoch": 1.1160182612147678,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004883909910088686,
      "loss": 5.2511,
      "mean_token_accuracy": 0.16492237895727158,
      "num_tokens": 24646705.0,
      "step": 11245
    },
    {
      "entropy": 5.453462362289429,
      "epoch": 1.1165144898769352,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883797543388515,
      "loss": 5.0633,
      "mean_token_accuracy": 0.17696111649274826,
      "num_tokens": 24655969.0,
      "step": 11250
    },
    {
      "entropy": 5.450081253051758,
      "epoch": 1.1170107185391027,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004883685123774764,
      "loss": 5.1108,
      "mean_token_accuracy": 0.16749437600374223,
      "num_tokens": 24666739.0,
      "step": 11255
    },
    {
      "entropy": 5.548746013641358,
      "epoch": 1.1175069472012704,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004883572651250218,
      "loss": 5.2281,
      "mean_token_accuracy": 0.16558442264795303,
      "num_tokens": 24677775.0,
      "step": 11260
    },
    {
      "entropy": 5.484425020217896,
      "epoch": 1.118003175863438,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004883460125817669,
      "loss": 5.1252,
      "mean_token_accuracy": 0.16872948855161668,
      "num_tokens": 24688781.0,
      "step": 11265
    },
    {
      "entropy": 5.5241766452789305,
      "epoch": 1.1184994045256054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004883347547479905,
      "loss": 5.1478,
      "mean_token_accuracy": 0.1699554905295372,
      "num_tokens": 24698979.0,
      "step": 11270
    },
    {
      "entropy": 5.530127811431885,
      "epoch": 1.1189956331877728,
      "grad_norm": 1.0,
      "learning_rate": 0.000488323491623972,
      "loss": 5.1523,
      "mean_token_accuracy": 0.16834699660539626,
      "num_tokens": 24709419.0,
      "step": 11275
    },
    {
      "entropy": 5.529350757598877,
      "epoch": 1.1194918618499405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004883122232099905,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16816397607326508,
      "num_tokens": 24720495.0,
      "step": 11280
    },
    {
      "entropy": 5.52017126083374,
      "epoch": 1.119988090512108,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004883009495063256,
      "loss": 5.1399,
      "mean_token_accuracy": 0.17000134587287902,
      "num_tokens": 24730714.0,
      "step": 11285
    },
    {
      "entropy": 5.521626281738281,
      "epoch": 1.1204843191742755,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004882896705132566,
      "loss": 5.1745,
      "mean_token_accuracy": 0.17010299861431122,
      "num_tokens": 24741953.0,
      "step": 11290
    },
    {
      "entropy": 5.501421880722046,
      "epoch": 1.120980547836443,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048827838623106355,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16589792519807817,
      "num_tokens": 24753090.0,
      "step": 11295
    },
    {
      "entropy": 5.537659645080566,
      "epoch": 1.1214767764986107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488267096660026,
      "loss": 5.1641,
      "mean_token_accuracy": 0.1641857385635376,
      "num_tokens": 24763738.0,
      "step": 11300
    },
    {
      "entropy": 5.505961513519287,
      "epoch": 1.1219730051607781,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048825580180042406,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16297164857387542,
      "num_tokens": 24774557.0,
      "step": 11305
    },
    {
      "entropy": 5.500089359283447,
      "epoch": 1.1224692338229456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004882445016525377,
      "loss": 5.2164,
      "mean_token_accuracy": 0.16500159800052644,
      "num_tokens": 24784815.0,
      "step": 11310
    },
    {
      "entropy": 5.555668115615845,
      "epoch": 1.122965462485113,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048823319621664723,
      "loss": 5.2105,
      "mean_token_accuracy": 0.16841561794281007,
      "num_tokens": 24796149.0,
      "step": 11315
    },
    {
      "entropy": 5.515928506851196,
      "epoch": 1.1234616911472806,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048822188549303296,
      "loss": 5.2499,
      "mean_token_accuracy": 0.16200900077819824,
      "num_tokens": 24806562.0,
      "step": 11320
    },
    {
      "entropy": 5.503495740890503,
      "epoch": 1.1239579198094483,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004882105694819754,
      "loss": 5.1528,
      "mean_token_accuracy": 0.16908366233110428,
      "num_tokens": 24817569.0,
      "step": 11325
    },
    {
      "entropy": 5.506762504577637,
      "epoch": 1.1244541484716157,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004881992481837551,
      "loss": 5.1742,
      "mean_token_accuracy": 0.16180469542741777,
      "num_tokens": 24828759.0,
      "step": 11330
    },
    {
      "entropy": 5.4972292423248295,
      "epoch": 1.1249503771337832,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004881879215986529,
      "loss": 5.153,
      "mean_token_accuracy": 0.164307077229023,
      "num_tokens": 24840891.0,
      "step": 11335
    },
    {
      "entropy": 5.498653173446655,
      "epoch": 1.1254466057959507,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048817658972694967,
      "loss": 5.1589,
      "mean_token_accuracy": 0.17596266269683838,
      "num_tokens": 24851740.0,
      "step": 11340
    },
    {
      "entropy": 5.544261646270752,
      "epoch": 1.1259428344581184,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004881652525689263,
      "loss": 5.1723,
      "mean_token_accuracy": 0.16702407747507095,
      "num_tokens": 24862484.0,
      "step": 11345
    },
    {
      "entropy": 5.569014596939087,
      "epoch": 1.1264390631202859,
      "grad_norm": 0.90625,
      "learning_rate": 0.000488153910124864,
      "loss": 5.2501,
      "mean_token_accuracy": 0.16689283400774002,
      "num_tokens": 24872897.0,
      "step": 11350
    },
    {
      "entropy": 5.599718236923218,
      "epoch": 1.1269352917824533,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000488142562395044,
      "loss": 5.3023,
      "mean_token_accuracy": 0.16407682299613952,
      "num_tokens": 24884318.0,
      "step": 11355
    },
    {
      "entropy": 5.4846443176269535,
      "epoch": 1.1274315204446208,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048813120937974774,
      "loss": 5.1617,
      "mean_token_accuracy": 0.1691942557692528,
      "num_tokens": 24894284.0,
      "step": 11360
    },
    {
      "entropy": 5.515476131439209,
      "epoch": 1.1279277491067885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048811985107925684,
      "loss": 5.1928,
      "mean_token_accuracy": 0.1655291348695755,
      "num_tokens": 24904971.0,
      "step": 11365
    },
    {
      "entropy": 5.492429590225219,
      "epoch": 1.128423977768956,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048810848749385286,
      "loss": 5.096,
      "mean_token_accuracy": 0.1778580978512764,
      "num_tokens": 24915650.0,
      "step": 11370
    },
    {
      "entropy": 5.497278785705566,
      "epoch": 1.1289202064311235,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048809711862381753,
      "loss": 5.1735,
      "mean_token_accuracy": 0.17566781491041183,
      "num_tokens": 24925430.0,
      "step": 11375
    },
    {
      "entropy": 5.513367605209351,
      "epoch": 1.129416435093291,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048808574446943274,
      "loss": 5.1307,
      "mean_token_accuracy": 0.16776469349861145,
      "num_tokens": 24935568.0,
      "step": 11380
    },
    {
      "entropy": 5.449744367599488,
      "epoch": 1.1299126637554586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004880743650309808,
      "loss": 5.2101,
      "mean_token_accuracy": 0.16414861381053925,
      "num_tokens": 24945761.0,
      "step": 11385
    },
    {
      "entropy": 5.573803663253784,
      "epoch": 1.130408892417626,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048806298030874373,
      "loss": 5.3001,
      "mean_token_accuracy": 0.15671519488096236,
      "num_tokens": 24958259.0,
      "step": 11390
    },
    {
      "entropy": 5.467913579940796,
      "epoch": 1.1309051210797936,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048805159030300383,
      "loss": 5.0666,
      "mean_token_accuracy": 0.17357375621795654,
      "num_tokens": 24967697.0,
      "step": 11395
    },
    {
      "entropy": 5.524194288253784,
      "epoch": 1.131401349741961,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048804019501404357,
      "loss": 5.1813,
      "mean_token_accuracy": 0.16415926665067673,
      "num_tokens": 24979010.0,
      "step": 11400
    },
    {
      "entropy": 5.48268756866455,
      "epoch": 1.1318975784041285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048802879444214553,
      "loss": 5.0949,
      "mean_token_accuracy": 0.17602596879005433,
      "num_tokens": 24989700.0,
      "step": 11405
    },
    {
      "entropy": 5.4382977962493895,
      "epoch": 1.1323938070662962,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048801738858759243,
      "loss": 5.1103,
      "mean_token_accuracy": 0.17243608087301254,
      "num_tokens": 25000851.0,
      "step": 11410
    },
    {
      "entropy": 5.489388036727905,
      "epoch": 1.1328900357284637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004880059774506671,
      "loss": 5.1339,
      "mean_token_accuracy": 0.16720753759145737,
      "num_tokens": 25010336.0,
      "step": 11415
    },
    {
      "entropy": 5.499615478515625,
      "epoch": 1.1333862643906312,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004879945610316525,
      "loss": 5.1215,
      "mean_token_accuracy": 0.17649332582950591,
      "num_tokens": 25019711.0,
      "step": 11420
    },
    {
      "entropy": 5.4892127990722654,
      "epoch": 1.1338824930527986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048798313933083175,
      "loss": 5.1303,
      "mean_token_accuracy": 0.16969332545995713,
      "num_tokens": 25030349.0,
      "step": 11425
    },
    {
      "entropy": 5.4398046970367435,
      "epoch": 1.1343787217149663,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004879717123484882,
      "loss": 5.043,
      "mean_token_accuracy": 0.18267804831266404,
      "num_tokens": 25041356.0,
      "step": 11430
    },
    {
      "entropy": 5.456723976135254,
      "epoch": 1.1348749503771338,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000487960280084905,
      "loss": 5.1514,
      "mean_token_accuracy": 0.16610136032104492,
      "num_tokens": 25051609.0,
      "step": 11435
    },
    {
      "entropy": 5.499270248413086,
      "epoch": 1.1353711790393013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004879488425403657,
      "loss": 5.1607,
      "mean_token_accuracy": 0.17203362733125688,
      "num_tokens": 25061813.0,
      "step": 11440
    },
    {
      "entropy": 5.558045768737793,
      "epoch": 1.1358674077014688,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048793739971515407,
      "loss": 5.1774,
      "mean_token_accuracy": 0.1645442694425583,
      "num_tokens": 25072750.0,
      "step": 11445
    },
    {
      "entropy": 5.582309865951538,
      "epoch": 1.1363636363636362,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048792595160955374,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1600268818438053,
      "num_tokens": 25083735.0,
      "step": 11450
    },
    {
      "entropy": 5.5603477478027346,
      "epoch": 1.136859865025804,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048791449822384856,
      "loss": 5.208,
      "mean_token_accuracy": 0.1655652865767479,
      "num_tokens": 25094037.0,
      "step": 11455
    },
    {
      "entropy": 5.471120119094849,
      "epoch": 1.1373560936879714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004879030395583226,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17280908823013305,
      "num_tokens": 25103497.0,
      "step": 11460
    },
    {
      "entropy": 5.50687689781189,
      "epoch": 1.137852322350139,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004878915756132601,
      "loss": 5.1335,
      "mean_token_accuracy": 0.1672196567058563,
      "num_tokens": 25114322.0,
      "step": 11465
    },
    {
      "entropy": 5.481862258911133,
      "epoch": 1.1383485510123066,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048788010638894523,
      "loss": 5.1691,
      "mean_token_accuracy": 0.16281620264053345,
      "num_tokens": 25125355.0,
      "step": 11470
    },
    {
      "entropy": 5.629963636398315,
      "epoch": 1.138844779674474,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004878686318856624,
      "loss": 5.3165,
      "mean_token_accuracy": 0.1579107567667961,
      "num_tokens": 25137264.0,
      "step": 11475
    },
    {
      "entropy": 5.529286289215088,
      "epoch": 1.1393410083366415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048785715210369635,
      "loss": 5.1841,
      "mean_token_accuracy": 0.1646794408559799,
      "num_tokens": 25148294.0,
      "step": 11480
    },
    {
      "entropy": 5.434118413925171,
      "epoch": 1.139837236998809,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004878456670433314,
      "loss": 5.1627,
      "mean_token_accuracy": 0.17090788185596467,
      "num_tokens": 25159123.0,
      "step": 11485
    },
    {
      "entropy": 5.557157039642334,
      "epoch": 1.1403334656609765,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004878341767048527,
      "loss": 5.256,
      "mean_token_accuracy": 0.15760297924280167,
      "num_tokens": 25168910.0,
      "step": 11490
    },
    {
      "entropy": 5.4846690654754635,
      "epoch": 1.1408296943231442,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048782268108854494,
      "loss": 5.0741,
      "mean_token_accuracy": 0.1724128395318985,
      "num_tokens": 25180433.0,
      "step": 11495
    },
    {
      "entropy": 5.492997169494629,
      "epoch": 1.1413259229853117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004878111801946933,
      "loss": 5.1144,
      "mean_token_accuracy": 0.17157799154520034,
      "num_tokens": 25191558.0,
      "step": 11500
    },
    {
      "entropy": 5.420406770706177,
      "epoch": 1.1418221516474791,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048779967402358296,
      "loss": 5.0923,
      "mean_token_accuracy": 0.16707315742969514,
      "num_tokens": 25201241.0,
      "step": 11505
    },
    {
      "entropy": 5.478773880004883,
      "epoch": 1.1423183803096466,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004877881625754993,
      "loss": 5.1079,
      "mean_token_accuracy": 0.17110174745321274,
      "num_tokens": 25212305.0,
      "step": 11510
    },
    {
      "entropy": 5.466380023956299,
      "epoch": 1.1428146089718143,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048777664585072765,
      "loss": 5.1511,
      "mean_token_accuracy": 0.16649417728185653,
      "num_tokens": 25224296.0,
      "step": 11515
    },
    {
      "entropy": 5.530935907363892,
      "epoch": 1.1433108376339818,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004877651238495537,
      "loss": 5.1709,
      "mean_token_accuracy": 0.16680178046226501,
      "num_tokens": 25235730.0,
      "step": 11520
    },
    {
      "entropy": 5.591672658920288,
      "epoch": 1.1438070662961493,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048775359657226316,
      "loss": 5.2383,
      "mean_token_accuracy": 0.16196312755346298,
      "num_tokens": 25247066.0,
      "step": 11525
    },
    {
      "entropy": 5.568566370010376,
      "epoch": 1.1443032949583167,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004877420640191418,
      "loss": 5.2495,
      "mean_token_accuracy": 0.16054814755916597,
      "num_tokens": 25258417.0,
      "step": 11530
    },
    {
      "entropy": 5.541493701934814,
      "epoch": 1.1447995236204842,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004877305261904758,
      "loss": 5.1795,
      "mean_token_accuracy": 0.1618010565638542,
      "num_tokens": 25268268.0,
      "step": 11535
    },
    {
      "entropy": 5.5383076667785645,
      "epoch": 1.145295752282652,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048771898308655107,
      "loss": 5.1453,
      "mean_token_accuracy": 0.17134673446416854,
      "num_tokens": 25278437.0,
      "step": 11540
    },
    {
      "entropy": 5.562634801864624,
      "epoch": 1.1457919809448194,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004877074347076539,
      "loss": 5.1773,
      "mean_token_accuracy": 0.166947041451931,
      "num_tokens": 25290515.0,
      "step": 11545
    },
    {
      "entropy": 5.500912857055664,
      "epoch": 1.1462882096069869,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048769588105407075,
      "loss": 5.1758,
      "mean_token_accuracy": 0.1606058046221733,
      "num_tokens": 25301990.0,
      "step": 11550
    },
    {
      "entropy": 5.505390644073486,
      "epoch": 1.1467844382691545,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048768432212608807,
      "loss": 5.1618,
      "mean_token_accuracy": 0.16852937191724776,
      "num_tokens": 25311917.0,
      "step": 11555
    },
    {
      "entropy": 5.53971676826477,
      "epoch": 1.147280666931322,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004876727579239925,
      "loss": 5.2645,
      "mean_token_accuracy": 0.1716856211423874,
      "num_tokens": 25322578.0,
      "step": 11560
    },
    {
      "entropy": 5.563260269165039,
      "epoch": 1.1477768955934895,
      "grad_norm": 1.0,
      "learning_rate": 0.0004876611884480708,
      "loss": 5.0819,
      "mean_token_accuracy": 0.1765405535697937,
      "num_tokens": 25332068.0,
      "step": 11565
    },
    {
      "entropy": 5.52223539352417,
      "epoch": 1.148273124255657,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048764961369860977,
      "loss": 5.193,
      "mean_token_accuracy": 0.1673763409256935,
      "num_tokens": 25342511.0,
      "step": 11570
    },
    {
      "entropy": 5.439308023452758,
      "epoch": 1.1487693529178244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048763803367589663,
      "loss": 5.1169,
      "mean_token_accuracy": 0.17009825855493546,
      "num_tokens": 25353602.0,
      "step": 11575
    },
    {
      "entropy": 5.51558289527893,
      "epoch": 1.1492655815799921,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048762644838021843,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1687605381011963,
      "num_tokens": 25364570.0,
      "step": 11580
    },
    {
      "entropy": 5.5021144390106205,
      "epoch": 1.1497618102421596,
      "grad_norm": 1.0,
      "learning_rate": 0.00048761485781186243,
      "loss": 5.1315,
      "mean_token_accuracy": 0.17038799375295638,
      "num_tokens": 25375280.0,
      "step": 11585
    },
    {
      "entropy": 5.558493661880493,
      "epoch": 1.150258038904327,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004876032619711162,
      "loss": 5.2589,
      "mean_token_accuracy": 0.1679903730750084,
      "num_tokens": 25386786.0,
      "step": 11590
    },
    {
      "entropy": 5.46271595954895,
      "epoch": 1.1507542675664946,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048759166085826714,
      "loss": 5.1459,
      "mean_token_accuracy": 0.1688641294836998,
      "num_tokens": 25397867.0,
      "step": 11595
    },
    {
      "entropy": 5.5997779846191404,
      "epoch": 1.1512504962286623,
      "grad_norm": 1.0,
      "learning_rate": 0.000487580054473603,
      "loss": 5.2628,
      "mean_token_accuracy": 0.16131413280963897,
      "num_tokens": 25408394.0,
      "step": 11600
    },
    {
      "entropy": 5.652062654495239,
      "epoch": 1.1517467248908297,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048756844281741156,
      "loss": 5.2675,
      "mean_token_accuracy": 0.16178132295608522,
      "num_tokens": 25418815.0,
      "step": 11605
    },
    {
      "entropy": 5.513944864273071,
      "epoch": 1.1522429535529972,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048755682588998085,
      "loss": 5.177,
      "mean_token_accuracy": 0.16539522707462312,
      "num_tokens": 25430262.0,
      "step": 11610
    },
    {
      "entropy": 5.526991510391236,
      "epoch": 1.1527391822151647,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004875452036915987,
      "loss": 5.2418,
      "mean_token_accuracy": 0.1612778902053833,
      "num_tokens": 25441082.0,
      "step": 11615
    },
    {
      "entropy": 5.53685417175293,
      "epoch": 1.1532354108773322,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004875335762225536,
      "loss": 5.2061,
      "mean_token_accuracy": 0.16473386585712432,
      "num_tokens": 25452096.0,
      "step": 11620
    },
    {
      "entropy": 5.574972867965698,
      "epoch": 1.1537316395394999,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048752194348313383,
      "loss": 5.2331,
      "mean_token_accuracy": 0.15927488952875138,
      "num_tokens": 25462872.0,
      "step": 11625
    },
    {
      "entropy": 5.530872392654419,
      "epoch": 1.1542278682016673,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004875103054736278,
      "loss": 5.2082,
      "mean_token_accuracy": 0.16245064437389373,
      "num_tokens": 25473538.0,
      "step": 11630
    },
    {
      "entropy": 5.492141151428223,
      "epoch": 1.1547240968638348,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004874986621943241,
      "loss": 5.131,
      "mean_token_accuracy": 0.1716034948825836,
      "num_tokens": 25484707.0,
      "step": 11635
    },
    {
      "entropy": 5.550400829315185,
      "epoch": 1.1552203255260023,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048748701364551147,
      "loss": 5.1455,
      "mean_token_accuracy": 0.17218492925167084,
      "num_tokens": 25496155.0,
      "step": 11640
    },
    {
      "entropy": 5.517294406890869,
      "epoch": 1.15571655418817,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874753598274787,
      "loss": 5.1503,
      "mean_token_accuracy": 0.17198813557624817,
      "num_tokens": 25507146.0,
      "step": 11645
    },
    {
      "entropy": 5.501380491256714,
      "epoch": 1.1562127828503375,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004874637007405149,
      "loss": 5.1283,
      "mean_token_accuracy": 0.16840028315782546,
      "num_tokens": 25519099.0,
      "step": 11650
    },
    {
      "entropy": 5.497886800765992,
      "epoch": 1.156709011512505,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048745203638490917,
      "loss": 5.0977,
      "mean_token_accuracy": 0.17796667963266372,
      "num_tokens": 25529491.0,
      "step": 11655
    },
    {
      "entropy": 5.467025470733643,
      "epoch": 1.1572052401746724,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00048744036676095067,
      "loss": 5.0722,
      "mean_token_accuracy": 0.18368176221847535,
      "num_tokens": 25540889.0,
      "step": 11660
    },
    {
      "entropy": 5.4720970630645756,
      "epoch": 1.15770146883684,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004874286918689289,
      "loss": 5.1142,
      "mean_token_accuracy": 0.17076132893562318,
      "num_tokens": 25551334.0,
      "step": 11665
    },
    {
      "entropy": 5.487093544006347,
      "epoch": 1.1581976974990076,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048741701170913326,
      "loss": 5.138,
      "mean_token_accuracy": 0.17133939117193223,
      "num_tokens": 25561538.0,
      "step": 11670
    },
    {
      "entropy": 5.447670125961304,
      "epoch": 1.158693926161175,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048740532628185353,
      "loss": 5.115,
      "mean_token_accuracy": 0.17361915111541748,
      "num_tokens": 25571652.0,
      "step": 11675
    },
    {
      "entropy": 5.526296615600586,
      "epoch": 1.1591901548233425,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004873936355873793,
      "loss": 5.2305,
      "mean_token_accuracy": 0.1599818602204323,
      "num_tokens": 25582799.0,
      "step": 11680
    },
    {
      "entropy": 5.5175347328186035,
      "epoch": 1.1596863834855102,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004873819396260007,
      "loss": 5.151,
      "mean_token_accuracy": 0.1707802101969719,
      "num_tokens": 25594540.0,
      "step": 11685
    },
    {
      "entropy": 5.586673259735107,
      "epoch": 1.1601826121476777,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048737023839800754,
      "loss": 5.2071,
      "mean_token_accuracy": 0.16723292618989943,
      "num_tokens": 25604656.0,
      "step": 11690
    },
    {
      "entropy": 5.624115657806397,
      "epoch": 1.1606788408098452,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004873585319036901,
      "loss": 5.2024,
      "mean_token_accuracy": 0.16829923987388612,
      "num_tokens": 25615385.0,
      "step": 11695
    },
    {
      "entropy": 5.522167205810547,
      "epoch": 1.1611750694720127,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004873468201433387,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17563265562057495,
      "num_tokens": 25626456.0,
      "step": 11700
    },
    {
      "entropy": 5.514829683303833,
      "epoch": 1.1616712981341801,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004873351031172437,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17421717196702957,
      "num_tokens": 25636098.0,
      "step": 11705
    },
    {
      "entropy": 5.465155410766601,
      "epoch": 1.1621675267963478,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048732338082569564,
      "loss": 5.1757,
      "mean_token_accuracy": 0.17029872685670852,
      "num_tokens": 25646578.0,
      "step": 11710
    },
    {
      "entropy": 5.510841512680054,
      "epoch": 1.1626637554585153,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004873116532689853,
      "loss": 5.2971,
      "mean_token_accuracy": 0.16602183282375335,
      "num_tokens": 25658587.0,
      "step": 11715
    },
    {
      "entropy": 5.540887641906738,
      "epoch": 1.1631599841206828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004872999204474035,
      "loss": 5.1347,
      "mean_token_accuracy": 0.17244531512260436,
      "num_tokens": 25667745.0,
      "step": 11720
    },
    {
      "entropy": 5.583864307403564,
      "epoch": 1.1636562127828503,
      "grad_norm": 0.953125,
      "learning_rate": 0.000487288182361241,
      "loss": 5.2674,
      "mean_token_accuracy": 0.16363503336906432,
      "num_tokens": 25679034.0,
      "step": 11725
    },
    {
      "entropy": 5.565138339996338,
      "epoch": 1.164152441445018,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004872764390107891,
      "loss": 5.2187,
      "mean_token_accuracy": 0.16876654624938964,
      "num_tokens": 25689352.0,
      "step": 11730
    },
    {
      "entropy": 5.504155778884888,
      "epoch": 1.1646486701071854,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048726469039633895,
      "loss": 5.0953,
      "mean_token_accuracy": 0.17124411463737488,
      "num_tokens": 25700425.0,
      "step": 11735
    },
    {
      "entropy": 5.568137693405151,
      "epoch": 1.165144898769353,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048725293651818183,
      "loss": 5.2104,
      "mean_token_accuracy": 0.16713183224201203,
      "num_tokens": 25711641.0,
      "step": 11740
    },
    {
      "entropy": 5.561224746704101,
      "epoch": 1.1656411274315204,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004872411773766093,
      "loss": 5.2119,
      "mean_token_accuracy": 0.17242039144039153,
      "num_tokens": 25721725.0,
      "step": 11745
    },
    {
      "entropy": 5.485946941375732,
      "epoch": 1.1661373560936878,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048722941297191284,
      "loss": 5.2267,
      "mean_token_accuracy": 0.17198195457458496,
      "num_tokens": 25731465.0,
      "step": 11750
    },
    {
      "entropy": 5.513245582580566,
      "epoch": 1.1666335847558555,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004872176433043843,
      "loss": 5.0934,
      "mean_token_accuracy": 0.16455240398645402,
      "num_tokens": 25741864.0,
      "step": 11755
    },
    {
      "entropy": 5.52078013420105,
      "epoch": 1.167129813418023,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004872058683743155,
      "loss": 5.2264,
      "mean_token_accuracy": 0.1639114648103714,
      "num_tokens": 25753486.0,
      "step": 11760
    },
    {
      "entropy": 5.517141580581665,
      "epoch": 1.1676260420801905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048719408818199836,
      "loss": 5.1362,
      "mean_token_accuracy": 0.16645919382572175,
      "num_tokens": 25763361.0,
      "step": 11765
    },
    {
      "entropy": 5.495981931686401,
      "epoch": 1.1681222707423582,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048718230272772517,
      "loss": 5.1276,
      "mean_token_accuracy": 0.16334474682807923,
      "num_tokens": 25772157.0,
      "step": 11770
    },
    {
      "entropy": 5.465791320800781,
      "epoch": 1.1686184994045257,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048717051201178804,
      "loss": 5.1886,
      "mean_token_accuracy": 0.16778605282306672,
      "num_tokens": 25783948.0,
      "step": 11775
    },
    {
      "entropy": 5.528016996383667,
      "epoch": 1.1691147280666931,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004871587160344794,
      "loss": 5.2079,
      "mean_token_accuracy": 0.1633235901594162,
      "num_tokens": 25794762.0,
      "step": 11780
    },
    {
      "entropy": 5.4909075736999515,
      "epoch": 1.1696109567288606,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004871469147960917,
      "loss": 5.2109,
      "mean_token_accuracy": 0.16864209473133088,
      "num_tokens": 25805519.0,
      "step": 11785
    },
    {
      "entropy": 5.605147409439087,
      "epoch": 1.170107185391028,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004871351082969177,
      "loss": 5.2664,
      "mean_token_accuracy": 0.15602922588586807,
      "num_tokens": 25816772.0,
      "step": 11790
    },
    {
      "entropy": 5.527482414245606,
      "epoch": 1.1706034140531958,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004871232965372501,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1655966028571129,
      "num_tokens": 25826649.0,
      "step": 11795
    },
    {
      "entropy": 5.559004402160644,
      "epoch": 1.1710996427153633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004871114795173819,
      "loss": 5.2188,
      "mean_token_accuracy": 0.16191770732402802,
      "num_tokens": 25837263.0,
      "step": 11800
    },
    {
      "entropy": 5.610972499847412,
      "epoch": 1.1715958713775307,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004870996572376061,
      "loss": 5.2603,
      "mean_token_accuracy": 0.16292236745357513,
      "num_tokens": 25848687.0,
      "step": 11805
    },
    {
      "entropy": 5.519275379180908,
      "epoch": 1.1720921000396982,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004870878296982157,
      "loss": 5.1361,
      "mean_token_accuracy": 0.17344751954078674,
      "num_tokens": 25857982.0,
      "step": 11810
    },
    {
      "entropy": 5.467006349563599,
      "epoch": 1.172588328701866,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004870759968995042,
      "loss": 5.2058,
      "mean_token_accuracy": 0.16411461383104325,
      "num_tokens": 25869772.0,
      "step": 11815
    },
    {
      "entropy": 5.5400474071502686,
      "epoch": 1.1730845573640334,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048706415884176493,
      "loss": 5.2399,
      "mean_token_accuracy": 0.16621620953083038,
      "num_tokens": 25881165.0,
      "step": 11820
    },
    {
      "entropy": 5.549607133865356,
      "epoch": 1.1735807860262009,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004870523155252915,
      "loss": 5.1346,
      "mean_token_accuracy": 0.17199407070875167,
      "num_tokens": 25891135.0,
      "step": 11825
    },
    {
      "entropy": 5.510217761993408,
      "epoch": 1.1740770146883683,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004870404669503776,
      "loss": 5.1831,
      "mean_token_accuracy": 0.1648363009095192,
      "num_tokens": 25901837.0,
      "step": 11830
    },
    {
      "entropy": 5.584711313247681,
      "epoch": 1.1745732433505358,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000487028613117317,
      "loss": 5.2585,
      "mean_token_accuracy": 0.1621185451745987,
      "num_tokens": 25913833.0,
      "step": 11835
    },
    {
      "entropy": 5.544169759750366,
      "epoch": 1.1750694720127035,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004870167540264037,
      "loss": 5.1429,
      "mean_token_accuracy": 0.16707085371017455,
      "num_tokens": 25924803.0,
      "step": 11840
    },
    {
      "entropy": 5.5706329345703125,
      "epoch": 1.175565700674871,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004870048896779317,
      "loss": 5.2178,
      "mean_token_accuracy": 0.16983820647001266,
      "num_tokens": 25935582.0,
      "step": 11845
    },
    {
      "entropy": 5.56692328453064,
      "epoch": 1.1760619293370385,
      "grad_norm": 0.875,
      "learning_rate": 0.00048699302007219533,
      "loss": 5.2546,
      "mean_token_accuracy": 0.16156228631734848,
      "num_tokens": 25947515.0,
      "step": 11850
    },
    {
      "entropy": 5.539269208908081,
      "epoch": 1.1765581579992062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004869811452094889,
      "loss": 5.1685,
      "mean_token_accuracy": 0.1720904603600502,
      "num_tokens": 25957238.0,
      "step": 11855
    },
    {
      "entropy": 5.452915525436401,
      "epoch": 1.1770543866613736,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048696926509010673,
      "loss": 5.1578,
      "mean_token_accuracy": 0.1702004849910736,
      "num_tokens": 25968315.0,
      "step": 11860
    },
    {
      "entropy": 5.512604236602783,
      "epoch": 1.177550615323541,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004869573797143437,
      "loss": 5.0172,
      "mean_token_accuracy": 0.17446357011795044,
      "num_tokens": 25978172.0,
      "step": 11865
    },
    {
      "entropy": 5.517088556289673,
      "epoch": 1.1780468439857086,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004869454890824942,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1639511078596115,
      "num_tokens": 25989720.0,
      "step": 11870
    },
    {
      "entropy": 5.581359004974365,
      "epoch": 1.178543072647876,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048693359319485337,
      "loss": 5.2963,
      "mean_token_accuracy": 0.16070622503757476,
      "num_tokens": 26000394.0,
      "step": 11875
    },
    {
      "entropy": 5.56655592918396,
      "epoch": 1.1790393013100438,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004869216920517161,
      "loss": 5.2661,
      "mean_token_accuracy": 0.15837783813476564,
      "num_tokens": 26012183.0,
      "step": 11880
    },
    {
      "entropy": 5.574172782897949,
      "epoch": 1.1795355299722112,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048690978565337753,
      "loss": 5.1997,
      "mean_token_accuracy": 0.16853897720575334,
      "num_tokens": 26022711.0,
      "step": 11885
    },
    {
      "entropy": 5.466795015335083,
      "epoch": 1.1800317586343787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048689787400013287,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17570223212242125,
      "num_tokens": 26033800.0,
      "step": 11890
    },
    {
      "entropy": 5.510940170288086,
      "epoch": 1.1805279872965462,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004868859570922775,
      "loss": 5.2559,
      "mean_token_accuracy": 0.16346777826547623,
      "num_tokens": 26045461.0,
      "step": 11895
    },
    {
      "entropy": 5.587993001937866,
      "epoch": 1.1810242159587139,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004868740349301071,
      "loss": 5.2891,
      "mean_token_accuracy": 0.16492349952459334,
      "num_tokens": 26057674.0,
      "step": 11900
    },
    {
      "entropy": 5.58078441619873,
      "epoch": 1.1815204446208813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000486862107513917,
      "loss": 5.2001,
      "mean_token_accuracy": 0.16333111226558686,
      "num_tokens": 26069995.0,
      "step": 11905
    },
    {
      "entropy": 5.573717069625855,
      "epoch": 1.1820166732830488,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004868501748440033,
      "loss": 5.1692,
      "mean_token_accuracy": 0.1641562759876251,
      "num_tokens": 26079969.0,
      "step": 11910
    },
    {
      "entropy": 5.545842599868775,
      "epoch": 1.1825129019452163,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004868382369206617,
      "loss": 5.2104,
      "mean_token_accuracy": 0.1687634214758873,
      "num_tokens": 26089865.0,
      "step": 11915
    },
    {
      "entropy": 5.504620552062988,
      "epoch": 1.1830091306073838,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004868262937441883,
      "loss": 5.1836,
      "mean_token_accuracy": 0.1693308413028717,
      "num_tokens": 26101349.0,
      "step": 11920
    },
    {
      "entropy": 5.56647629737854,
      "epoch": 1.1835053592695515,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048681434531487925,
      "loss": 5.2236,
      "mean_token_accuracy": 0.16495576798915862,
      "num_tokens": 26113933.0,
      "step": 11925
    },
    {
      "entropy": 5.628184270858765,
      "epoch": 1.184001587931719,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048680239163303085,
      "loss": 5.2226,
      "mean_token_accuracy": 0.16842087209224701,
      "num_tokens": 26125106.0,
      "step": 11930
    },
    {
      "entropy": 5.575946950912476,
      "epoch": 1.1844978165938864,
      "grad_norm": 1.0,
      "learning_rate": 0.0004867904326989396,
      "loss": 5.2281,
      "mean_token_accuracy": 0.1631712019443512,
      "num_tokens": 26135131.0,
      "step": 11935
    },
    {
      "entropy": 5.5840850353240965,
      "epoch": 1.184994045256054,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004867784685129019,
      "loss": 5.2933,
      "mean_token_accuracy": 0.1589612513780594,
      "num_tokens": 26147053.0,
      "step": 11940
    },
    {
      "entropy": 5.56238808631897,
      "epoch": 1.1854902739182216,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048676649907521455,
      "loss": 5.1707,
      "mean_token_accuracy": 0.16851481646299363,
      "num_tokens": 26157465.0,
      "step": 11945
    },
    {
      "entropy": 5.536026096343994,
      "epoch": 1.185986502580389,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048675452438617434,
      "loss": 5.1519,
      "mean_token_accuracy": 0.16985128372907637,
      "num_tokens": 26168383.0,
      "step": 11950
    },
    {
      "entropy": 5.558902788162231,
      "epoch": 1.1864827312425565,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004867425444460782,
      "loss": 5.2271,
      "mean_token_accuracy": 0.16657354384660722,
      "num_tokens": 26178316.0,
      "step": 11955
    },
    {
      "entropy": 5.516727018356323,
      "epoch": 1.186978959904724,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004867305592552233,
      "loss": 5.1223,
      "mean_token_accuracy": 0.1706480175256729,
      "num_tokens": 26189494.0,
      "step": 11960
    },
    {
      "entropy": 5.489030647277832,
      "epoch": 1.1874751885668917,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004867185688139067,
      "loss": 5.2047,
      "mean_token_accuracy": 0.16336651146411896,
      "num_tokens": 26200865.0,
      "step": 11965
    },
    {
      "entropy": 5.51828031539917,
      "epoch": 1.1879714172290592,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004867065731224258,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16768025308847428,
      "num_tokens": 26212187.0,
      "step": 11970
    },
    {
      "entropy": 5.567694091796875,
      "epoch": 1.1884676458912267,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004866945721810781,
      "loss": 5.1746,
      "mean_token_accuracy": 0.1714477777481079,
      "num_tokens": 26222059.0,
      "step": 11975
    },
    {
      "entropy": 5.546338796615601,
      "epoch": 1.1889638745533941,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004866825659901612,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1669100672006607,
      "num_tokens": 26232630.0,
      "step": 11980
    },
    {
      "entropy": 5.541977787017823,
      "epoch": 1.1894601032155618,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048667055454997276,
      "loss": 5.2078,
      "mean_token_accuracy": 0.16799766421318055,
      "num_tokens": 26244232.0,
      "step": 11985
    },
    {
      "entropy": 5.607252407073974,
      "epoch": 1.1899563318777293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048665853786081063,
      "loss": 5.303,
      "mean_token_accuracy": 0.16935265064239502,
      "num_tokens": 26254721.0,
      "step": 11990
    },
    {
      "entropy": 5.56548056602478,
      "epoch": 1.1904525605398968,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048664651592297294,
      "loss": 5.1954,
      "mean_token_accuracy": 0.1668042689561844,
      "num_tokens": 26266174.0,
      "step": 11995
    },
    {
      "entropy": 5.593805599212646,
      "epoch": 1.1909487892020643,
      "grad_norm": 1.0,
      "learning_rate": 0.00048663448873675766,
      "loss": 5.2515,
      "mean_token_accuracy": 0.16460499316453933,
      "num_tokens": 26277501.0,
      "step": 12000
    },
    {
      "epoch": 1.1909487892020643,
      "eval_entropy": 5.354918643283459,
      "eval_loss": 5.2936177253723145,
      "eval_mean_token_accuracy": 0.16768150058030432,
      "eval_num_tokens": 26277501.0,
      "eval_runtime": 20.1864,
      "eval_samples_per_second": 1667.903,
      "eval_steps_per_second": 208.506,
      "step": 12000
    },
    {
      "entropy": 5.521618366241455,
      "epoch": 1.1914450178642317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004866224563024631,
      "loss": 5.1003,
      "mean_token_accuracy": 0.16884034872055054,
      "num_tokens": 26287291.0,
      "step": 12005
    },
    {
      "entropy": 5.487085056304932,
      "epoch": 1.1919412465263994,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004866104186203877,
      "loss": 5.064,
      "mean_token_accuracy": 0.17710066586732864,
      "num_tokens": 26297259.0,
      "step": 12010
    },
    {
      "entropy": 5.431519174575806,
      "epoch": 1.192437475188567,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048659837569082977,
      "loss": 5.1478,
      "mean_token_accuracy": 0.170035682618618,
      "num_tokens": 26307873.0,
      "step": 12015
    },
    {
      "entropy": 5.546828413009644,
      "epoch": 1.1929337038507344,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004865863275140882,
      "loss": 5.1929,
      "mean_token_accuracy": 0.16605871617794038,
      "num_tokens": 26320421.0,
      "step": 12020
    },
    {
      "entropy": 5.57897629737854,
      "epoch": 1.1934299325129019,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004865742740904615,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1601196214556694,
      "num_tokens": 26330932.0,
      "step": 12025
    },
    {
      "entropy": 5.583005380630493,
      "epoch": 1.1939261611750696,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004865622154202488,
      "loss": 5.2109,
      "mean_token_accuracy": 0.15890599340200423,
      "num_tokens": 26342286.0,
      "step": 12030
    },
    {
      "entropy": 5.495851993560791,
      "epoch": 1.194422389837237,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048655015150374904,
      "loss": 5.0916,
      "mean_token_accuracy": 0.17101497203111649,
      "num_tokens": 26352210.0,
      "step": 12035
    },
    {
      "entropy": 5.517520427703857,
      "epoch": 1.1949186184994045,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004865380823412613,
      "loss": 5.2359,
      "mean_token_accuracy": 0.16564253121614456,
      "num_tokens": 26363630.0,
      "step": 12040
    },
    {
      "entropy": 5.478893756866455,
      "epoch": 1.195414847161572,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000486526007933085,
      "loss": 5.1959,
      "mean_token_accuracy": 0.1663554683327675,
      "num_tokens": 26375723.0,
      "step": 12045
    },
    {
      "entropy": 5.5307151794433596,
      "epoch": 1.1959110758237395,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048651392827951944,
      "loss": 5.1481,
      "mean_token_accuracy": 0.16902241259813308,
      "num_tokens": 26387551.0,
      "step": 12050
    },
    {
      "entropy": 5.512584543228149,
      "epoch": 1.1964073044859072,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048650184338086436,
      "loss": 5.1013,
      "mean_token_accuracy": 0.1736832395195961,
      "num_tokens": 26399015.0,
      "step": 12055
    },
    {
      "entropy": 5.439255475997925,
      "epoch": 1.1969035331480746,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048648975323741916,
      "loss": 5.0679,
      "mean_token_accuracy": 0.17607573419809341,
      "num_tokens": 26409242.0,
      "step": 12060
    },
    {
      "entropy": 5.515980672836304,
      "epoch": 1.197399761810242,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048647765784948395,
      "loss": 5.2446,
      "mean_token_accuracy": 0.16756718903779982,
      "num_tokens": 26419189.0,
      "step": 12065
    },
    {
      "entropy": 5.543720293045044,
      "epoch": 1.1978959904724098,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004864655572173584,
      "loss": 5.1073,
      "mean_token_accuracy": 0.17223026007413864,
      "num_tokens": 26429957.0,
      "step": 12070
    },
    {
      "entropy": 5.478489208221435,
      "epoch": 1.1983922191345773,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048645345134134274,
      "loss": 5.1428,
      "mean_token_accuracy": 0.17450497448444366,
      "num_tokens": 26441088.0,
      "step": 12075
    },
    {
      "entropy": 5.514760971069336,
      "epoch": 1.1988884477967447,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048644134022173707,
      "loss": 5.204,
      "mean_token_accuracy": 0.16578360199928283,
      "num_tokens": 26450891.0,
      "step": 12080
    },
    {
      "entropy": 5.481167984008789,
      "epoch": 1.1993846764589122,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004864292238588419,
      "loss": 5.125,
      "mean_token_accuracy": 0.16866031885147095,
      "num_tokens": 26460445.0,
      "step": 12085
    },
    {
      "entropy": 5.5044684410095215,
      "epoch": 1.1998809051210797,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048641710225295746,
      "loss": 5.1764,
      "mean_token_accuracy": 0.16643801033496858,
      "num_tokens": 26472110.0,
      "step": 12090
    },
    {
      "entropy": 5.499518537521363,
      "epoch": 1.2003771337832474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048640497540438455,
      "loss": 5.098,
      "mean_token_accuracy": 0.17853525131940842,
      "num_tokens": 26483122.0,
      "step": 12095
    },
    {
      "entropy": 5.458695602416992,
      "epoch": 1.2008733624454149,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004863928433134237,
      "loss": 5.1524,
      "mean_token_accuracy": 0.1684826448559761,
      "num_tokens": 26495541.0,
      "step": 12100
    },
    {
      "entropy": 5.470108556747436,
      "epoch": 1.2013695911075823,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048638070598037585,
      "loss": 5.069,
      "mean_token_accuracy": 0.18441873490810395,
      "num_tokens": 26504904.0,
      "step": 12105
    },
    {
      "entropy": 5.548913240432739,
      "epoch": 1.2018658197697498,
      "grad_norm": 0.890625,
      "learning_rate": 0.000486368563405542,
      "loss": 5.1923,
      "mean_token_accuracy": 0.16773575395345688,
      "num_tokens": 26515850.0,
      "step": 12110
    },
    {
      "entropy": 5.585134029388428,
      "epoch": 1.2023620484319175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004863564155892232,
      "loss": 5.1706,
      "mean_token_accuracy": 0.1738531693816185,
      "num_tokens": 26526671.0,
      "step": 12115
    },
    {
      "entropy": 5.485021781921387,
      "epoch": 1.202858277094085,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004863442625317208,
      "loss": 5.2133,
      "mean_token_accuracy": 0.16620536297559738,
      "num_tokens": 26537919.0,
      "step": 12120
    },
    {
      "entropy": 5.570797967910766,
      "epoch": 1.2033545057562525,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048633210423333607,
      "loss": 5.221,
      "mean_token_accuracy": 0.16755713373422623,
      "num_tokens": 26547903.0,
      "step": 12125
    },
    {
      "entropy": 5.589312410354614,
      "epoch": 1.20385073441842,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004863199406943705,
      "loss": 5.2726,
      "mean_token_accuracy": 0.16519963443279267,
      "num_tokens": 26559253.0,
      "step": 12130
    },
    {
      "entropy": 5.572986268997193,
      "epoch": 1.2043469630805874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048630777191512584,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17038536667823792,
      "num_tokens": 26569562.0,
      "step": 12135
    },
    {
      "entropy": 5.494859743118286,
      "epoch": 1.2048431917427551,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048629559789590366,
      "loss": 5.1307,
      "mean_token_accuracy": 0.17446403652429582,
      "num_tokens": 26580701.0,
      "step": 12140
    },
    {
      "entropy": 5.462882232666016,
      "epoch": 1.2053394204049226,
      "grad_norm": 0.90625,
      "learning_rate": 0.000486283418637006,
      "loss": 5.1993,
      "mean_token_accuracy": 0.16541835069656372,
      "num_tokens": 26592233.0,
      "step": 12145
    },
    {
      "entropy": 5.5236142635345455,
      "epoch": 1.20583564906709,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004862712341387348,
      "loss": 5.1735,
      "mean_token_accuracy": 0.1667939767241478,
      "num_tokens": 26603073.0,
      "step": 12150
    },
    {
      "entropy": 5.522187995910644,
      "epoch": 1.2063318777292578,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004862590444013923,
      "loss": 5.1624,
      "mean_token_accuracy": 0.1626284494996071,
      "num_tokens": 26612497.0,
      "step": 12155
    },
    {
      "entropy": 5.544393920898438,
      "epoch": 1.2068281063914252,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004862468494252806,
      "loss": 5.1656,
      "mean_token_accuracy": 0.17306784838438033,
      "num_tokens": 26623616.0,
      "step": 12160
    },
    {
      "entropy": 5.583105087280273,
      "epoch": 1.2073243350535927,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004862346492107024,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1678294137120247,
      "num_tokens": 26634571.0,
      "step": 12165
    },
    {
      "entropy": 5.4941822528839115,
      "epoch": 1.2078205637157602,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048622244375796,
      "loss": 5.1455,
      "mean_token_accuracy": 0.17040181756019593,
      "num_tokens": 26646635.0,
      "step": 12170
    },
    {
      "entropy": 5.497052145004273,
      "epoch": 1.2083167923779277,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004862102330673561,
      "loss": 5.1605,
      "mean_token_accuracy": 0.16388365626335144,
      "num_tokens": 26657923.0,
      "step": 12175
    },
    {
      "entropy": 5.523563051223755,
      "epoch": 1.2088130210400954,
      "grad_norm": 1.0,
      "learning_rate": 0.00048619801713919357,
      "loss": 5.13,
      "mean_token_accuracy": 0.16854281425476075,
      "num_tokens": 26666989.0,
      "step": 12180
    },
    {
      "entropy": 5.512526988983154,
      "epoch": 1.2093092497022628,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004861857959737753,
      "loss": 5.2306,
      "mean_token_accuracy": 0.15994245707988738,
      "num_tokens": 26677325.0,
      "step": 12185
    },
    {
      "entropy": 5.477494430541992,
      "epoch": 1.2098054783644303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004861735695714044,
      "loss": 5.1383,
      "mean_token_accuracy": 0.17155715078115463,
      "num_tokens": 26688306.0,
      "step": 12190
    },
    {
      "entropy": 5.52174334526062,
      "epoch": 1.2103017070265978,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048616133793238394,
      "loss": 5.1435,
      "mean_token_accuracy": 0.16651806682348252,
      "num_tokens": 26698059.0,
      "step": 12195
    },
    {
      "entropy": 5.530380296707153,
      "epoch": 1.2107979356887655,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048614910105701737,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1674140140414238,
      "num_tokens": 26708165.0,
      "step": 12200
    },
    {
      "entropy": 5.568740272521973,
      "epoch": 1.211294164350933,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004861368589456081,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1701131835579872,
      "num_tokens": 26720010.0,
      "step": 12205
    },
    {
      "entropy": 5.519538927078247,
      "epoch": 1.2117903930131004,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004861246115984596,
      "loss": 5.1995,
      "mean_token_accuracy": 0.16924939006567002,
      "num_tokens": 26732756.0,
      "step": 12210
    },
    {
      "entropy": 5.571823835372925,
      "epoch": 1.212286621675268,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048611235901587575,
      "loss": 5.1768,
      "mean_token_accuracy": 0.16840581893920897,
      "num_tokens": 26744656.0,
      "step": 12215
    },
    {
      "entropy": 5.6199322700500485,
      "epoch": 1.2127828503374354,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004861001011981602,
      "loss": 5.222,
      "mean_token_accuracy": 0.16871834993362428,
      "num_tokens": 26756676.0,
      "step": 12220
    },
    {
      "entropy": 5.506667852401733,
      "epoch": 1.213279078999603,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004860878381456171,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1728111743927002,
      "num_tokens": 26767671.0,
      "step": 12225
    },
    {
      "entropy": 5.495825719833374,
      "epoch": 1.2137753076617706,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048607556985855047,
      "loss": 5.1376,
      "mean_token_accuracy": 0.17638952285051346,
      "num_tokens": 26779149.0,
      "step": 12230
    },
    {
      "entropy": 5.50913519859314,
      "epoch": 1.214271536323938,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048606329633726447,
      "loss": 5.1233,
      "mean_token_accuracy": 0.17241134196519853,
      "num_tokens": 26789825.0,
      "step": 12235
    },
    {
      "entropy": 5.528254222869873,
      "epoch": 1.2147677649861055,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004860510175820636,
      "loss": 5.1428,
      "mean_token_accuracy": 0.17063958197832108,
      "num_tokens": 26802145.0,
      "step": 12240
    },
    {
      "entropy": 5.55825629234314,
      "epoch": 1.2152639936482732,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004860387335932522,
      "loss": 5.1801,
      "mean_token_accuracy": 0.16894905269145966,
      "num_tokens": 26811925.0,
      "step": 12245
    },
    {
      "entropy": 5.545732736587524,
      "epoch": 1.2157602223104407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004860264443711351,
      "loss": 5.1695,
      "mean_token_accuracy": 0.16742797046899796,
      "num_tokens": 26821178.0,
      "step": 12250
    },
    {
      "entropy": 5.514078950881958,
      "epoch": 1.2162564509726081,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004860141499160167,
      "loss": 5.1562,
      "mean_token_accuracy": 0.17165687978267669,
      "num_tokens": 26831563.0,
      "step": 12255
    },
    {
      "entropy": 5.6370336532592775,
      "epoch": 1.2167526796347756,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004860018502282022,
      "loss": 5.2835,
      "mean_token_accuracy": 0.16029767245054244,
      "num_tokens": 26844596.0,
      "step": 12260
    },
    {
      "entropy": 5.589813327789306,
      "epoch": 1.2172489082969433,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048598954530799644,
      "loss": 5.1915,
      "mean_token_accuracy": 0.16703647673130034,
      "num_tokens": 26855149.0,
      "step": 12265
    },
    {
      "entropy": 5.5655313491821286,
      "epoch": 1.2177451369591108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048597723515570465,
      "loss": 5.1789,
      "mean_token_accuracy": 0.16997957080602646,
      "num_tokens": 26866200.0,
      "step": 12270
    },
    {
      "entropy": 5.493567752838135,
      "epoch": 1.2182413656212783,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000485964919771632,
      "loss": 5.1244,
      "mean_token_accuracy": 0.1723329558968544,
      "num_tokens": 26876834.0,
      "step": 12275
    },
    {
      "entropy": 5.5262291431427,
      "epoch": 1.2187375942834457,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048595259915608396,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17046998143196107,
      "num_tokens": 26887675.0,
      "step": 12280
    },
    {
      "entropy": 5.504010200500488,
      "epoch": 1.2192338229456134,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048594027330936594,
      "loss": 5.0861,
      "mean_token_accuracy": 0.17439922392368318,
      "num_tokens": 26897650.0,
      "step": 12285
    },
    {
      "entropy": 5.493280220031738,
      "epoch": 1.219730051607781,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048592794223178374,
      "loss": 5.1776,
      "mean_token_accuracy": 0.16597894579172134,
      "num_tokens": 26908404.0,
      "step": 12290
    },
    {
      "entropy": 5.474131011962891,
      "epoch": 1.2202262802699484,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048591560592364314,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1711941957473755,
      "num_tokens": 26917184.0,
      "step": 12295
    },
    {
      "entropy": 5.532330894470215,
      "epoch": 1.2207225089321159,
      "grad_norm": 1.0,
      "learning_rate": 0.0004859032643852499,
      "loss": 5.1572,
      "mean_token_accuracy": 0.17474822252988814,
      "num_tokens": 26927757.0,
      "step": 12300
    },
    {
      "entropy": 5.571223640441895,
      "epoch": 1.2212187375942833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004858909176169102,
      "loss": 5.1313,
      "mean_token_accuracy": 0.16809676140546798,
      "num_tokens": 26939036.0,
      "step": 12305
    },
    {
      "entropy": 5.503966426849365,
      "epoch": 1.221714966256451,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048587856561893014,
      "loss": 5.1287,
      "mean_token_accuracy": 0.17459919303655624,
      "num_tokens": 26949637.0,
      "step": 12310
    },
    {
      "entropy": 5.474580860137939,
      "epoch": 1.2222111949186185,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004858662083916161,
      "loss": 5.1642,
      "mean_token_accuracy": 0.17053653448820114,
      "num_tokens": 26961333.0,
      "step": 12315
    },
    {
      "entropy": 5.499805593490601,
      "epoch": 1.222707423580786,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004858538459352744,
      "loss": 5.1337,
      "mean_token_accuracy": 0.16748412400484086,
      "num_tokens": 26971830.0,
      "step": 12320
    },
    {
      "entropy": 5.505438327789307,
      "epoch": 1.2232036522429535,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004858414782502117,
      "loss": 5.1789,
      "mean_token_accuracy": 0.1688440665602684,
      "num_tokens": 26982808.0,
      "step": 12325
    },
    {
      "entropy": 5.5412757396698,
      "epoch": 1.2236998809051212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048582910533673474,
      "loss": 5.1368,
      "mean_token_accuracy": 0.16816569417715072,
      "num_tokens": 26992593.0,
      "step": 12330
    },
    {
      "entropy": 5.559628486633301,
      "epoch": 1.2241961095672886,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004858167271951502,
      "loss": 5.1269,
      "mean_token_accuracy": 0.16682473719120025,
      "num_tokens": 27003107.0,
      "step": 12335
    },
    {
      "entropy": 5.50968189239502,
      "epoch": 1.2246923382294561,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004858043438257651,
      "loss": 5.1462,
      "mean_token_accuracy": 0.1625799357891083,
      "num_tokens": 27013841.0,
      "step": 12340
    },
    {
      "entropy": 5.57605094909668,
      "epoch": 1.2251885668916236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004857919552288865,
      "loss": 5.1678,
      "mean_token_accuracy": 0.1665867656469345,
      "num_tokens": 27026236.0,
      "step": 12345
    },
    {
      "entropy": 5.546645450592041,
      "epoch": 1.225684795553791,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004857795614048217,
      "loss": 5.1736,
      "mean_token_accuracy": 0.17138618379831314,
      "num_tokens": 27037517.0,
      "step": 12350
    },
    {
      "entropy": 5.469513988494873,
      "epoch": 1.2261810242159588,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048576716235387786,
      "loss": 5.0776,
      "mean_token_accuracy": 0.18037131130695344,
      "num_tokens": 27048974.0,
      "step": 12355
    },
    {
      "entropy": 5.471656227111817,
      "epoch": 1.2266772528781262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048575475807636267,
      "loss": 5.1239,
      "mean_token_accuracy": 0.17492700815200807,
      "num_tokens": 27059637.0,
      "step": 12360
    },
    {
      "entropy": 5.55221266746521,
      "epoch": 1.2271734815402937,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048574234857258356,
      "loss": 5.1195,
      "mean_token_accuracy": 0.1669531837105751,
      "num_tokens": 27069584.0,
      "step": 12365
    },
    {
      "entropy": 5.544053554534912,
      "epoch": 1.2276697102024614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048572993384284833,
      "loss": 5.1585,
      "mean_token_accuracy": 0.1728387549519539,
      "num_tokens": 27080078.0,
      "step": 12370
    },
    {
      "entropy": 5.547524356842041,
      "epoch": 1.2281659388646289,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857175138874648,
      "loss": 5.1848,
      "mean_token_accuracy": 0.17266012728214264,
      "num_tokens": 27090396.0,
      "step": 12375
    },
    {
      "entropy": 5.5682628631591795,
      "epoch": 1.2286621675267964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048570508870674097,
      "loss": 5.1812,
      "mean_token_accuracy": 0.17109360098838805,
      "num_tokens": 27101052.0,
      "step": 12380
    },
    {
      "entropy": 5.522511148452759,
      "epoch": 1.2291583961889638,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048569265830098493,
      "loss": 5.1885,
      "mean_token_accuracy": 0.16452860683202744,
      "num_tokens": 27111352.0,
      "step": 12385
    },
    {
      "entropy": 5.490081548690796,
      "epoch": 1.2296546248511313,
      "grad_norm": 1.0,
      "learning_rate": 0.000485680222670505,
      "loss": 5.1662,
      "mean_token_accuracy": 0.1711289554834366,
      "num_tokens": 27122589.0,
      "step": 12390
    },
    {
      "entropy": 5.516182947158813,
      "epoch": 1.230150853513299,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004856677818156096,
      "loss": 5.1154,
      "mean_token_accuracy": 0.17326458245515824,
      "num_tokens": 27133445.0,
      "step": 12395
    },
    {
      "entropy": 5.5415243148803714,
      "epoch": 1.2306470821754665,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048565533573660707,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16823286414146424,
      "num_tokens": 27144043.0,
      "step": 12400
    },
    {
      "entropy": 5.599020195007324,
      "epoch": 1.231143310837634,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004856428844338062,
      "loss": 5.2169,
      "mean_token_accuracy": 0.16516635715961456,
      "num_tokens": 27155235.0,
      "step": 12405
    },
    {
      "entropy": 5.556141519546509,
      "epoch": 1.2316395394998014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004856304279075156,
      "loss": 5.1566,
      "mean_token_accuracy": 0.1692817509174347,
      "num_tokens": 27166360.0,
      "step": 12410
    },
    {
      "entropy": 5.517143297195434,
      "epoch": 1.2321357681619691,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004856179661580444,
      "loss": 5.2133,
      "mean_token_accuracy": 0.16420414596796035,
      "num_tokens": 27178018.0,
      "step": 12415
    },
    {
      "entropy": 5.602080917358398,
      "epoch": 1.2326319968241366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004856054991857013,
      "loss": 5.1755,
      "mean_token_accuracy": 0.1686885491013527,
      "num_tokens": 27187609.0,
      "step": 12420
    },
    {
      "entropy": 5.463791751861573,
      "epoch": 1.233128225486304,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048559302699079576,
      "loss": 5.1024,
      "mean_token_accuracy": 0.17678465098142623,
      "num_tokens": 27197477.0,
      "step": 12425
    },
    {
      "entropy": 5.504228019714356,
      "epoch": 1.2336244541484715,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004855805495736369,
      "loss": 5.1421,
      "mean_token_accuracy": 0.16893352419137955,
      "num_tokens": 27209276.0,
      "step": 12430
    },
    {
      "entropy": 5.580841493606568,
      "epoch": 1.234120682810639,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004855680669345342,
      "loss": 5.2084,
      "mean_token_accuracy": 0.16770119965076447,
      "num_tokens": 27221355.0,
      "step": 12435
    },
    {
      "entropy": 5.530679035186767,
      "epoch": 1.2346169114728067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048555557907379714,
      "loss": 5.1699,
      "mean_token_accuracy": 0.16668629348278047,
      "num_tokens": 27231408.0,
      "step": 12440
    },
    {
      "entropy": 5.533014917373658,
      "epoch": 1.2351131401349742,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048554308599173544,
      "loss": 5.1331,
      "mean_token_accuracy": 0.1693246692419052,
      "num_tokens": 27242577.0,
      "step": 12445
    },
    {
      "entropy": 5.568971824645996,
      "epoch": 1.2356093687971417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004855305876886589,
      "loss": 5.2606,
      "mean_token_accuracy": 0.16149116903543473,
      "num_tokens": 27252101.0,
      "step": 12450
    },
    {
      "entropy": 5.496117162704468,
      "epoch": 1.2361055974593094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048551808416487743,
      "loss": 5.2275,
      "mean_token_accuracy": 0.1620436504483223,
      "num_tokens": 27263281.0,
      "step": 12455
    },
    {
      "entropy": 5.483275413513184,
      "epoch": 1.2366018261214768,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004855055754207011,
      "loss": 5.0928,
      "mean_token_accuracy": 0.1683959648013115,
      "num_tokens": 27273747.0,
      "step": 12460
    },
    {
      "entropy": 5.542902660369873,
      "epoch": 1.2370980547836443,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004854930614564401,
      "loss": 5.2787,
      "mean_token_accuracy": 0.16046934723854064,
      "num_tokens": 27285282.0,
      "step": 12465
    },
    {
      "entropy": 5.535215520858765,
      "epoch": 1.2375942834458118,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004854805422724047,
      "loss": 5.1219,
      "mean_token_accuracy": 0.17540245056152343,
      "num_tokens": 27297258.0,
      "step": 12470
    },
    {
      "entropy": 5.556577491760254,
      "epoch": 1.2380905121079793,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004854680178689055,
      "loss": 5.1982,
      "mean_token_accuracy": 0.16466570049524307,
      "num_tokens": 27309366.0,
      "step": 12475
    },
    {
      "entropy": 5.50343770980835,
      "epoch": 1.238586740770147,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048545548824625287,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17359208315610886,
      "num_tokens": 27319878.0,
      "step": 12480
    },
    {
      "entropy": 5.472994804382324,
      "epoch": 1.2390829694323144,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004854429534047577,
      "loss": 5.0744,
      "mean_token_accuracy": 0.17389891147613526,
      "num_tokens": 27330056.0,
      "step": 12485
    },
    {
      "entropy": 5.533177185058594,
      "epoch": 1.239579198094482,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004854304133447307,
      "loss": 5.1639,
      "mean_token_accuracy": 0.17321743220090866,
      "num_tokens": 27342364.0,
      "step": 12490
    },
    {
      "entropy": 5.520839262008667,
      "epoch": 1.2400754267566494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048541786806648293,
      "loss": 5.1079,
      "mean_token_accuracy": 0.17276813685894013,
      "num_tokens": 27352651.0,
      "step": 12495
    },
    {
      "entropy": 5.539780902862549,
      "epoch": 1.240571655418817,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004854053175703254,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1699444830417633,
      "num_tokens": 27364161.0,
      "step": 12500
    },
    {
      "entropy": 5.514006948471069,
      "epoch": 1.2410678840809846,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004853927618565694,
      "loss": 5.1639,
      "mean_token_accuracy": 0.17667832225561142,
      "num_tokens": 27376461.0,
      "step": 12505
    },
    {
      "entropy": 5.53059606552124,
      "epoch": 1.241564112743152,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048538020092552626,
      "loss": 5.2072,
      "mean_token_accuracy": 0.1636827602982521,
      "num_tokens": 27386312.0,
      "step": 12510
    },
    {
      "entropy": 5.474186277389526,
      "epoch": 1.2420603414053195,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048536763477750747,
      "loss": 5.15,
      "mean_token_accuracy": 0.17513466626405716,
      "num_tokens": 27397971.0,
      "step": 12515
    },
    {
      "entropy": 5.5338935375213625,
      "epoch": 1.242556570067487,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004853550634128246,
      "loss": 5.0727,
      "mean_token_accuracy": 0.16713001281023027,
      "num_tokens": 27409942.0,
      "step": 12520
    },
    {
      "entropy": 5.5587090969085695,
      "epoch": 1.2430527987296547,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004853424868317895,
      "loss": 5.1791,
      "mean_token_accuracy": 0.16530852019786835,
      "num_tokens": 27421828.0,
      "step": 12525
    },
    {
      "entropy": 5.461774015426636,
      "epoch": 1.2435490273918222,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004853299050347139,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17822690606117247,
      "num_tokens": 27433321.0,
      "step": 12530
    },
    {
      "entropy": 5.493717098236084,
      "epoch": 1.2440452560539896,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048531731802190994,
      "loss": 5.132,
      "mean_token_accuracy": 0.16739080250263214,
      "num_tokens": 27443030.0,
      "step": 12535
    },
    {
      "entropy": 5.5233134746551515,
      "epoch": 1.244541484716157,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004853047257936897,
      "loss": 5.1337,
      "mean_token_accuracy": 0.17250906229019164,
      "num_tokens": 27453920.0,
      "step": 12540
    },
    {
      "entropy": 5.4648560047149655,
      "epoch": 1.2450377133783248,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004852921283503653,
      "loss": 5.1419,
      "mean_token_accuracy": 0.17465052306652068,
      "num_tokens": 27464014.0,
      "step": 12545
    },
    {
      "entropy": 5.606508064270019,
      "epoch": 1.2455339420404923,
      "grad_norm": 0.875,
      "learning_rate": 0.0004852795256922494,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16888390183448793,
      "num_tokens": 27475555.0,
      "step": 12550
    },
    {
      "entropy": 5.490652704238892,
      "epoch": 1.2460301707026598,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048526691781965425,
      "loss": 5.1191,
      "mean_token_accuracy": 0.17270503044128419,
      "num_tokens": 27486624.0,
      "step": 12555
    },
    {
      "entropy": 5.553306245803833,
      "epoch": 1.2465263993648272,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048525430473289274,
      "loss": 5.1776,
      "mean_token_accuracy": 0.16670886278152466,
      "num_tokens": 27497105.0,
      "step": 12560
    },
    {
      "entropy": 5.490859127044677,
      "epoch": 1.247022628026995,
      "grad_norm": 1.0,
      "learning_rate": 0.00048524168643227747,
      "loss": 5.118,
      "mean_token_accuracy": 0.17173615097999573,
      "num_tokens": 27508157.0,
      "step": 12565
    },
    {
      "entropy": 5.4995989322662355,
      "epoch": 1.2475188566891624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048522906291812145,
      "loss": 5.1291,
      "mean_token_accuracy": 0.16573092490434646,
      "num_tokens": 27519128.0,
      "step": 12570
    },
    {
      "entropy": 5.533367776870728,
      "epoch": 1.2480150853513299,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048521643419073763,
      "loss": 5.1508,
      "mean_token_accuracy": 0.17376940548419953,
      "num_tokens": 27529151.0,
      "step": 12575
    },
    {
      "entropy": 5.522754383087158,
      "epoch": 1.2485113140134974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004852038002504392,
      "loss": 5.123,
      "mean_token_accuracy": 0.17236966639757156,
      "num_tokens": 27538822.0,
      "step": 12580
    },
    {
      "entropy": 5.515516567230224,
      "epoch": 1.249007542675665,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048519116109753955,
      "loss": 5.1078,
      "mean_token_accuracy": 0.1759345442056656,
      "num_tokens": 27548779.0,
      "step": 12585
    },
    {
      "entropy": 5.493427515029907,
      "epoch": 1.2495037713378325,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048517851673235196,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17244986742734908,
      "num_tokens": 27559467.0,
      "step": 12590
    },
    {
      "entropy": 5.4921619415283205,
      "epoch": 1.25,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004851658671551901,
      "loss": 5.124,
      "mean_token_accuracy": 0.1739724650979042,
      "num_tokens": 27569518.0,
      "step": 12595
    },
    {
      "entropy": 5.501033449172974,
      "epoch": 1.2504962286621675,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048515321236636754,
      "loss": 5.116,
      "mean_token_accuracy": 0.1715035542845726,
      "num_tokens": 27580926.0,
      "step": 12600
    },
    {
      "entropy": 5.464759492874146,
      "epoch": 1.250992457324335,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004851405523661982,
      "loss": 5.1017,
      "mean_token_accuracy": 0.17170210033655167,
      "num_tokens": 27591354.0,
      "step": 12605
    },
    {
      "entropy": 5.49543080329895,
      "epoch": 1.2514886859865026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048512788715499597,
      "loss": 5.1549,
      "mean_token_accuracy": 0.16743143051862716,
      "num_tokens": 27601950.0,
      "step": 12610
    },
    {
      "entropy": 5.501472043991089,
      "epoch": 1.2519849146486701,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004851152167330749,
      "loss": 5.1185,
      "mean_token_accuracy": 0.17506213039159774,
      "num_tokens": 27614215.0,
      "step": 12615
    },
    {
      "entropy": 5.56895079612732,
      "epoch": 1.2524811433108376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048510254110074923,
      "loss": 5.1611,
      "mean_token_accuracy": 0.16871336251497268,
      "num_tokens": 27625486.0,
      "step": 12620
    },
    {
      "entropy": 5.5685515880584715,
      "epoch": 1.2529773719730053,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004850898602583333,
      "loss": 5.1369,
      "mean_token_accuracy": 0.17361843436956406,
      "num_tokens": 27636028.0,
      "step": 12625
    },
    {
      "entropy": 5.5593852519989015,
      "epoch": 1.2534736006351728,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048507717420614153,
      "loss": 5.1901,
      "mean_token_accuracy": 0.16501169353723527,
      "num_tokens": 27646368.0,
      "step": 12630
    },
    {
      "entropy": 5.456397438049317,
      "epoch": 1.2539698292973402,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004850644829444884,
      "loss": 5.1368,
      "mean_token_accuracy": 0.1622335880994797,
      "num_tokens": 27656337.0,
      "step": 12635
    },
    {
      "entropy": 5.491011285781861,
      "epoch": 1.2544660579595077,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048505178647368887,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1689670503139496,
      "num_tokens": 27666174.0,
      "step": 12640
    },
    {
      "entropy": 5.485871171951294,
      "epoch": 1.2549622866216752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004850390847940576,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17582833170890808,
      "num_tokens": 27675908.0,
      "step": 12645
    },
    {
      "entropy": 5.546670627593994,
      "epoch": 1.2554585152838427,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004850263779059097,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1675795942544937,
      "num_tokens": 27686746.0,
      "step": 12650
    },
    {
      "entropy": 5.564828729629516,
      "epoch": 1.2559547439460104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004850136658095601,
      "loss": 5.1471,
      "mean_token_accuracy": 0.17846752554178238,
      "num_tokens": 27699911.0,
      "step": 12655
    },
    {
      "entropy": 5.5291656970977785,
      "epoch": 1.2564509726081778,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048500094850532417,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17287509888410568,
      "num_tokens": 27709572.0,
      "step": 12660
    },
    {
      "entropy": 5.458728885650634,
      "epoch": 1.2569472012703453,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004849882259935172,
      "loss": 5.0717,
      "mean_token_accuracy": 0.1701683893799782,
      "num_tokens": 27720357.0,
      "step": 12665
    },
    {
      "entropy": 5.499853038787842,
      "epoch": 1.257443429932513,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004849754982744548,
      "loss": 5.1736,
      "mean_token_accuracy": 0.16684239953756333,
      "num_tokens": 27731284.0,
      "step": 12670
    },
    {
      "entropy": 5.563919401168823,
      "epoch": 1.2579396585946805,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004849627653484524,
      "loss": 5.2302,
      "mean_token_accuracy": 0.16523496955633163,
      "num_tokens": 27743619.0,
      "step": 12675
    },
    {
      "entropy": 5.52667989730835,
      "epoch": 1.258435887256848,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004849500272158259,
      "loss": 5.1412,
      "mean_token_accuracy": 0.1722117006778717,
      "num_tokens": 27753440.0,
      "step": 12680
    },
    {
      "entropy": 5.515864419937134,
      "epoch": 1.2589321159190154,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004849372838768911,
      "loss": 5.1646,
      "mean_token_accuracy": 0.17276352792978286,
      "num_tokens": 27764765.0,
      "step": 12685
    },
    {
      "entropy": 5.553107404708863,
      "epoch": 1.259428344581183,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000484924535331964,
      "loss": 5.2231,
      "mean_token_accuracy": 0.1613525055348873,
      "num_tokens": 27776817.0,
      "step": 12690
    },
    {
      "entropy": 5.556576204299927,
      "epoch": 1.2599245732433506,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048491178158136086,
      "loss": 5.1533,
      "mean_token_accuracy": 0.17035652846097946,
      "num_tokens": 27788276.0,
      "step": 12695
    },
    {
      "entropy": 5.41622371673584,
      "epoch": 1.260420801905518,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004848990226253978,
      "loss": 5.0481,
      "mean_token_accuracy": 0.17962032109498977,
      "num_tokens": 27798493.0,
      "step": 12700
    },
    {
      "entropy": 5.497841644287109,
      "epoch": 1.2609170305676856,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048488625846439135,
      "loss": 5.1403,
      "mean_token_accuracy": 0.16607593446969987,
      "num_tokens": 27809908.0,
      "step": 12705
    },
    {
      "entropy": 5.465724468231201,
      "epoch": 1.261413259229853,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004848734890986578,
      "loss": 5.1372,
      "mean_token_accuracy": 0.16898632943630218,
      "num_tokens": 27820954.0,
      "step": 12710
    },
    {
      "entropy": 5.569362545013428,
      "epoch": 1.2619094878920207,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004848607145285141,
      "loss": 5.2125,
      "mean_token_accuracy": 0.17060350477695466,
      "num_tokens": 27833645.0,
      "step": 12715
    },
    {
      "entropy": 5.518893527984619,
      "epoch": 1.2624057165541882,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004848479347542768,
      "loss": 5.102,
      "mean_token_accuracy": 0.17680168002843857,
      "num_tokens": 27844532.0,
      "step": 12720
    },
    {
      "entropy": 5.535592937469483,
      "epoch": 1.2629019452163557,
      "grad_norm": 0.90625,
      "learning_rate": 0.000484835149776263,
      "loss": 5.1855,
      "mean_token_accuracy": 0.17127904593944548,
      "num_tokens": 27855832.0,
      "step": 12725
    },
    {
      "entropy": 5.540751838684082,
      "epoch": 1.2633981738785232,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048482235959478953,
      "loss": 5.2469,
      "mean_token_accuracy": 0.1651556223630905,
      "num_tokens": 27867087.0,
      "step": 12730
    },
    {
      "entropy": 5.537172460556031,
      "epoch": 1.2638944025406906,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004848095642101737,
      "loss": 5.1417,
      "mean_token_accuracy": 0.1704070344567299,
      "num_tokens": 27877576.0,
      "step": 12735
    },
    {
      "entropy": 5.443470096588134,
      "epoch": 1.2643906312028583,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048479676362273277,
      "loss": 5.0312,
      "mean_token_accuracy": 0.17454352974891663,
      "num_tokens": 27888998.0,
      "step": 12740
    },
    {
      "entropy": 5.487547779083252,
      "epoch": 1.2648868598650258,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048478395783278425,
      "loss": 5.0638,
      "mean_token_accuracy": 0.17406281977891921,
      "num_tokens": 27899611.0,
      "step": 12745
    },
    {
      "entropy": 5.455127763748169,
      "epoch": 1.2653830885271933,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004847711468406455,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17804909199476243,
      "num_tokens": 27911787.0,
      "step": 12750
    },
    {
      "entropy": 5.589468717575073,
      "epoch": 1.265879317189361,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048475833064663443,
      "loss": 5.2531,
      "mean_token_accuracy": 0.15997117832303048,
      "num_tokens": 27922729.0,
      "step": 12755
    },
    {
      "entropy": 5.5026556015014645,
      "epoch": 1.2663755458515285,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004847455092510687,
      "loss": 5.1217,
      "mean_token_accuracy": 0.17359943836927413,
      "num_tokens": 27934357.0,
      "step": 12760
    },
    {
      "entropy": 5.547998380661011,
      "epoch": 1.266871774513696,
      "grad_norm": 1.0,
      "learning_rate": 0.0004847326826542663,
      "loss": 5.17,
      "mean_token_accuracy": 0.16755862683057784,
      "num_tokens": 27945355.0,
      "step": 12765
    },
    {
      "entropy": 5.485854387283325,
      "epoch": 1.2673680031758634,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048471985085654537,
      "loss": 5.1251,
      "mean_token_accuracy": 0.1764381483197212,
      "num_tokens": 27956426.0,
      "step": 12770
    },
    {
      "entropy": 5.546267175674439,
      "epoch": 1.2678642318380309,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048470701385822394,
      "loss": 5.1762,
      "mean_token_accuracy": 0.16643107831478118,
      "num_tokens": 27968431.0,
      "step": 12775
    },
    {
      "entropy": 5.557998943328857,
      "epoch": 1.2683604605001986,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048469417165962047,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1629766881465912,
      "num_tokens": 27978903.0,
      "step": 12780
    },
    {
      "entropy": 5.496155595779419,
      "epoch": 1.268856689162366,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004846813242610535,
      "loss": 5.1033,
      "mean_token_accuracy": 0.17364907264709473,
      "num_tokens": 27990791.0,
      "step": 12785
    },
    {
      "entropy": 5.504463529586792,
      "epoch": 1.2693529178245335,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004846684716628414,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17230179607868196,
      "num_tokens": 28000534.0,
      "step": 12790
    },
    {
      "entropy": 5.481282949447632,
      "epoch": 1.269849146486701,
      "grad_norm": 1.046875,
      "learning_rate": 0.000484655613865303,
      "loss": 5.1408,
      "mean_token_accuracy": 0.17063341438770294,
      "num_tokens": 28010573.0,
      "step": 12795
    },
    {
      "entropy": 5.523792123794555,
      "epoch": 1.2703453751488687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048464275086875717,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1632524013519287,
      "num_tokens": 28021255.0,
      "step": 12800
    },
    {
      "entropy": 5.534746313095093,
      "epoch": 1.2708416038110362,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004846298826735229,
      "loss": 5.1971,
      "mean_token_accuracy": 0.16938788294792176,
      "num_tokens": 28032529.0,
      "step": 12805
    },
    {
      "entropy": 5.528180646896362,
      "epoch": 1.2713378324732036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004846170092799192,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1700103461742401,
      "num_tokens": 28044225.0,
      "step": 12810
    },
    {
      "entropy": 5.456730127334595,
      "epoch": 1.2718340611353711,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048460413068826536,
      "loss": 5.1031,
      "mean_token_accuracy": 0.17549630701541902,
      "num_tokens": 28056175.0,
      "step": 12815
    },
    {
      "entropy": 5.543406391143799,
      "epoch": 1.2723302897975386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004845912468988807,
      "loss": 5.2176,
      "mean_token_accuracy": 0.16502409428358078,
      "num_tokens": 28067496.0,
      "step": 12820
    },
    {
      "entropy": 5.553228807449341,
      "epoch": 1.2728265184597063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004845783579120849,
      "loss": 5.1679,
      "mean_token_accuracy": 0.16465123742818832,
      "num_tokens": 28078140.0,
      "step": 12825
    },
    {
      "entropy": 5.523446655273437,
      "epoch": 1.2733227471218738,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004845654637281973,
      "loss": 5.1224,
      "mean_token_accuracy": 0.16985026597976685,
      "num_tokens": 28089612.0,
      "step": 12830
    },
    {
      "entropy": 5.4162962436676025,
      "epoch": 1.2738189757840412,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048455256434753774,
      "loss": 5.0095,
      "mean_token_accuracy": 0.17128804475069045,
      "num_tokens": 28100537.0,
      "step": 12835
    },
    {
      "entropy": 5.561084747314453,
      "epoch": 1.274315204446209,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004845396597704262,
      "loss": 5.2987,
      "mean_token_accuracy": 0.16176768243312836,
      "num_tokens": 28113317.0,
      "step": 12840
    },
    {
      "entropy": 5.4283397674560545,
      "epoch": 1.2748114331083764,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048452674999718257,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17267561703920364,
      "num_tokens": 28125098.0,
      "step": 12845
    },
    {
      "entropy": 5.554773139953613,
      "epoch": 1.2753076617705439,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048451383502812705,
      "loss": 5.196,
      "mean_token_accuracy": 0.1729342371225357,
      "num_tokens": 28135099.0,
      "step": 12850
    },
    {
      "entropy": 5.549209022521973,
      "epoch": 1.2758038904327114,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048450091486357987,
      "loss": 5.1789,
      "mean_token_accuracy": 0.16738216876983641,
      "num_tokens": 28146470.0,
      "step": 12855
    },
    {
      "entropy": 5.490698766708374,
      "epoch": 1.2763001190948788,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048448798950386147,
      "loss": 5.0988,
      "mean_token_accuracy": 0.17214040756225585,
      "num_tokens": 28156886.0,
      "step": 12860
    },
    {
      "entropy": 5.458677005767822,
      "epoch": 1.2767963477570463,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048447505894929233,
      "loss": 5.0969,
      "mean_token_accuracy": 0.17236390411853791,
      "num_tokens": 28166907.0,
      "step": 12865
    },
    {
      "entropy": 5.505178070068359,
      "epoch": 1.277292576419214,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048446212320019306,
      "loss": 5.1172,
      "mean_token_accuracy": 0.17355813682079316,
      "num_tokens": 28178015.0,
      "step": 12870
    },
    {
      "entropy": 5.56055474281311,
      "epoch": 1.2777888050813815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004844491822568846,
      "loss": 5.174,
      "mean_token_accuracy": 0.17019443064928055,
      "num_tokens": 28188518.0,
      "step": 12875
    },
    {
      "entropy": 5.493287038803101,
      "epoch": 1.278285033743549,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048443623611968764,
      "loss": 5.1257,
      "mean_token_accuracy": 0.16767860651016236,
      "num_tokens": 28199782.0,
      "step": 12880
    },
    {
      "entropy": 5.488153266906738,
      "epoch": 1.2787812624057167,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048442328478892334,
      "loss": 5.1564,
      "mean_token_accuracy": 0.16849966943264008,
      "num_tokens": 28211455.0,
      "step": 12885
    },
    {
      "entropy": 5.544581747055053,
      "epoch": 1.2792774910678841,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004844103282649128,
      "loss": 5.208,
      "mean_token_accuracy": 0.16500588059425353,
      "num_tokens": 28222599.0,
      "step": 12890
    },
    {
      "entropy": 5.551724052429199,
      "epoch": 1.2797737197300516,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004843973665479774,
      "loss": 5.1554,
      "mean_token_accuracy": 0.1657870501279831,
      "num_tokens": 28233129.0,
      "step": 12895
    },
    {
      "entropy": 5.50454158782959,
      "epoch": 1.280269948392219,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048438439963843853,
      "loss": 5.1414,
      "mean_token_accuracy": 0.1705342322587967,
      "num_tokens": 28243603.0,
      "step": 12900
    },
    {
      "entropy": 5.557707786560059,
      "epoch": 1.2807661770543866,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048437142753661775,
      "loss": 5.1665,
      "mean_token_accuracy": 0.16510963290929795,
      "num_tokens": 28254695.0,
      "step": 12905
    },
    {
      "entropy": 5.534022331237793,
      "epoch": 1.2812624057165543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843584502428367,
      "loss": 5.1404,
      "mean_token_accuracy": 0.16867119520902635,
      "num_tokens": 28266486.0,
      "step": 12910
    },
    {
      "entropy": 5.47842493057251,
      "epoch": 1.2817586343787217,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004843454677574173,
      "loss": 5.1391,
      "mean_token_accuracy": 0.17227717041969298,
      "num_tokens": 28277931.0,
      "step": 12915
    },
    {
      "entropy": 5.520468473434448,
      "epoch": 1.2822548630408892,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048433248008068123,
      "loss": 5.2657,
      "mean_token_accuracy": 0.16511921286582948,
      "num_tokens": 28288807.0,
      "step": 12920
    },
    {
      "entropy": 5.530875158309937,
      "epoch": 1.282751091703057,
      "grad_norm": 1.0,
      "learning_rate": 0.00048431948721295085,
      "loss": 5.1642,
      "mean_token_accuracy": 0.1646326258778572,
      "num_tokens": 28299577.0,
      "step": 12925
    },
    {
      "entropy": 5.549858903884887,
      "epoch": 1.2832473203652244,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004843064891545482,
      "loss": 5.179,
      "mean_token_accuracy": 0.16486040651798248,
      "num_tokens": 28310645.0,
      "step": 12930
    },
    {
      "entropy": 5.523878574371338,
      "epoch": 1.2837435490273919,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004842934859057956,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17844093590974808,
      "num_tokens": 28321634.0,
      "step": 12935
    },
    {
      "entropy": 5.5439189910888675,
      "epoch": 1.2842397776895593,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004842804774670156,
      "loss": 5.094,
      "mean_token_accuracy": 0.17429423928260804,
      "num_tokens": 28333325.0,
      "step": 12940
    },
    {
      "entropy": 5.528505134582519,
      "epoch": 1.2847360063517268,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048426746383853066,
      "loss": 5.1582,
      "mean_token_accuracy": 0.1718052878975868,
      "num_tokens": 28344163.0,
      "step": 12945
    },
    {
      "entropy": 5.554698753356933,
      "epoch": 1.2852322350138943,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048425444502066354,
      "loss": 5.2156,
      "mean_token_accuracy": 0.1697764053940773,
      "num_tokens": 28355820.0,
      "step": 12950
    },
    {
      "entropy": 5.562047243118286,
      "epoch": 1.285728463676062,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048424142101373706,
      "loss": 5.1597,
      "mean_token_accuracy": 0.1732676938176155,
      "num_tokens": 28367850.0,
      "step": 12955
    },
    {
      "entropy": 5.5454872131347654,
      "epoch": 1.2862246923382294,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048422839181807426,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1603301763534546,
      "num_tokens": 28379013.0,
      "step": 12960
    },
    {
      "entropy": 5.580005073547364,
      "epoch": 1.286720921000397,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004842153574339982,
      "loss": 5.1925,
      "mean_token_accuracy": 0.17116967737674713,
      "num_tokens": 28389649.0,
      "step": 12965
    },
    {
      "entropy": 5.451429462432861,
      "epoch": 1.2872171496625646,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048420231786183216,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18472816050052643,
      "num_tokens": 28400239.0,
      "step": 12970
    },
    {
      "entropy": 5.475468540191651,
      "epoch": 1.287713378324732,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004841892731018994,
      "loss": 5.1037,
      "mean_token_accuracy": 0.17194812446832658,
      "num_tokens": 28410504.0,
      "step": 12975
    },
    {
      "entropy": 5.501543712615967,
      "epoch": 1.2882096069868996,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004841762231545233,
      "loss": 5.138,
      "mean_token_accuracy": 0.17121327370405198,
      "num_tokens": 28422549.0,
      "step": 12980
    },
    {
      "entropy": 5.56064019203186,
      "epoch": 1.288705835649067,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048416316802002776,
      "loss": 5.1299,
      "mean_token_accuracy": 0.16979871839284896,
      "num_tokens": 28433019.0,
      "step": 12985
    },
    {
      "entropy": 5.537854957580566,
      "epoch": 1.2892020643112345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048415010769873625,
      "loss": 5.2066,
      "mean_token_accuracy": 0.16343292593955994,
      "num_tokens": 28443814.0,
      "step": 12990
    },
    {
      "entropy": 5.538273477554322,
      "epoch": 1.2896982929734022,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004841370421909728,
      "loss": 5.0778,
      "mean_token_accuracy": 0.17417672425508499,
      "num_tokens": 28453825.0,
      "step": 12995
    },
    {
      "entropy": 5.543932151794434,
      "epoch": 1.2901945216355697,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048412397149706133,
      "loss": 5.2204,
      "mean_token_accuracy": 0.16969054639339448,
      "num_tokens": 28464583.0,
      "step": 13000
    },
    {
      "entropy": 5.610096406936646,
      "epoch": 1.2906907502977372,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000484110895617326,
      "loss": 5.1988,
      "mean_token_accuracy": 0.16563909649848937,
      "num_tokens": 28475764.0,
      "step": 13005
    },
    {
      "entropy": 5.653671646118164,
      "epoch": 1.2911869789599046,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048409781455209104,
      "loss": 5.1651,
      "mean_token_accuracy": 0.16442917734384538,
      "num_tokens": 28486451.0,
      "step": 13010
    },
    {
      "entropy": 5.560187530517578,
      "epoch": 1.2916832076220723,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004840847283016809,
      "loss": 5.1948,
      "mean_token_accuracy": 0.16885197907686234,
      "num_tokens": 28498012.0,
      "step": 13015
    },
    {
      "entropy": 5.481222867965698,
      "epoch": 1.2921794362842398,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048407163686642,
      "loss": 5.151,
      "mean_token_accuracy": 0.17008824944496154,
      "num_tokens": 28509058.0,
      "step": 13020
    },
    {
      "entropy": 5.538768577575683,
      "epoch": 1.2926756649464073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000484058540246633,
      "loss": 5.1802,
      "mean_token_accuracy": 0.17426546663045883,
      "num_tokens": 28519001.0,
      "step": 13025
    },
    {
      "entropy": 5.542804193496704,
      "epoch": 1.2931718936085748,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004840454384426447,
      "loss": 5.1726,
      "mean_token_accuracy": 0.17041651010513306,
      "num_tokens": 28529606.0,
      "step": 13030
    },
    {
      "entropy": 5.541600847244263,
      "epoch": 1.2936681222707422,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004840323314547799,
      "loss": 5.1754,
      "mean_token_accuracy": 0.17077397108078002,
      "num_tokens": 28540723.0,
      "step": 13035
    },
    {
      "entropy": 5.5346386432647705,
      "epoch": 1.29416435093291,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048401921928336385,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17532050907611846,
      "num_tokens": 28549817.0,
      "step": 13040
    },
    {
      "entropy": 5.522007417678833,
      "epoch": 1.2946605795950774,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004840061019287215,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16576242744922637,
      "num_tokens": 28560519.0,
      "step": 13045
    },
    {
      "entropy": 5.520895862579346,
      "epoch": 1.2951568082572449,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048399297939117817,
      "loss": 5.1943,
      "mean_token_accuracy": 0.16789402365684508,
      "num_tokens": 28573234.0,
      "step": 13050
    },
    {
      "entropy": 5.571822214126587,
      "epoch": 1.2956530369194126,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004839798516710593,
      "loss": 5.2618,
      "mean_token_accuracy": 0.16378962993621826,
      "num_tokens": 28582954.0,
      "step": 13055
    },
    {
      "entropy": 5.509437799453735,
      "epoch": 1.29614926558158,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004839667187686904,
      "loss": 5.0307,
      "mean_token_accuracy": 0.17818064987659454,
      "num_tokens": 28593704.0,
      "step": 13060
    },
    {
      "entropy": 5.514905166625977,
      "epoch": 1.2966454942437475,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004839535806843972,
      "loss": 5.1283,
      "mean_token_accuracy": 0.16850632280111313,
      "num_tokens": 28604023.0,
      "step": 13065
    },
    {
      "entropy": 5.5849100112915036,
      "epoch": 1.297141722905915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004839404374185054,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16459398865699768,
      "num_tokens": 28614634.0,
      "step": 13070
    },
    {
      "entropy": 5.468796920776367,
      "epoch": 1.2976379515680825,
      "grad_norm": 0.953125,
      "learning_rate": 0.000483927288971341,
      "loss": 5.1128,
      "mean_token_accuracy": 0.1724123775959015,
      "num_tokens": 28624436.0,
      "step": 13075
    },
    {
      "entropy": 5.531090974807739,
      "epoch": 1.2981341802302502,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004839141353432301,
      "loss": 5.1658,
      "mean_token_accuracy": 0.1689655900001526,
      "num_tokens": 28636372.0,
      "step": 13080
    },
    {
      "entropy": 5.587186622619629,
      "epoch": 1.2986304088924177,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004839009765344987,
      "loss": 5.171,
      "mean_token_accuracy": 0.16937105655670165,
      "num_tokens": 28647495.0,
      "step": 13085
    },
    {
      "entropy": 5.533828592300415,
      "epoch": 1.2991266375545851,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048388781254547324,
      "loss": 5.1561,
      "mean_token_accuracy": 0.17131630778312684,
      "num_tokens": 28657692.0,
      "step": 13090
    },
    {
      "entropy": 5.5454336643219,
      "epoch": 1.2996228662167526,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048387464337648026,
      "loss": 5.1675,
      "mean_token_accuracy": 0.16941652595996856,
      "num_tokens": 28668580.0,
      "step": 13095
    },
    {
      "entropy": 5.696693086624146,
      "epoch": 1.3001190948789203,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048386146902784606,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1641668900847435,
      "num_tokens": 28680274.0,
      "step": 13100
    },
    {
      "entropy": 5.605729389190674,
      "epoch": 1.3006153235410878,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048384828949989757,
      "loss": 5.1935,
      "mean_token_accuracy": 0.17092992961406708,
      "num_tokens": 28691572.0,
      "step": 13105
    },
    {
      "entropy": 5.554057693481445,
      "epoch": 1.3011115522032553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048383510479296153,
      "loss": 5.2076,
      "mean_token_accuracy": 0.165431971848011,
      "num_tokens": 28702153.0,
      "step": 13110
    },
    {
      "entropy": 5.514198112487793,
      "epoch": 1.3016077808654227,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004838219149073649,
      "loss": 5.1696,
      "mean_token_accuracy": 0.17142996490001677,
      "num_tokens": 28712626.0,
      "step": 13115
    },
    {
      "entropy": 5.559290075302124,
      "epoch": 1.3021040095275902,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004838087198434347,
      "loss": 5.2051,
      "mean_token_accuracy": 0.16787269711494446,
      "num_tokens": 28723912.0,
      "step": 13120
    },
    {
      "entropy": 5.527219963073731,
      "epoch": 1.302600238189758,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004837955196014982,
      "loss": 5.1046,
      "mean_token_accuracy": 0.16919633597135544,
      "num_tokens": 28734675.0,
      "step": 13125
    },
    {
      "entropy": 5.469776678085327,
      "epoch": 1.3030964668519254,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048378231418188275,
      "loss": 5.1108,
      "mean_token_accuracy": 0.17067650556564332,
      "num_tokens": 28746470.0,
      "step": 13130
    },
    {
      "entropy": 5.4805673122406,
      "epoch": 1.3035926955140928,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048376910358491576,
      "loss": 5.1266,
      "mean_token_accuracy": 0.17402711212635041,
      "num_tokens": 28757825.0,
      "step": 13135
    },
    {
      "entropy": 5.546200513839722,
      "epoch": 1.3040889241762605,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004837558878109249,
      "loss": 5.1654,
      "mean_token_accuracy": 0.1679050073027611,
      "num_tokens": 28768305.0,
      "step": 13140
    },
    {
      "entropy": 5.5245466232299805,
      "epoch": 1.304585152838428,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004837426668602378,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1634801596403122,
      "num_tokens": 28779294.0,
      "step": 13145
    },
    {
      "entropy": 5.478153705596924,
      "epoch": 1.3050813815005955,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004837294407331825,
      "loss": 5.1401,
      "mean_token_accuracy": 0.16333397179841996,
      "num_tokens": 28789911.0,
      "step": 13150
    },
    {
      "entropy": 5.546548795700073,
      "epoch": 1.305577610162763,
      "grad_norm": 1.0,
      "learning_rate": 0.0004837162094300867,
      "loss": 5.2251,
      "mean_token_accuracy": 0.16128335893154144,
      "num_tokens": 28800705.0,
      "step": 13155
    },
    {
      "entropy": 5.484284257888794,
      "epoch": 1.3060738388249304,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004837029729512787,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17480751872062683,
      "num_tokens": 28812453.0,
      "step": 13160
    },
    {
      "entropy": 5.535871267318726,
      "epoch": 1.306570067487098,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048368973129708667,
      "loss": 5.1087,
      "mean_token_accuracy": 0.17170681655406952,
      "num_tokens": 28823257.0,
      "step": 13165
    },
    {
      "entropy": 5.470357084274292,
      "epoch": 1.3070662961492656,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048367648446783894,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17567323744297028,
      "num_tokens": 28834344.0,
      "step": 13170
    },
    {
      "entropy": 5.493726205825806,
      "epoch": 1.307562524811433,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004836632324638642,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17612750828266144,
      "num_tokens": 28844372.0,
      "step": 13175
    },
    {
      "entropy": 5.494200897216797,
      "epoch": 1.3080587534736006,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048364997528549077,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17299483418464662,
      "num_tokens": 28855384.0,
      "step": 13180
    },
    {
      "entropy": 5.490035915374756,
      "epoch": 1.3085549821357683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048363671293304763,
      "loss": 5.1937,
      "mean_token_accuracy": 0.16737659871578217,
      "num_tokens": 28865170.0,
      "step": 13185
    },
    {
      "entropy": 5.541040706634521,
      "epoch": 1.3090512107979357,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004836234454068636,
      "loss": 5.2005,
      "mean_token_accuracy": 0.16840395480394363,
      "num_tokens": 28876681.0,
      "step": 13190
    },
    {
      "entropy": 5.568867349624634,
      "epoch": 1.3095474394601032,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048361017270726764,
      "loss": 5.2031,
      "mean_token_accuracy": 0.175581693649292,
      "num_tokens": 28887033.0,
      "step": 13195
    },
    {
      "entropy": 5.545899486541748,
      "epoch": 1.3100436681222707,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004835968948345889,
      "loss": 5.1855,
      "mean_token_accuracy": 0.17130817025899886,
      "num_tokens": 28898009.0,
      "step": 13200
    },
    {
      "entropy": 5.5262704372406,
      "epoch": 1.3105398967844382,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048358361178915666,
      "loss": 5.1455,
      "mean_token_accuracy": 0.1710340455174446,
      "num_tokens": 28909261.0,
      "step": 13205
    },
    {
      "entropy": 5.545828771591187,
      "epoch": 1.3110361254466059,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004835703235713003,
      "loss": 5.1901,
      "mean_token_accuracy": 0.16538542956113816,
      "num_tokens": 28920789.0,
      "step": 13210
    },
    {
      "entropy": 5.572131776809693,
      "epoch": 1.3115323541087733,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048355703018134935,
      "loss": 5.198,
      "mean_token_accuracy": 0.1682153657078743,
      "num_tokens": 28931964.0,
      "step": 13215
    },
    {
      "entropy": 5.51666293144226,
      "epoch": 1.3120285827709408,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004835437316196335,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17420334815979005,
      "num_tokens": 28942961.0,
      "step": 13220
    },
    {
      "entropy": 5.552600383758545,
      "epoch": 1.3125248114331085,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004835304278864824,
      "loss": 5.1792,
      "mean_token_accuracy": 0.17196974754333497,
      "num_tokens": 28953449.0,
      "step": 13225
    },
    {
      "entropy": 5.532246828079224,
      "epoch": 1.313021040095276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004835171189822261,
      "loss": 5.1827,
      "mean_token_accuracy": 0.1719557151198387,
      "num_tokens": 28963261.0,
      "step": 13230
    },
    {
      "entropy": 5.498179721832275,
      "epoch": 1.3135172687574435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004835038049071945,
      "loss": 5.1932,
      "mean_token_accuracy": 0.16926682442426683,
      "num_tokens": 28973611.0,
      "step": 13235
    },
    {
      "entropy": 5.598708915710449,
      "epoch": 1.314013497419611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004834904856617178,
      "loss": 5.2282,
      "mean_token_accuracy": 0.16912354826927184,
      "num_tokens": 28983502.0,
      "step": 13240
    },
    {
      "entropy": 5.5761579990386965,
      "epoch": 1.3145097260817784,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048347716124612643,
      "loss": 5.2001,
      "mean_token_accuracy": 0.17045961320400238,
      "num_tokens": 28994586.0,
      "step": 13245
    },
    {
      "entropy": 5.5188216209411625,
      "epoch": 1.3150059547439459,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004834638316607506,
      "loss": 5.1202,
      "mean_token_accuracy": 0.17566194385290146,
      "num_tokens": 29005316.0,
      "step": 13250
    },
    {
      "entropy": 5.50579047203064,
      "epoch": 1.3155021834061136,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048345049690592105,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1772515669465065,
      "num_tokens": 29016007.0,
      "step": 13255
    },
    {
      "entropy": 5.532780361175537,
      "epoch": 1.315998412068281,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048343715698196814,
      "loss": 5.132,
      "mean_token_accuracy": 0.16415969729423524,
      "num_tokens": 29025256.0,
      "step": 13260
    },
    {
      "entropy": 5.548705577850342,
      "epoch": 1.3164946407304485,
      "grad_norm": 1.0,
      "learning_rate": 0.0004834238118892231,
      "loss": 5.1787,
      "mean_token_accuracy": 0.16985909193754195,
      "num_tokens": 29035271.0,
      "step": 13265
    },
    {
      "entropy": 5.625109624862671,
      "epoch": 1.3169908693926162,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004834104616280165,
      "loss": 5.1999,
      "mean_token_accuracy": 0.1702074646949768,
      "num_tokens": 29045151.0,
      "step": 13270
    },
    {
      "entropy": 5.5009297847747805,
      "epoch": 1.3174870980547837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004833971061986796,
      "loss": 5.1069,
      "mean_token_accuracy": 0.17399545013904572,
      "num_tokens": 29056346.0,
      "step": 13275
    },
    {
      "entropy": 5.460687303543091,
      "epoch": 1.3179833267169512,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048338374560154344,
      "loss": 5.082,
      "mean_token_accuracy": 0.17241844534873962,
      "num_tokens": 29065609.0,
      "step": 13280
    },
    {
      "entropy": 5.4931214332580565,
      "epoch": 1.3184795553791187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004833703798369395,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1735379219055176,
      "num_tokens": 29075308.0,
      "step": 13285
    },
    {
      "entropy": 5.616419982910156,
      "epoch": 1.3189757840412861,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004833570089051992,
      "loss": 5.1824,
      "mean_token_accuracy": 0.16608849614858628,
      "num_tokens": 29086488.0,
      "step": 13290
    },
    {
      "entropy": 5.485676050186157,
      "epoch": 1.3194720127034538,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048334363280665393,
      "loss": 5.1037,
      "mean_token_accuracy": 0.16841933876276016,
      "num_tokens": 29098502.0,
      "step": 13295
    },
    {
      "entropy": 5.470865631103516,
      "epoch": 1.3199682413656213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004833302515416355,
      "loss": 5.0686,
      "mean_token_accuracy": 0.17796107083559037,
      "num_tokens": 29108991.0,
      "step": 13300
    },
    {
      "entropy": 5.508537006378174,
      "epoch": 1.3204644700277888,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048331686511047584,
      "loss": 5.1648,
      "mean_token_accuracy": 0.1707780435681343,
      "num_tokens": 29119436.0,
      "step": 13305
    },
    {
      "entropy": 5.529358053207398,
      "epoch": 1.3209606986899565,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048330347351350676,
      "loss": 5.1879,
      "mean_token_accuracy": 0.16615227311849595,
      "num_tokens": 29130627.0,
      "step": 13310
    },
    {
      "entropy": 5.4674736022949215,
      "epoch": 1.321456927352124,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048329007675106036,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17261235862970353,
      "num_tokens": 29141386.0,
      "step": 13315
    },
    {
      "entropy": 5.53117265701294,
      "epoch": 1.3219531560142914,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048327667482346895,
      "loss": 5.1459,
      "mean_token_accuracy": 0.167472967505455,
      "num_tokens": 29151562.0,
      "step": 13320
    },
    {
      "entropy": 5.511061716079712,
      "epoch": 1.322449384676459,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004832632677310647,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17501041442155837,
      "num_tokens": 29161571.0,
      "step": 13325
    },
    {
      "entropy": 5.560285043716431,
      "epoch": 1.3229456133386264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048324985547418026,
      "loss": 5.1454,
      "mean_token_accuracy": 0.17272383719682693,
      "num_tokens": 29172221.0,
      "step": 13330
    },
    {
      "entropy": 5.498801422119141,
      "epoch": 1.3234418420007938,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004832364380531482,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16998904794454575,
      "num_tokens": 29182434.0,
      "step": 13335
    },
    {
      "entropy": 5.474870729446411,
      "epoch": 1.3239380706629615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048322301546830103,
      "loss": 5.0798,
      "mean_token_accuracy": 0.1778826043009758,
      "num_tokens": 29191614.0,
      "step": 13340
    },
    {
      "entropy": 5.503689098358154,
      "epoch": 1.324434299325129,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004832095877199718,
      "loss": 5.0532,
      "mean_token_accuracy": 0.17769275605678558,
      "num_tokens": 29203105.0,
      "step": 13345
    },
    {
      "entropy": 5.4954931259155275,
      "epoch": 1.3249305279872965,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048319615480849347,
      "loss": 5.1244,
      "mean_token_accuracy": 0.1754699870944023,
      "num_tokens": 29213074.0,
      "step": 13350
    },
    {
      "entropy": 5.486136102676392,
      "epoch": 1.3254267566494642,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048318271673419914,
      "loss": 5.1681,
      "mean_token_accuracy": 0.17383600026369095,
      "num_tokens": 29223421.0,
      "step": 13355
    },
    {
      "entropy": 5.498506784439087,
      "epoch": 1.3259229853116317,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048316927349742207,
      "loss": 5.0744,
      "mean_token_accuracy": 0.17757874578237534,
      "num_tokens": 29233330.0,
      "step": 13360
    },
    {
      "entropy": 5.523985195159912,
      "epoch": 1.3264192139737991,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004831558250984955,
      "loss": 5.1858,
      "mean_token_accuracy": 0.16627840399742128,
      "num_tokens": 29244679.0,
      "step": 13365
    },
    {
      "entropy": 5.527511930465698,
      "epoch": 1.3269154426359666,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048314237153775303,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17034837305545808,
      "num_tokens": 29254840.0,
      "step": 13370
    },
    {
      "entropy": 5.523643541336059,
      "epoch": 1.327411671298134,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004831289128155283,
      "loss": 5.1081,
      "mean_token_accuracy": 0.1641836829483509,
      "num_tokens": 29265794.0,
      "step": 13375
    },
    {
      "entropy": 5.572889804840088,
      "epoch": 1.3279078999603018,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048311544893215496,
      "loss": 5.14,
      "mean_token_accuracy": 0.17148406952619552,
      "num_tokens": 29275675.0,
      "step": 13380
    },
    {
      "entropy": 5.514709854125977,
      "epoch": 1.3284041286224693,
      "grad_norm": 1.015625,
      "learning_rate": 0.000483101979887967,
      "loss": 5.1891,
      "mean_token_accuracy": 0.16753035336732863,
      "num_tokens": 29286809.0,
      "step": 13385
    },
    {
      "entropy": 5.5328590869903564,
      "epoch": 1.3289003572846367,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048308850568329825,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17310334593057633,
      "num_tokens": 29297097.0,
      "step": 13390
    },
    {
      "entropy": 5.517452239990234,
      "epoch": 1.3293965859468042,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048307502631848306,
      "loss": 5.1644,
      "mean_token_accuracy": 0.16858946084976195,
      "num_tokens": 29309148.0,
      "step": 13395
    },
    {
      "entropy": 5.615398740768432,
      "epoch": 1.329892814608972,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048306154179385546,
      "loss": 5.2671,
      "mean_token_accuracy": 0.166430102288723,
      "num_tokens": 29320461.0,
      "step": 13400
    },
    {
      "entropy": 5.485873556137085,
      "epoch": 1.3303890432711394,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004830480521097501,
      "loss": 5.032,
      "mean_token_accuracy": 0.17394491732120515,
      "num_tokens": 29330442.0,
      "step": 13405
    },
    {
      "entropy": 5.421484041213989,
      "epoch": 1.3308852719333069,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004830345572665012,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1745743438601494,
      "num_tokens": 29340401.0,
      "step": 13410
    },
    {
      "entropy": 5.5289122581481935,
      "epoch": 1.3313815005954743,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048302105726444366,
      "loss": 5.1928,
      "mean_token_accuracy": 0.16846967488527298,
      "num_tokens": 29351481.0,
      "step": 13415
    },
    {
      "entropy": 5.576769495010376,
      "epoch": 1.3318777292576418,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048300755210391205,
      "loss": 5.1654,
      "mean_token_accuracy": 0.16950989067554473,
      "num_tokens": 29361642.0,
      "step": 13420
    },
    {
      "entropy": 5.511451292037964,
      "epoch": 1.3323739579198095,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004829940417852415,
      "loss": 5.1352,
      "mean_token_accuracy": 0.16299306750297546,
      "num_tokens": 29373210.0,
      "step": 13425
    },
    {
      "entropy": 5.521296691894531,
      "epoch": 1.332870186581977,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004829805263087668,
      "loss": 5.1423,
      "mean_token_accuracy": 0.17030459493398667,
      "num_tokens": 29382772.0,
      "step": 13430
    },
    {
      "entropy": 5.551665544509888,
      "epoch": 1.3333664152441445,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048296700567482324,
      "loss": 5.1573,
      "mean_token_accuracy": 0.16794985085725783,
      "num_tokens": 29393909.0,
      "step": 13435
    },
    {
      "entropy": 5.540644216537475,
      "epoch": 1.3338626439063122,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004829534798837461,
      "loss": 5.1478,
      "mean_token_accuracy": 0.17121424227952958,
      "num_tokens": 29406046.0,
      "step": 13440
    },
    {
      "entropy": 5.515796375274658,
      "epoch": 1.3343588725684796,
      "grad_norm": 0.875,
      "learning_rate": 0.0004829399489358707,
      "loss": 5.126,
      "mean_token_accuracy": 0.1778121381998062,
      "num_tokens": 29415785.0,
      "step": 13445
    },
    {
      "entropy": 5.480230665206909,
      "epoch": 1.334855101230647,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004829264128315326,
      "loss": 5.0293,
      "mean_token_accuracy": 0.17756011486053466,
      "num_tokens": 29426083.0,
      "step": 13450
    },
    {
      "entropy": 5.455508852005005,
      "epoch": 1.3353513298928146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004829128715710676,
      "loss": 5.115,
      "mean_token_accuracy": 0.17279589176177979,
      "num_tokens": 29437342.0,
      "step": 13455
    },
    {
      "entropy": 5.544042015075684,
      "epoch": 1.335847558554982,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004828993251548114,
      "loss": 5.1397,
      "mean_token_accuracy": 0.1719125047326088,
      "num_tokens": 29448973.0,
      "step": 13460
    },
    {
      "entropy": 5.537853050231933,
      "epoch": 1.3363437872171495,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004828857735830999,
      "loss": 5.1672,
      "mean_token_accuracy": 0.17549205869436263,
      "num_tokens": 29460308.0,
      "step": 13465
    },
    {
      "entropy": 5.557877492904663,
      "epoch": 1.3368400158793172,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004828722168562691,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1651998743414879,
      "num_tokens": 29470245.0,
      "step": 13470
    },
    {
      "entropy": 5.512295722961426,
      "epoch": 1.3373362445414847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004828586549746553,
      "loss": 5.159,
      "mean_token_accuracy": 0.171791772544384,
      "num_tokens": 29480072.0,
      "step": 13475
    },
    {
      "entropy": 5.4983409404754635,
      "epoch": 1.3378324732036522,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004828450879385947,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17460366636514663,
      "num_tokens": 29489724.0,
      "step": 13480
    },
    {
      "entropy": 5.52481427192688,
      "epoch": 1.3383287018658199,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004828315157484239,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17248978167772294,
      "num_tokens": 29500845.0,
      "step": 13485
    },
    {
      "entropy": 5.537868642807007,
      "epoch": 1.3388249305279873,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004828179384044793,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1708867609500885,
      "num_tokens": 29511721.0,
      "step": 13490
    },
    {
      "entropy": 5.526830673217773,
      "epoch": 1.3393211591901548,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048280435590709756,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17017228603363038,
      "num_tokens": 29522236.0,
      "step": 13495
    },
    {
      "entropy": 5.506719875335693,
      "epoch": 1.3398173878523223,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004827907682566157,
      "loss": 5.1848,
      "mean_token_accuracy": 0.1706010565161705,
      "num_tokens": 29533929.0,
      "step": 13500
    },
    {
      "entropy": 5.477182674407959,
      "epoch": 1.3403136165144898,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004827771754533704,
      "loss": 5.1134,
      "mean_token_accuracy": 0.17094102948904039,
      "num_tokens": 29544805.0,
      "step": 13505
    },
    {
      "entropy": 5.61255955696106,
      "epoch": 1.3408098451766575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048276357749769895,
      "loss": 5.2424,
      "mean_token_accuracy": 0.1604885220527649,
      "num_tokens": 29555334.0,
      "step": 13510
    },
    {
      "entropy": 5.544547986984253,
      "epoch": 1.341306073838825,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048274997438993845,
      "loss": 5.1861,
      "mean_token_accuracy": 0.1690140962600708,
      "num_tokens": 29567681.0,
      "step": 13515
    },
    {
      "entropy": 5.57725157737732,
      "epoch": 1.3418023025009924,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048273636613042626,
      "loss": 5.1913,
      "mean_token_accuracy": 0.16328493654727935,
      "num_tokens": 29579626.0,
      "step": 13520
    },
    {
      "entropy": 5.4765912055969235,
      "epoch": 1.3422985311631601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004827227527194998,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18251999020576476,
      "num_tokens": 29590312.0,
      "step": 13525
    },
    {
      "entropy": 5.523817443847657,
      "epoch": 1.3427947598253276,
      "grad_norm": 1.0,
      "learning_rate": 0.0004827091341574967,
      "loss": 5.1213,
      "mean_token_accuracy": 0.17335151731967927,
      "num_tokens": 29599846.0,
      "step": 13530
    },
    {
      "entropy": 5.510282182693482,
      "epoch": 1.343290988487495,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004826955104447546,
      "loss": 5.1385,
      "mean_token_accuracy": 0.16947751641273498,
      "num_tokens": 29610777.0,
      "step": 13535
    },
    {
      "entropy": 5.561915159225464,
      "epoch": 1.3437872171496625,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004826818815816114,
      "loss": 5.2171,
      "mean_token_accuracy": 0.16449571400880814,
      "num_tokens": 29623622.0,
      "step": 13540
    },
    {
      "entropy": 5.60267744064331,
      "epoch": 1.34428344581183,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048266824756840513,
      "loss": 5.2049,
      "mean_token_accuracy": 0.16709287315607071,
      "num_tokens": 29635390.0,
      "step": 13545
    },
    {
      "entropy": 5.584821987152099,
      "epoch": 1.3447796744739975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004826546084054737,
      "loss": 5.2319,
      "mean_token_accuracy": 0.16799997836351394,
      "num_tokens": 29645443.0,
      "step": 13550
    },
    {
      "entropy": 5.7164435386657715,
      "epoch": 1.3452759031361652,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004826409640931555,
      "loss": 5.3864,
      "mean_token_accuracy": 0.15693028941750525,
      "num_tokens": 29658027.0,
      "step": 13555
    },
    {
      "entropy": 5.494174480438232,
      "epoch": 1.3457721317983327,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048262731463178884,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1700786143541336,
      "num_tokens": 29668809.0,
      "step": 13560
    },
    {
      "entropy": 5.567457246780395,
      "epoch": 1.3462683604605001,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004826136600217121,
      "loss": 5.2371,
      "mean_token_accuracy": 0.16236980482935906,
      "num_tokens": 29680036.0,
      "step": 13565
    },
    {
      "entropy": 5.521518754959106,
      "epoch": 1.3467645891226678,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000482600000263264,
      "loss": 5.1605,
      "mean_token_accuracy": 0.16954350769519805,
      "num_tokens": 29692014.0,
      "step": 13570
    },
    {
      "entropy": 5.512815475463867,
      "epoch": 1.3472608177848353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825863353567832,
      "loss": 5.1561,
      "mean_token_accuracy": 0.1672423928976059,
      "num_tokens": 29703168.0,
      "step": 13575
    },
    {
      "entropy": 5.526180267333984,
      "epoch": 1.3477570464470028,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048257266530260865,
      "loss": 5.1684,
      "mean_token_accuracy": 0.16634728461503984,
      "num_tokens": 29715532.0,
      "step": 13580
    },
    {
      "entropy": 5.490181827545166,
      "epoch": 1.3482532751091703,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004825589901010792,
      "loss": 5.0546,
      "mean_token_accuracy": 0.167274209856987,
      "num_tokens": 29725697.0,
      "step": 13585
    },
    {
      "entropy": 5.549990701675415,
      "epoch": 1.3487495037713377,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004825453097525341,
      "loss": 5.19,
      "mean_token_accuracy": 0.16230688393115997,
      "num_tokens": 29736870.0,
      "step": 13590
    },
    {
      "entropy": 5.592596960067749,
      "epoch": 1.3492457324335054,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048253162425731253,
      "loss": 5.2707,
      "mean_token_accuracy": 0.1674803152680397,
      "num_tokens": 29747504.0,
      "step": 13595
    },
    {
      "entropy": 5.56365270614624,
      "epoch": 1.349741961095673,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004825179336157539,
      "loss": 5.1571,
      "mean_token_accuracy": 0.16669014394283294,
      "num_tokens": 29759182.0,
      "step": 13600
    },
    {
      "entropy": 5.496691131591797,
      "epoch": 1.3502381897578404,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048250423782819765,
      "loss": 5.0799,
      "mean_token_accuracy": 0.17564546465873718,
      "num_tokens": 29770430.0,
      "step": 13605
    },
    {
      "entropy": 5.543774795532227,
      "epoch": 1.350734418420008,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004824905368949834,
      "loss": 5.1864,
      "mean_token_accuracy": 0.17038919627666474,
      "num_tokens": 29782196.0,
      "step": 13610
    },
    {
      "entropy": 5.528673267364502,
      "epoch": 1.3512306470821756,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004824768308164509,
      "loss": 5.0881,
      "mean_token_accuracy": 0.17860938906669616,
      "num_tokens": 29792739.0,
      "step": 13615
    },
    {
      "entropy": 5.5504292964935305,
      "epoch": 1.351726875744343,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004824631195929402,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16704537123441696,
      "num_tokens": 29804384.0,
      "step": 13620
    },
    {
      "entropy": 5.546772241592407,
      "epoch": 1.3522231044065105,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004824494032247911,
      "loss": 5.2168,
      "mean_token_accuracy": 0.16944082230329513,
      "num_tokens": 29815817.0,
      "step": 13625
    },
    {
      "entropy": 5.547857236862183,
      "epoch": 1.352719333068678,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004824356817123439,
      "loss": 5.3048,
      "mean_token_accuracy": 0.16037378162145616,
      "num_tokens": 29827811.0,
      "step": 13630
    },
    {
      "entropy": 5.536314821243286,
      "epoch": 1.3532155617308455,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048242195505593865,
      "loss": 5.0436,
      "mean_token_accuracy": 0.17336487919092178,
      "num_tokens": 29836776.0,
      "step": 13635
    },
    {
      "entropy": 5.450568294525146,
      "epoch": 1.3537117903930131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048240822325591596,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1781477376818657,
      "num_tokens": 29847127.0,
      "step": 13640
    },
    {
      "entropy": 5.467622947692871,
      "epoch": 1.3542080190551806,
      "grad_norm": 0.875,
      "learning_rate": 0.0004823944863126162,
      "loss": 5.1382,
      "mean_token_accuracy": 0.16629139035940171,
      "num_tokens": 29859056.0,
      "step": 13645
    },
    {
      "entropy": 5.533165693283081,
      "epoch": 1.354704247717348,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004823807442263801,
      "loss": 5.1309,
      "mean_token_accuracy": 0.17545454502105712,
      "num_tokens": 29868468.0,
      "step": 13650
    },
    {
      "entropy": 5.5396342277526855,
      "epoch": 1.3552004763795158,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048236699699754845,
      "loss": 5.1458,
      "mean_token_accuracy": 0.17624143958091737,
      "num_tokens": 29878653.0,
      "step": 13655
    },
    {
      "entropy": 5.493823528289795,
      "epoch": 1.3556967050416833,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000482353244626462,
      "loss": 5.0738,
      "mean_token_accuracy": 0.17160923033952713,
      "num_tokens": 29889298.0,
      "step": 13660
    },
    {
      "entropy": 5.503366613388062,
      "epoch": 1.3561929337038507,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048233948711346196,
      "loss": 5.1517,
      "mean_token_accuracy": 0.16796502321958542,
      "num_tokens": 29900712.0,
      "step": 13665
    },
    {
      "entropy": 5.530891466140747,
      "epoch": 1.3566891623660182,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048232572445888945,
      "loss": 5.0993,
      "mean_token_accuracy": 0.17405882924795152,
      "num_tokens": 29910528.0,
      "step": 13670
    },
    {
      "entropy": 5.581868982315063,
      "epoch": 1.3571853910281857,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004823119566630857,
      "loss": 5.1531,
      "mean_token_accuracy": 0.17459705770015715,
      "num_tokens": 29920832.0,
      "step": 13675
    },
    {
      "entropy": 5.491753435134887,
      "epoch": 1.3576816196903534,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048229818372639216,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17498027086257933,
      "num_tokens": 29930692.0,
      "step": 13680
    },
    {
      "entropy": 5.58777928352356,
      "epoch": 1.3581778483525209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048228440564915026,
      "loss": 5.2501,
      "mean_token_accuracy": 0.16773310601711272,
      "num_tokens": 29941381.0,
      "step": 13685
    },
    {
      "entropy": 5.617674350738525,
      "epoch": 1.3586740770146883,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048227062243170186,
      "loss": 5.187,
      "mean_token_accuracy": 0.17110847681760788,
      "num_tokens": 29952297.0,
      "step": 13690
    },
    {
      "entropy": 5.509106588363648,
      "epoch": 1.3591703056768558,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004822568340743886,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17964437156915664,
      "num_tokens": 29963011.0,
      "step": 13695
    },
    {
      "entropy": 5.48095932006836,
      "epoch": 1.3596665343390235,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004822430405775525,
      "loss": 5.1318,
      "mean_token_accuracy": 0.17092402875423432,
      "num_tokens": 29975575.0,
      "step": 13700
    },
    {
      "entropy": 5.573531007766723,
      "epoch": 1.360162763001191,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004822292419415355,
      "loss": 5.2445,
      "mean_token_accuracy": 0.16997918486595154,
      "num_tokens": 29987088.0,
      "step": 13705
    },
    {
      "entropy": 5.485914707183838,
      "epoch": 1.3606589916633585,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048221543816667984,
      "loss": 5.0283,
      "mean_token_accuracy": 0.17769981920719147,
      "num_tokens": 29996471.0,
      "step": 13710
    },
    {
      "entropy": 5.558653736114502,
      "epoch": 1.361155220325526,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048220162925332786,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1676446408033371,
      "num_tokens": 30007342.0,
      "step": 13715
    },
    {
      "entropy": 5.5465394973754885,
      "epoch": 1.3616514489876934,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000482187815201822,
      "loss": 5.119,
      "mean_token_accuracy": 0.1724363699555397,
      "num_tokens": 30018043.0,
      "step": 13720
    },
    {
      "entropy": 5.477905702590943,
      "epoch": 1.362147677649861,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048217399601250464,
      "loss": 5.0884,
      "mean_token_accuracy": 0.17167914658784866,
      "num_tokens": 30029033.0,
      "step": 13725
    },
    {
      "entropy": 5.510889053344727,
      "epoch": 1.3626439063120286,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004821601716857187,
      "loss": 5.1401,
      "mean_token_accuracy": 0.17578934878110886,
      "num_tokens": 30039460.0,
      "step": 13730
    },
    {
      "entropy": 5.521335458755493,
      "epoch": 1.363140134974196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048214634222180686,
      "loss": 5.1239,
      "mean_token_accuracy": 0.17328353226184845,
      "num_tokens": 30049337.0,
      "step": 13735
    },
    {
      "entropy": 5.524208307266235,
      "epoch": 1.3636363636363638,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004821325076211121,
      "loss": 5.1515,
      "mean_token_accuracy": 0.17382630258798598,
      "num_tokens": 30060852.0,
      "step": 13740
    },
    {
      "entropy": 5.53268666267395,
      "epoch": 1.3641325922985312,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048211866788397753,
      "loss": 5.1392,
      "mean_token_accuracy": 0.17411574572324753,
      "num_tokens": 30071202.0,
      "step": 13745
    },
    {
      "entropy": 5.496741676330567,
      "epoch": 1.3646288209606987,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004821048230107462,
      "loss": 5.1461,
      "mean_token_accuracy": 0.1678459659218788,
      "num_tokens": 30082523.0,
      "step": 13750
    },
    {
      "entropy": 5.533251476287842,
      "epoch": 1.3651250496228662,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004820909730017616,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1741952583193779,
      "num_tokens": 30093521.0,
      "step": 13755
    },
    {
      "entropy": 5.57225022315979,
      "epoch": 1.3656212782850337,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048207711785736707,
      "loss": 5.166,
      "mean_token_accuracy": 0.1680004522204399,
      "num_tokens": 30104085.0,
      "step": 13760
    },
    {
      "entropy": 5.535334682464599,
      "epoch": 1.3661175069472011,
      "grad_norm": 1.0,
      "learning_rate": 0.00048206325757790624,
      "loss": 5.0701,
      "mean_token_accuracy": 0.17707088440656663,
      "num_tokens": 30114566.0,
      "step": 13765
    },
    {
      "entropy": 5.553226804733276,
      "epoch": 1.3666137356093688,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048204939216372285,
      "loss": 5.1806,
      "mean_token_accuracy": 0.1680713638663292,
      "num_tokens": 30124643.0,
      "step": 13770
    },
    {
      "entropy": 5.5201417922973635,
      "epoch": 1.3671099642715363,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004820355216151607,
      "loss": 5.1342,
      "mean_token_accuracy": 0.16902111023664473,
      "num_tokens": 30135657.0,
      "step": 13775
    },
    {
      "entropy": 5.519287347793579,
      "epoch": 1.3676061929337038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048202164593256363,
      "loss": 5.098,
      "mean_token_accuracy": 0.171302430331707,
      "num_tokens": 30146751.0,
      "step": 13780
    },
    {
      "entropy": 5.515567684173584,
      "epoch": 1.3681024215958715,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048200776511627585,
      "loss": 5.1993,
      "mean_token_accuracy": 0.1614388644695282,
      "num_tokens": 30157854.0,
      "step": 13785
    },
    {
      "entropy": 5.561240816116333,
      "epoch": 1.368598650258039,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004819938791666416,
      "loss": 5.181,
      "mean_token_accuracy": 0.17131547778844833,
      "num_tokens": 30169726.0,
      "step": 13790
    },
    {
      "entropy": 5.555774831771851,
      "epoch": 1.3690948789202064,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048197998808400524,
      "loss": 5.2218,
      "mean_token_accuracy": 0.17072439342737197,
      "num_tokens": 30182006.0,
      "step": 13795
    },
    {
      "entropy": 5.566875267028808,
      "epoch": 1.369591107582374,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004819660918687111,
      "loss": 5.1765,
      "mean_token_accuracy": 0.16438425183296204,
      "num_tokens": 30192975.0,
      "step": 13800
    },
    {
      "entropy": 5.612747526168823,
      "epoch": 1.3700873362445414,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004819521905211039,
      "loss": 5.2556,
      "mean_token_accuracy": 0.16691527962684632,
      "num_tokens": 30205742.0,
      "step": 13805
    },
    {
      "entropy": 5.537242794036866,
      "epoch": 1.370583564906709,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048193828404152826,
      "loss": 5.1343,
      "mean_token_accuracy": 0.17988501936197282,
      "num_tokens": 30215850.0,
      "step": 13810
    },
    {
      "entropy": 5.502561807632446,
      "epoch": 1.3710797935688765,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004819243724303291,
      "loss": 5.1019,
      "mean_token_accuracy": 0.16895131766796112,
      "num_tokens": 30227206.0,
      "step": 13815
    },
    {
      "entropy": 5.477277755737305,
      "epoch": 1.371576022231044,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048191045568785144,
      "loss": 5.076,
      "mean_token_accuracy": 0.17830151319503784,
      "num_tokens": 30238946.0,
      "step": 13820
    },
    {
      "entropy": 5.51090931892395,
      "epoch": 1.3720722508932117,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048189653381444024,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16935242414474488,
      "num_tokens": 30249381.0,
      "step": 13825
    },
    {
      "entropy": 5.514531421661377,
      "epoch": 1.3725684795553792,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004818826068104409,
      "loss": 5.1882,
      "mean_token_accuracy": 0.17324302196502686,
      "num_tokens": 30259877.0,
      "step": 13830
    },
    {
      "entropy": 5.5309834480285645,
      "epoch": 1.3730647082175467,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004818686746761987,
      "loss": 5.1946,
      "mean_token_accuracy": 0.16190171539783477,
      "num_tokens": 30270401.0,
      "step": 13835
    },
    {
      "entropy": 5.615086507797241,
      "epoch": 1.3735609368797141,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004818547374120591,
      "loss": 5.0691,
      "mean_token_accuracy": 0.18310218453407287,
      "num_tokens": 30281173.0,
      "step": 13840
    },
    {
      "entropy": 5.482473754882813,
      "epoch": 1.3740571655418816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004818407950183678,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17558681219816208,
      "num_tokens": 30291792.0,
      "step": 13845
    },
    {
      "entropy": 5.55519700050354,
      "epoch": 1.374553394204049,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004818268474954704,
      "loss": 5.1812,
      "mean_token_accuracy": 0.17166556119918824,
      "num_tokens": 30302830.0,
      "step": 13850
    },
    {
      "entropy": 5.5665501117706295,
      "epoch": 1.3750496228662168,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004818128948437129,
      "loss": 5.1385,
      "mean_token_accuracy": 0.16834837198257446,
      "num_tokens": 30314888.0,
      "step": 13855
    },
    {
      "entropy": 5.542465734481811,
      "epoch": 1.3755458515283843,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004817989370634412,
      "loss": 5.1397,
      "mean_token_accuracy": 0.17269478589296341,
      "num_tokens": 30325324.0,
      "step": 13860
    },
    {
      "entropy": 5.5032360553741455,
      "epoch": 1.3760420801905517,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048178497415500153,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17447879761457444,
      "num_tokens": 30336012.0,
      "step": 13865
    },
    {
      "entropy": 5.6025666236877445,
      "epoch": 1.3765383088527194,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004817710061187401,
      "loss": 5.294,
      "mean_token_accuracy": 0.1592933475971222,
      "num_tokens": 30346983.0,
      "step": 13870
    },
    {
      "entropy": 5.556367492675781,
      "epoch": 1.377034537514887,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048175703295500323,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17695105969905853,
      "num_tokens": 30357522.0,
      "step": 13875
    },
    {
      "entropy": 5.532705402374267,
      "epoch": 1.3775307661770544,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048174305466413755,
      "loss": 5.1212,
      "mean_token_accuracy": 0.168430894613266,
      "num_tokens": 30368778.0,
      "step": 13880
    },
    {
      "entropy": 5.525425052642822,
      "epoch": 1.3780269948392219,
      "grad_norm": 1.0,
      "learning_rate": 0.00048172907124648947,
      "loss": 5.1577,
      "mean_token_accuracy": 0.1739433452486992,
      "num_tokens": 30379651.0,
      "step": 13885
    },
    {
      "entropy": 5.5520384311676025,
      "epoch": 1.3785232235013893,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004817150827024059,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16684924960136413,
      "num_tokens": 30391288.0,
      "step": 13890
    },
    {
      "entropy": 5.53673152923584,
      "epoch": 1.379019452163557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004817010890322338,
      "loss": 5.1447,
      "mean_token_accuracy": 0.16979329586029052,
      "num_tokens": 30400672.0,
      "step": 13895
    },
    {
      "entropy": 5.5540846824646,
      "epoch": 1.3795156808257245,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048168709023632006,
      "loss": 5.2262,
      "mean_token_accuracy": 0.172475603222847,
      "num_tokens": 30411953.0,
      "step": 13900
    },
    {
      "entropy": 5.600684452056885,
      "epoch": 1.380011909487892,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048167308631501184,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16767453402280807,
      "num_tokens": 30422813.0,
      "step": 13905
    },
    {
      "entropy": 5.4950345993042,
      "epoch": 1.3805081381500597,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048165907726865643,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1735315278172493,
      "num_tokens": 30434088.0,
      "step": 13910
    },
    {
      "entropy": 5.574574184417725,
      "epoch": 1.3810043668122272,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004816450630976012,
      "loss": 5.2437,
      "mean_token_accuracy": 0.16598855778574945,
      "num_tokens": 30446167.0,
      "step": 13915
    },
    {
      "entropy": 5.532061767578125,
      "epoch": 1.3815005954743946,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004816310438021937,
      "loss": 5.0275,
      "mean_token_accuracy": 0.17829238325357438,
      "num_tokens": 30456713.0,
      "step": 13920
    },
    {
      "entropy": 5.541204357147217,
      "epoch": 1.381996824136562,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048161701938278156,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17077284604310988,
      "num_tokens": 30468183.0,
      "step": 13925
    },
    {
      "entropy": 5.51575083732605,
      "epoch": 1.3824930527987296,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048160298983971256,
      "loss": 5.0935,
      "mean_token_accuracy": 0.175765261054039,
      "num_tokens": 30478545.0,
      "step": 13930
    },
    {
      "entropy": 5.563014888763428,
      "epoch": 1.382989281460897,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004815889551733346,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1648734003305435,
      "num_tokens": 30490843.0,
      "step": 13935
    },
    {
      "entropy": 5.585333299636841,
      "epoch": 1.3834855101230648,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048157491538399573,
      "loss": 5.1967,
      "mean_token_accuracy": 0.1735122710466385,
      "num_tokens": 30502301.0,
      "step": 13940
    },
    {
      "entropy": 5.568125200271607,
      "epoch": 1.3839817387852322,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048156087047204407,
      "loss": 5.2418,
      "mean_token_accuracy": 0.16805353462696077,
      "num_tokens": 30513842.0,
      "step": 13945
    },
    {
      "entropy": 5.655458068847656,
      "epoch": 1.3844779674473997,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004815468204378279,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16372652649879454,
      "num_tokens": 30526867.0,
      "step": 13950
    },
    {
      "entropy": 5.532332849502564,
      "epoch": 1.3849741961095674,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004815327652816957,
      "loss": 5.2346,
      "mean_token_accuracy": 0.16646682918071748,
      "num_tokens": 30539493.0,
      "step": 13955
    },
    {
      "entropy": 5.501109504699707,
      "epoch": 1.3854704247717349,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004815187050039959,
      "loss": 5.1683,
      "mean_token_accuracy": 0.1747711941599846,
      "num_tokens": 30550978.0,
      "step": 13960
    },
    {
      "entropy": 5.596608304977417,
      "epoch": 1.3859666534339024,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048150463960507724,
      "loss": 5.1476,
      "mean_token_accuracy": 0.1720774620771408,
      "num_tokens": 30561873.0,
      "step": 13965
    },
    {
      "entropy": 5.569955825805664,
      "epoch": 1.3864628820960698,
      "grad_norm": 1.0,
      "learning_rate": 0.00048149056908528855,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17407867908477784,
      "num_tokens": 30573053.0,
      "step": 13970
    },
    {
      "entropy": 5.483378124237061,
      "epoch": 1.3869591107582373,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004814764934449786,
      "loss": 5.0718,
      "mean_token_accuracy": 0.17751435041427613,
      "num_tokens": 30583455.0,
      "step": 13975
    },
    {
      "entropy": 5.473579454421997,
      "epoch": 1.387455339420405,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048146241268449665,
      "loss": 5.144,
      "mean_token_accuracy": 0.17275312393903733,
      "num_tokens": 30593908.0,
      "step": 13980
    },
    {
      "entropy": 5.549531030654907,
      "epoch": 1.3879515680825725,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004814483268041917,
      "loss": 5.1158,
      "mean_token_accuracy": 0.1676690846681595,
      "num_tokens": 30604870.0,
      "step": 13985
    },
    {
      "entropy": 5.527051305770874,
      "epoch": 1.38844779674474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004814342358044131,
      "loss": 5.0762,
      "mean_token_accuracy": 0.17296316027641295,
      "num_tokens": 30615048.0,
      "step": 13990
    },
    {
      "entropy": 5.5568427562713625,
      "epoch": 1.3889440254069074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048142013968551026,
      "loss": 5.1868,
      "mean_token_accuracy": 0.17081286907196044,
      "num_tokens": 30626652.0,
      "step": 13995
    },
    {
      "entropy": 5.487878656387329,
      "epoch": 1.3894402540690751,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048140603844783273,
      "loss": 5.1408,
      "mean_token_accuracy": 0.1716690942645073,
      "num_tokens": 30637890.0,
      "step": 14000
    },
    {
      "entropy": 5.554536056518555,
      "epoch": 1.3899364827312426,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004813919320917303,
      "loss": 5.1574,
      "mean_token_accuracy": 0.17045267224311828,
      "num_tokens": 30649267.0,
      "step": 14005
    },
    {
      "entropy": 5.605321645736694,
      "epoch": 1.39043271139341,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048137782061755265,
      "loss": 5.1447,
      "mean_token_accuracy": 0.1738915503025055,
      "num_tokens": 30661003.0,
      "step": 14010
    },
    {
      "entropy": 5.516326141357422,
      "epoch": 1.3909289400555775,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048136370402564974,
      "loss": 5.1229,
      "mean_token_accuracy": 0.17721529603004454,
      "num_tokens": 30671895.0,
      "step": 14015
    },
    {
      "entropy": 5.5314960956573485,
      "epoch": 1.391425168717745,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004813495823163716,
      "loss": 5.1151,
      "mean_token_accuracy": 0.1696929857134819,
      "num_tokens": 30683246.0,
      "step": 14020
    },
    {
      "entropy": 5.631370735168457,
      "epoch": 1.3919213973799127,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004813354554900685,
      "loss": 5.2609,
      "mean_token_accuracy": 0.17019764631986617,
      "num_tokens": 30694078.0,
      "step": 14025
    },
    {
      "entropy": 5.528982305526734,
      "epoch": 1.3924176260420802,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004813213235470907,
      "loss": 5.182,
      "mean_token_accuracy": 0.16784007996320724,
      "num_tokens": 30705270.0,
      "step": 14030
    },
    {
      "entropy": 5.523030090332031,
      "epoch": 1.3929138547042477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004813071864877887,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1687418892979622,
      "num_tokens": 30716482.0,
      "step": 14035
    },
    {
      "entropy": 5.544445562362671,
      "epoch": 1.3934100833664154,
      "grad_norm": 0.953125,
      "learning_rate": 0.000481293044312513,
      "loss": 5.164,
      "mean_token_accuracy": 0.16594312638044356,
      "num_tokens": 30727257.0,
      "step": 14040
    },
    {
      "entropy": 5.507503938674927,
      "epoch": 1.3939063120285828,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048127889702161435,
      "loss": 5.1379,
      "mean_token_accuracy": 0.17083055526018143,
      "num_tokens": 30739100.0,
      "step": 14045
    },
    {
      "entropy": 5.525509595870972,
      "epoch": 1.3944025406907503,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004812647446154435,
      "loss": 5.0392,
      "mean_token_accuracy": 0.1764340505003929,
      "num_tokens": 30749002.0,
      "step": 14050
    },
    {
      "entropy": 5.537301445007325,
      "epoch": 1.3948987693529178,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004812505870943515,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17912777215242387,
      "num_tokens": 30760056.0,
      "step": 14055
    },
    {
      "entropy": 5.520949125289917,
      "epoch": 1.3953949980150853,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004812364244586893,
      "loss": 5.169,
      "mean_token_accuracy": 0.15960632860660554,
      "num_tokens": 30771409.0,
      "step": 14060
    },
    {
      "entropy": 5.456146621704102,
      "epoch": 1.395891226677253,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004812222567088082,
      "loss": 5.1428,
      "mean_token_accuracy": 0.1718702033162117,
      "num_tokens": 30782976.0,
      "step": 14065
    },
    {
      "entropy": 5.501482439041138,
      "epoch": 1.3963874553394204,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004812080838450596,
      "loss": 5.0718,
      "mean_token_accuracy": 0.174846151471138,
      "num_tokens": 30794745.0,
      "step": 14070
    },
    {
      "entropy": 5.55799617767334,
      "epoch": 1.396883684001588,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004811939058677947,
      "loss": 5.1301,
      "mean_token_accuracy": 0.17275519222021102,
      "num_tokens": 30807034.0,
      "step": 14075
    },
    {
      "entropy": 5.581147336959839,
      "epoch": 1.3973799126637554,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004811797227773654,
      "loss": 5.1669,
      "mean_token_accuracy": 0.16470855325460435,
      "num_tokens": 30818117.0,
      "step": 14080
    },
    {
      "entropy": 5.478305625915527,
      "epoch": 1.397876141325923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048116553457412317,
      "loss": 5.0583,
      "mean_token_accuracy": 0.17510392069816588,
      "num_tokens": 30829222.0,
      "step": 14085
    },
    {
      "entropy": 5.506008100509644,
      "epoch": 1.3983723699880906,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048115134125841993,
      "loss": 5.1216,
      "mean_token_accuracy": 0.16889320462942123,
      "num_tokens": 30839766.0,
      "step": 14090
    },
    {
      "entropy": 5.498178434371948,
      "epoch": 1.398868598650258,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004811371428306076,
      "loss": 5.1482,
      "mean_token_accuracy": 0.17122518718242646,
      "num_tokens": 30850381.0,
      "step": 14095
    },
    {
      "entropy": 5.494927263259887,
      "epoch": 1.3993648273124255,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004811229392910384,
      "loss": 5.1146,
      "mean_token_accuracy": 0.16960859745740892,
      "num_tokens": 30860851.0,
      "step": 14100
    },
    {
      "entropy": 5.548106288909912,
      "epoch": 1.399861055974593,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004811087306400644,
      "loss": 5.1067,
      "mean_token_accuracy": 0.16686815172433853,
      "num_tokens": 30871246.0,
      "step": 14105
    },
    {
      "entropy": 5.535901737213135,
      "epoch": 1.4003572846367607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048109451687803804,
      "loss": 5.1778,
      "mean_token_accuracy": 0.1746618002653122,
      "num_tokens": 30883092.0,
      "step": 14110
    },
    {
      "entropy": 5.583242893218994,
      "epoch": 1.4008535132989282,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004810802980053117,
      "loss": 5.1858,
      "mean_token_accuracy": 0.17568325996398926,
      "num_tokens": 30895848.0,
      "step": 14115
    },
    {
      "entropy": 5.563458728790283,
      "epoch": 1.4013497419610956,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048106607402223815,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16984723657369613,
      "num_tokens": 30907085.0,
      "step": 14120
    },
    {
      "entropy": 5.542504215240479,
      "epoch": 1.4018459706232633,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048105184492916995,
      "loss": 5.1515,
      "mean_token_accuracy": 0.16743867397308348,
      "num_tokens": 30917735.0,
      "step": 14125
    },
    {
      "entropy": 5.60797438621521,
      "epoch": 1.4023421992854308,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048103761072645993,
      "loss": 5.0919,
      "mean_token_accuracy": 0.17118586599826813,
      "num_tokens": 30927736.0,
      "step": 14130
    },
    {
      "entropy": 5.552013206481933,
      "epoch": 1.4028384279475983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048102337141446116,
      "loss": 5.128,
      "mean_token_accuracy": 0.1735871195793152,
      "num_tokens": 30938224.0,
      "step": 14135
    },
    {
      "entropy": 5.486589860916138,
      "epoch": 1.4033346566097658,
      "grad_norm": 1.0,
      "learning_rate": 0.00048100912699352666,
      "loss": 5.0923,
      "mean_token_accuracy": 0.17759290486574172,
      "num_tokens": 30948537.0,
      "step": 14140
    },
    {
      "entropy": 5.505899906158447,
      "epoch": 1.4038308852719332,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004809948774640098,
      "loss": 5.1184,
      "mean_token_accuracy": 0.1655098244547844,
      "num_tokens": 30959545.0,
      "step": 14145
    },
    {
      "entropy": 5.503247928619385,
      "epoch": 1.4043271139341007,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004809806228262639,
      "loss": 5.1676,
      "mean_token_accuracy": 0.17667951583862304,
      "num_tokens": 30970026.0,
      "step": 14150
    },
    {
      "entropy": 5.547444772720337,
      "epoch": 1.4048233425962684,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048096636308064236,
      "loss": 5.1256,
      "mean_token_accuracy": 0.16523968130350114,
      "num_tokens": 30981191.0,
      "step": 14155
    },
    {
      "entropy": 5.488926649093628,
      "epoch": 1.4053195712584359,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004809520982274989,
      "loss": 5.09,
      "mean_token_accuracy": 0.1786476731300354,
      "num_tokens": 30992427.0,
      "step": 14160
    },
    {
      "entropy": 5.538772535324097,
      "epoch": 1.4058157999206033,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048093782826718706,
      "loss": 5.2315,
      "mean_token_accuracy": 0.16562466472387313,
      "num_tokens": 31004972.0,
      "step": 14165
    },
    {
      "entropy": 5.534318113327027,
      "epoch": 1.406312028582771,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048092355320006086,
      "loss": 5.1718,
      "mean_token_accuracy": 0.1665011689066887,
      "num_tokens": 31015043.0,
      "step": 14170
    },
    {
      "entropy": 5.511624193191528,
      "epoch": 1.4068082572449385,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004809092730264744,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17948632538318635,
      "num_tokens": 31026241.0,
      "step": 14175
    },
    {
      "entropy": 5.509392833709716,
      "epoch": 1.407304485907106,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048089498774678155,
      "loss": 5.0962,
      "mean_token_accuracy": 0.1711428478360176,
      "num_tokens": 31036548.0,
      "step": 14180
    },
    {
      "entropy": 5.533528423309326,
      "epoch": 1.4078007145692735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048088069736133664,
      "loss": 5.0911,
      "mean_token_accuracy": 0.16683247834444045,
      "num_tokens": 31046948.0,
      "step": 14185
    },
    {
      "entropy": 5.5596293926239015,
      "epoch": 1.408296943231441,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004808664018704941,
      "loss": 5.1588,
      "mean_token_accuracy": 0.17533683329820632,
      "num_tokens": 31057679.0,
      "step": 14190
    },
    {
      "entropy": 5.534587240219116,
      "epoch": 1.4087931718936086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004808521012746085,
      "loss": 5.1468,
      "mean_token_accuracy": 0.17012311965227128,
      "num_tokens": 31068108.0,
      "step": 14195
    },
    {
      "entropy": 5.530222702026367,
      "epoch": 1.4092894005557761,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048083779557403424,
      "loss": 5.119,
      "mean_token_accuracy": 0.1764849156141281,
      "num_tokens": 31079229.0,
      "step": 14200
    },
    {
      "entropy": 5.511088943481445,
      "epoch": 1.4097856292179436,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004808234847691262,
      "loss": 5.138,
      "mean_token_accuracy": 0.16866187900304794,
      "num_tokens": 31090581.0,
      "step": 14205
    },
    {
      "entropy": 5.594992160797119,
      "epoch": 1.4102818578801113,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004808091688602393,
      "loss": 5.1903,
      "mean_token_accuracy": 0.1654011994600296,
      "num_tokens": 31102904.0,
      "step": 14210
    },
    {
      "entropy": 5.658291625976562,
      "epoch": 1.4107780865422788,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048079484784772854,
      "loss": 5.1666,
      "mean_token_accuracy": 0.1697390556335449,
      "num_tokens": 31115123.0,
      "step": 14215
    },
    {
      "entropy": 5.533841943740844,
      "epoch": 1.4112743152044462,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048078052173194886,
      "loss": 5.191,
      "mean_token_accuracy": 0.17303943037986755,
      "num_tokens": 31126374.0,
      "step": 14220
    },
    {
      "entropy": 5.5108418464660645,
      "epoch": 1.4117705438666137,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004807661905132557,
      "loss": 5.1693,
      "mean_token_accuracy": 0.17447694689035415,
      "num_tokens": 31137699.0,
      "step": 14225
    },
    {
      "entropy": 5.557755041122436,
      "epoch": 1.4122667725287812,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004807518541920045,
      "loss": 5.149,
      "mean_token_accuracy": 0.16915134936571122,
      "num_tokens": 31149202.0,
      "step": 14230
    },
    {
      "entropy": 5.544007205963135,
      "epoch": 1.4127630011909487,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048073751276855056,
      "loss": 5.1594,
      "mean_token_accuracy": 0.1740089923143387,
      "num_tokens": 31162624.0,
      "step": 14235
    },
    {
      "entropy": 5.488327217102051,
      "epoch": 1.4132592298531164,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004807231662432497,
      "loss": 5.0744,
      "mean_token_accuracy": 0.1746911808848381,
      "num_tokens": 31172604.0,
      "step": 14240
    },
    {
      "entropy": 5.591342973709106,
      "epoch": 1.4137554585152838,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004807088146164575,
      "loss": 5.1864,
      "mean_token_accuracy": 0.17179159820079803,
      "num_tokens": 31181899.0,
      "step": 14245
    },
    {
      "entropy": 5.536600637435913,
      "epoch": 1.4142516871774513,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004806944578885301,
      "loss": 5.1698,
      "mean_token_accuracy": 0.16746268570423126,
      "num_tokens": 31193110.0,
      "step": 14250
    },
    {
      "entropy": 5.498945665359497,
      "epoch": 1.414747915839619,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048068009605982333,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17832186371088027,
      "num_tokens": 31202901.0,
      "step": 14255
    },
    {
      "entropy": 5.524302387237549,
      "epoch": 1.4152441445017865,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048066572913069334,
      "loss": 5.177,
      "mean_token_accuracy": 0.1741760790348053,
      "num_tokens": 31213371.0,
      "step": 14260
    },
    {
      "entropy": 5.593499231338501,
      "epoch": 1.415740373163954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004806513571014965,
      "loss": 5.1331,
      "mean_token_accuracy": 0.173665913939476,
      "num_tokens": 31224043.0,
      "step": 14265
    },
    {
      "entropy": 5.582223749160766,
      "epoch": 1.4162366018261214,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004806369799725891,
      "loss": 5.1952,
      "mean_token_accuracy": 0.16848590672016145,
      "num_tokens": 31236459.0,
      "step": 14270
    },
    {
      "entropy": 5.511090183258057,
      "epoch": 1.416732830488289,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004806225977443277,
      "loss": 5.0917,
      "mean_token_accuracy": 0.1703210011124611,
      "num_tokens": 31247669.0,
      "step": 14275
    },
    {
      "entropy": 5.52247166633606,
      "epoch": 1.4172290591504566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004806082104170689,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1730809897184372,
      "num_tokens": 31258886.0,
      "step": 14280
    },
    {
      "entropy": 5.606154298782348,
      "epoch": 1.417725287812624,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048059381799116967,
      "loss": 5.2412,
      "mean_token_accuracy": 0.17074717879295348,
      "num_tokens": 31271798.0,
      "step": 14285
    },
    {
      "entropy": 5.567326354980469,
      "epoch": 1.4182215164747916,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004805794204669867,
      "loss": 5.1491,
      "mean_token_accuracy": 0.17151427567005156,
      "num_tokens": 31283075.0,
      "step": 14290
    },
    {
      "entropy": 5.581582069396973,
      "epoch": 1.418717745136959,
      "grad_norm": 1.0,
      "learning_rate": 0.00048056501784487707,
      "loss": 5.1497,
      "mean_token_accuracy": 0.17057183533906936,
      "num_tokens": 31292791.0,
      "step": 14295
    },
    {
      "entropy": 5.543632841110229,
      "epoch": 1.4192139737991267,
      "grad_norm": 0.890625,
      "learning_rate": 0.000480550610125198,
      "loss": 5.1737,
      "mean_token_accuracy": 0.1698571503162384,
      "num_tokens": 31302803.0,
      "step": 14300
    },
    {
      "entropy": 5.521845006942749,
      "epoch": 1.4197102024612942,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004805361973083066,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1769335612654686,
      "num_tokens": 31314620.0,
      "step": 14305
    },
    {
      "entropy": 5.545644664764405,
      "epoch": 1.4202064311234617,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004805217793945605,
      "loss": 5.124,
      "mean_token_accuracy": 0.1670493647456169,
      "num_tokens": 31324861.0,
      "step": 14310
    },
    {
      "entropy": 5.554840898513794,
      "epoch": 1.4207026597856292,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004805073563843171,
      "loss": 5.124,
      "mean_token_accuracy": 0.1709902822971344,
      "num_tokens": 31337264.0,
      "step": 14315
    },
    {
      "entropy": 5.506866216659546,
      "epoch": 1.4211988884477966,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048049292827793413,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17300688475370407,
      "num_tokens": 31347584.0,
      "step": 14320
    },
    {
      "entropy": 5.610779476165772,
      "epoch": 1.4216951171099643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004804784950757693,
      "loss": 5.2625,
      "mean_token_accuracy": 0.17245173454284668,
      "num_tokens": 31358871.0,
      "step": 14325
    },
    {
      "entropy": 5.626703071594238,
      "epoch": 1.4221913457721318,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048046405677818057,
      "loss": 5.2429,
      "mean_token_accuracy": 0.16590566188097,
      "num_tokens": 31371088.0,
      "step": 14330
    },
    {
      "entropy": 5.5352006435394285,
      "epoch": 1.4226875744342993,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048044961338552595,
      "loss": 5.1202,
      "mean_token_accuracy": 0.1746457576751709,
      "num_tokens": 31382608.0,
      "step": 14335
    },
    {
      "entropy": 5.474417591094971,
      "epoch": 1.423183803096467,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004804351648981636,
      "loss": 5.0573,
      "mean_token_accuracy": 0.17789312154054643,
      "num_tokens": 31394098.0,
      "step": 14340
    },
    {
      "entropy": 5.634558582305909,
      "epoch": 1.4236800317586344,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004804207113164519,
      "loss": 5.208,
      "mean_token_accuracy": 0.16504755914211272,
      "num_tokens": 31403927.0,
      "step": 14345
    },
    {
      "entropy": 5.535257959365845,
      "epoch": 1.424176260420802,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048040625264074915,
      "loss": 5.1366,
      "mean_token_accuracy": 0.17728502452373504,
      "num_tokens": 31413922.0,
      "step": 14350
    },
    {
      "entropy": 5.6522280216217045,
      "epoch": 1.4246724890829694,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048039178887141404,
      "loss": 5.303,
      "mean_token_accuracy": 0.15894700288772584,
      "num_tokens": 31425020.0,
      "step": 14355
    },
    {
      "entropy": 5.506766223907471,
      "epoch": 1.4251687177451369,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048037732000880503,
      "loss": 5.077,
      "mean_token_accuracy": 0.17465019971132278,
      "num_tokens": 31435763.0,
      "step": 14360
    },
    {
      "entropy": 5.550914478302002,
      "epoch": 1.4256649464073046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048036284605328106,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17789830565452575,
      "num_tokens": 31445112.0,
      "step": 14365
    },
    {
      "entropy": 5.588475227355957,
      "epoch": 1.426161175069472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004803483670052011,
      "loss": 5.1785,
      "mean_token_accuracy": 0.16956028640270232,
      "num_tokens": 31456380.0,
      "step": 14370
    },
    {
      "entropy": 5.5639513492584225,
      "epoch": 1.4266574037316395,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000480333882864924,
      "loss": 5.2307,
      "mean_token_accuracy": 0.16576272994279861,
      "num_tokens": 31468764.0,
      "step": 14375
    },
    {
      "entropy": 5.431308269500732,
      "epoch": 1.427153632393807,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004803193936328092,
      "loss": 4.9744,
      "mean_token_accuracy": 0.17800817340612413,
      "num_tokens": 31478982.0,
      "step": 14380
    },
    {
      "entropy": 5.590399074554443,
      "epoch": 1.4276498610559747,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004803048993092158,
      "loss": 5.2568,
      "mean_token_accuracy": 0.16126177310943604,
      "num_tokens": 31489615.0,
      "step": 14385
    },
    {
      "entropy": 5.578668832778931,
      "epoch": 1.4281460897181422,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048029039989450317,
      "loss": 5.1043,
      "mean_token_accuracy": 0.17662750333547592,
      "num_tokens": 31500688.0,
      "step": 14390
    },
    {
      "entropy": 5.482921838760376,
      "epoch": 1.4286423183803096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048027589538903115,
      "loss": 5.0932,
      "mean_token_accuracy": 0.1722172111272812,
      "num_tokens": 31512044.0,
      "step": 14395
    },
    {
      "entropy": 5.544270467758179,
      "epoch": 1.4291385470424771,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004802613857931592,
      "loss": 5.1879,
      "mean_token_accuracy": 0.17271738052368163,
      "num_tokens": 31523786.0,
      "step": 14400
    },
    {
      "entropy": 5.552690649032593,
      "epoch": 1.4296347757046446,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004802468711072472,
      "loss": 5.1512,
      "mean_token_accuracy": 0.1727266415953636,
      "num_tokens": 31534799.0,
      "step": 14405
    },
    {
      "entropy": 5.533667707443238,
      "epoch": 1.4301310043668123,
      "grad_norm": 1.015625,
      "learning_rate": 0.000480232351331655,
      "loss": 5.0784,
      "mean_token_accuracy": 0.18035546988248824,
      "num_tokens": 31544607.0,
      "step": 14410
    },
    {
      "entropy": 5.525924396514893,
      "epoch": 1.4306272330289798,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048021782646674277,
      "loss": 5.1199,
      "mean_token_accuracy": 0.16886428743600845,
      "num_tokens": 31556173.0,
      "step": 14415
    },
    {
      "entropy": 5.4796257495880125,
      "epoch": 1.4311234616911472,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048020329651287064,
      "loss": 5.0743,
      "mean_token_accuracy": 0.1735415369272232,
      "num_tokens": 31568206.0,
      "step": 14420
    },
    {
      "entropy": 5.5719257354736325,
      "epoch": 1.431619690353315,
      "grad_norm": 1.03125,
      "learning_rate": 0.000480188761470399,
      "loss": 5.2058,
      "mean_token_accuracy": 0.1626433923840523,
      "num_tokens": 31580021.0,
      "step": 14425
    },
    {
      "entropy": 5.487243890762329,
      "epoch": 1.4321159190154824,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004801742213396881,
      "loss": 5.0199,
      "mean_token_accuracy": 0.17503135204315184,
      "num_tokens": 31590343.0,
      "step": 14430
    },
    {
      "entropy": 5.509610271453857,
      "epoch": 1.4326121476776499,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048015967612109875,
      "loss": 5.1258,
      "mean_token_accuracy": 0.16812596023082732,
      "num_tokens": 31603387.0,
      "step": 14435
    },
    {
      "entropy": 5.483160972595215,
      "epoch": 1.4331083763398174,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048014512581499136,
      "loss": 5.0626,
      "mean_token_accuracy": 0.17583634555339814,
      "num_tokens": 31613640.0,
      "step": 14440
    },
    {
      "entropy": 5.558779907226563,
      "epoch": 1.4336046050019848,
      "grad_norm": 1.0,
      "learning_rate": 0.000480130570421727,
      "loss": 5.1395,
      "mean_token_accuracy": 0.17684323191642762,
      "num_tokens": 31624299.0,
      "step": 14445
    },
    {
      "entropy": 5.623220014572143,
      "epoch": 1.4341008336641523,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004801160099416665,
      "loss": 5.1715,
      "mean_token_accuracy": 0.1724609836935997,
      "num_tokens": 31635878.0,
      "step": 14450
    },
    {
      "entropy": 5.496553087234497,
      "epoch": 1.43459706232632,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004801014443751709,
      "loss": 5.056,
      "mean_token_accuracy": 0.17775179147720338,
      "num_tokens": 31648132.0,
      "step": 14455
    },
    {
      "entropy": 5.500878143310547,
      "epoch": 1.4350932909884875,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004800868737226015,
      "loss": 5.1384,
      "mean_token_accuracy": 0.17200473546981812,
      "num_tokens": 31657791.0,
      "step": 14460
    },
    {
      "entropy": 5.501968717575073,
      "epoch": 1.435589519650655,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048007229798431946,
      "loss": 5.0535,
      "mean_token_accuracy": 0.17453721314668655,
      "num_tokens": 31669561.0,
      "step": 14465
    },
    {
      "entropy": 5.4895820140838625,
      "epoch": 1.4360857483128227,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004800577171606864,
      "loss": 5.0782,
      "mean_token_accuracy": 0.17714058607816696,
      "num_tokens": 31680099.0,
      "step": 14470
    },
    {
      "entropy": 5.527725791931152,
      "epoch": 1.4365819769749901,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004800431312520638,
      "loss": 5.1484,
      "mean_token_accuracy": 0.17269737869501114,
      "num_tokens": 31692047.0,
      "step": 14475
    },
    {
      "entropy": 5.565553379058838,
      "epoch": 1.4370782056371576,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004800285402588134,
      "loss": 5.1694,
      "mean_token_accuracy": 0.17151716202497483,
      "num_tokens": 31703376.0,
      "step": 14480
    },
    {
      "entropy": 5.524068880081177,
      "epoch": 1.437574434299325,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004800139441812969,
      "loss": 5.1668,
      "mean_token_accuracy": 0.16686049997806549,
      "num_tokens": 31714201.0,
      "step": 14485
    },
    {
      "entropy": 5.629807853698731,
      "epoch": 1.4380706629614926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047999934301987644,
      "loss": 5.2024,
      "mean_token_accuracy": 0.1608566462993622,
      "num_tokens": 31725357.0,
      "step": 14490
    },
    {
      "entropy": 5.629396867752075,
      "epoch": 1.4385668916236602,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000479984736774914,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1676245927810669,
      "num_tokens": 31736311.0,
      "step": 14495
    },
    {
      "entropy": 5.5444141864776615,
      "epoch": 1.4390631202858277,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047997012544677176,
      "loss": 5.0828,
      "mean_token_accuracy": 0.1759336322546005,
      "num_tokens": 31747446.0,
      "step": 14500
    },
    {
      "entropy": 5.565255117416382,
      "epoch": 1.4395593489479952,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004799555090358121,
      "loss": 5.2645,
      "mean_token_accuracy": 0.16581302881240845,
      "num_tokens": 31758176.0,
      "step": 14505
    },
    {
      "entropy": 5.539243364334107,
      "epoch": 1.440055577610163,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004799408875423975,
      "loss": 5.0943,
      "mean_token_accuracy": 0.17127877920866014,
      "num_tokens": 31767597.0,
      "step": 14510
    },
    {
      "entropy": 5.573805618286133,
      "epoch": 1.4405518062723304,
      "grad_norm": 1.0,
      "learning_rate": 0.0004799262609668904,
      "loss": 5.2209,
      "mean_token_accuracy": 0.17043735235929489,
      "num_tokens": 31779521.0,
      "step": 14515
    },
    {
      "entropy": 5.545178747177124,
      "epoch": 1.4410480349344978,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004799116293096537,
      "loss": 5.1194,
      "mean_token_accuracy": 0.1731815055012703,
      "num_tokens": 31791780.0,
      "step": 14520
    },
    {
      "entropy": 5.446946144104004,
      "epoch": 1.4415442635966653,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004798969925710501,
      "loss": 5.0485,
      "mean_token_accuracy": 0.1771108999848366,
      "num_tokens": 31802280.0,
      "step": 14525
    },
    {
      "entropy": 5.501938772201538,
      "epoch": 1.4420404922588328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047988235075144254,
      "loss": 5.0386,
      "mean_token_accuracy": 0.1782764747738838,
      "num_tokens": 31812515.0,
      "step": 14530
    },
    {
      "entropy": 5.522060394287109,
      "epoch": 1.4425367209210003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004798677038511942,
      "loss": 5.1675,
      "mean_token_accuracy": 0.17112843543291092,
      "num_tokens": 31823483.0,
      "step": 14535
    },
    {
      "entropy": 5.620578193664551,
      "epoch": 1.443032949583168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004798530518706683,
      "loss": 5.2572,
      "mean_token_accuracy": 0.16600303053855897,
      "num_tokens": 31834978.0,
      "step": 14540
    },
    {
      "entropy": 5.614982080459595,
      "epoch": 1.4435291782453354,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004798383948102281,
      "loss": 5.1047,
      "mean_token_accuracy": 0.18115582466125488,
      "num_tokens": 31845375.0,
      "step": 14545
    },
    {
      "entropy": 5.5310853004455565,
      "epoch": 1.444025406907503,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004798237326702371,
      "loss": 5.189,
      "mean_token_accuracy": 0.1711132064461708,
      "num_tokens": 31856682.0,
      "step": 14550
    },
    {
      "entropy": 5.498862266540527,
      "epoch": 1.4445216355696706,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047980906545105884,
      "loss": 5.0939,
      "mean_token_accuracy": 0.17913826704025268,
      "num_tokens": 31867904.0,
      "step": 14555
    },
    {
      "entropy": 5.604802083969116,
      "epoch": 1.445017864231838,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004797943931530571,
      "loss": 5.2102,
      "mean_token_accuracy": 0.16719010323286057,
      "num_tokens": 31878929.0,
      "step": 14560
    },
    {
      "entropy": 5.556772375106812,
      "epoch": 1.4455140928940056,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004797797157765958,
      "loss": 5.1404,
      "mean_token_accuracy": 0.1752620071172714,
      "num_tokens": 31890117.0,
      "step": 14565
    },
    {
      "entropy": 5.576875257492065,
      "epoch": 1.446010321556173,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004797650333220387,
      "loss": 5.1864,
      "mean_token_accuracy": 0.16601977795362471,
      "num_tokens": 31901669.0,
      "step": 14570
    },
    {
      "entropy": 5.482005596160889,
      "epoch": 1.4465065502183405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047975034578975006,
      "loss": 5.1429,
      "mean_token_accuracy": 0.17350834012031555,
      "num_tokens": 31913774.0,
      "step": 14575
    },
    {
      "entropy": 5.536506271362304,
      "epoch": 1.4470027788805082,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000479735653180094,
      "loss": 5.1031,
      "mean_token_accuracy": 0.17331975251436232,
      "num_tokens": 31924435.0,
      "step": 14580
    },
    {
      "entropy": 5.546610593795776,
      "epoch": 1.4474990075426757,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004797209554934349,
      "loss": 5.1531,
      "mean_token_accuracy": 0.16817228347063065,
      "num_tokens": 31935359.0,
      "step": 14585
    },
    {
      "entropy": 5.590019798278808,
      "epoch": 1.4479952362048432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004797062527301373,
      "loss": 5.1786,
      "mean_token_accuracy": 0.16719991713762283,
      "num_tokens": 31946224.0,
      "step": 14590
    },
    {
      "entropy": 5.530066919326782,
      "epoch": 1.4484914648670106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004796915448905657,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17805466651916504,
      "num_tokens": 31956458.0,
      "step": 14595
    },
    {
      "entropy": 5.592698383331299,
      "epoch": 1.4489876935291783,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047967683197508486,
      "loss": 5.1594,
      "mean_token_accuracy": 0.17314376533031464,
      "num_tokens": 31966984.0,
      "step": 14600
    },
    {
      "entropy": 5.55301308631897,
      "epoch": 1.4494839221913458,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047966211398405966,
      "loss": 5.1459,
      "mean_token_accuracy": 0.1707082763314247,
      "num_tokens": 31978058.0,
      "step": 14605
    },
    {
      "entropy": 5.56214656829834,
      "epoch": 1.4499801508535133,
      "grad_norm": 1.0,
      "learning_rate": 0.000479647390917855,
      "loss": 5.1674,
      "mean_token_accuracy": 0.16170575171709062,
      "num_tokens": 31987968.0,
      "step": 14610
    },
    {
      "entropy": 5.5077613353729244,
      "epoch": 1.4504763795156808,
      "grad_norm": 0.9375,
      "learning_rate": 0.000479632662776836,
      "loss": 5.1098,
      "mean_token_accuracy": 0.1746418669819832,
      "num_tokens": 31999005.0,
      "step": 14615
    },
    {
      "entropy": 5.607811832427979,
      "epoch": 1.4509726081778482,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000479617929561368,
      "loss": 5.2294,
      "mean_token_accuracy": 0.1620844542980194,
      "num_tokens": 32011997.0,
      "step": 14620
    },
    {
      "entropy": 5.638015222549439,
      "epoch": 1.451468836840016,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004796031912718162,
      "loss": 5.2162,
      "mean_token_accuracy": 0.16827848106622695,
      "num_tokens": 32024429.0,
      "step": 14625
    },
    {
      "entropy": 5.533307361602783,
      "epoch": 1.4519650655021834,
      "grad_norm": 1.0,
      "learning_rate": 0.00047958844790854606,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1770411491394043,
      "num_tokens": 32034176.0,
      "step": 14630
    },
    {
      "entropy": 5.4563640594482425,
      "epoch": 1.4524612941643509,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047957369947192337,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1756252244114876,
      "num_tokens": 32045482.0,
      "step": 14635
    },
    {
      "entropy": 5.4972411632537845,
      "epoch": 1.4529575228265186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004795589459623136,
      "loss": 5.0863,
      "mean_token_accuracy": 0.1720438554883003,
      "num_tokens": 32056890.0,
      "step": 14640
    },
    {
      "entropy": 5.471979665756225,
      "epoch": 1.453453751488686,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047954418738008295,
      "loss": 5.0331,
      "mean_token_accuracy": 0.18008808940649032,
      "num_tokens": 32067915.0,
      "step": 14645
    },
    {
      "entropy": 5.499617958068848,
      "epoch": 1.4539499801508535,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004795294237255971,
      "loss": 5.052,
      "mean_token_accuracy": 0.18036505430936814,
      "num_tokens": 32080026.0,
      "step": 14650
    },
    {
      "entropy": 5.486179494857788,
      "epoch": 1.454446208813021,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047951465499922234,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1756055936217308,
      "num_tokens": 32090150.0,
      "step": 14655
    },
    {
      "entropy": 5.581918525695801,
      "epoch": 1.4549424374751885,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004794998812013248,
      "loss": 5.205,
      "mean_token_accuracy": 0.16557288765907288,
      "num_tokens": 32101858.0,
      "step": 14660
    },
    {
      "entropy": 5.460342884063721,
      "epoch": 1.4554386661373562,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047948510233227084,
      "loss": 5.0582,
      "mean_token_accuracy": 0.17734010368585587,
      "num_tokens": 32112078.0,
      "step": 14665
    },
    {
      "entropy": 5.548183345794678,
      "epoch": 1.4559348947995236,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004794703183924269,
      "loss": 5.139,
      "mean_token_accuracy": 0.17388161569833754,
      "num_tokens": 32123539.0,
      "step": 14670
    },
    {
      "entropy": 5.619879817962646,
      "epoch": 1.4564311234616911,
      "grad_norm": 1.0,
      "learning_rate": 0.0004794555293821597,
      "loss": 5.2243,
      "mean_token_accuracy": 0.17040760219097137,
      "num_tokens": 32135063.0,
      "step": 14675
    },
    {
      "entropy": 5.56371283531189,
      "epoch": 1.4569273521238586,
      "grad_norm": 0.96875,
      "learning_rate": 0.000479440735301836,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1653165802359581,
      "num_tokens": 32145407.0,
      "step": 14680
    },
    {
      "entropy": 5.499918031692505,
      "epoch": 1.4574235807860263,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004794259361518226,
      "loss": 5.1351,
      "mean_token_accuracy": 0.17262995541095733,
      "num_tokens": 32156340.0,
      "step": 14685
    },
    {
      "entropy": 5.511840152740478,
      "epoch": 1.4579198094481938,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047941113193248643,
      "loss": 5.1193,
      "mean_token_accuracy": 0.1778529942035675,
      "num_tokens": 32168627.0,
      "step": 14690
    },
    {
      "entropy": 5.5513897895812985,
      "epoch": 1.4584160381103612,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004793963226441947,
      "loss": 5.1305,
      "mean_token_accuracy": 0.1675900086760521,
      "num_tokens": 32179770.0,
      "step": 14695
    },
    {
      "entropy": 5.604520082473755,
      "epoch": 1.4589122667725287,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047938150828731455,
      "loss": 5.1123,
      "mean_token_accuracy": 0.17456894218921662,
      "num_tokens": 32190309.0,
      "step": 14700
    },
    {
      "entropy": 5.551539373397827,
      "epoch": 1.4594084954346962,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004793666888622134,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17042281329631806,
      "num_tokens": 32201113.0,
      "step": 14705
    },
    {
      "entropy": 5.525229454040527,
      "epoch": 1.459904724096864,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004793518643692587,
      "loss": 5.148,
      "mean_token_accuracy": 0.173648901283741,
      "num_tokens": 32211412.0,
      "step": 14710
    },
    {
      "entropy": 5.620937156677246,
      "epoch": 1.4604009527590314,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004793370348088182,
      "loss": 5.1255,
      "mean_token_accuracy": 0.17041030377149582,
      "num_tokens": 32222992.0,
      "step": 14715
    },
    {
      "entropy": 5.5358561992645265,
      "epoch": 1.4608971814211988,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004793222001812595,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17519469559192657,
      "num_tokens": 32234600.0,
      "step": 14720
    },
    {
      "entropy": 5.5811577320098875,
      "epoch": 1.4613934100833665,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047930736048695045,
      "loss": 5.2889,
      "mean_token_accuracy": 0.16633660197257996,
      "num_tokens": 32245835.0,
      "step": 14725
    },
    {
      "entropy": 5.545107793807984,
      "epoch": 1.461889638745534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004792925157262592,
      "loss": 5.1785,
      "mean_token_accuracy": 0.17525752931833266,
      "num_tokens": 32257006.0,
      "step": 14730
    },
    {
      "entropy": 5.5801488876342775,
      "epoch": 1.4623858674077015,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004792776658995537,
      "loss": 5.1389,
      "mean_token_accuracy": 0.16772967875003814,
      "num_tokens": 32268049.0,
      "step": 14735
    },
    {
      "entropy": 5.517350864410401,
      "epoch": 1.462882096069869,
      "grad_norm": 1.0,
      "learning_rate": 0.0004792628110072023,
      "loss": 5.1102,
      "mean_token_accuracy": 0.17724338918924332,
      "num_tokens": 32278035.0,
      "step": 14740
    },
    {
      "entropy": 5.5452831268310545,
      "epoch": 1.4633783247320364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004792479510495734,
      "loss": 5.1316,
      "mean_token_accuracy": 0.16876960843801497,
      "num_tokens": 32289029.0,
      "step": 14745
    },
    {
      "entropy": 5.544247055053711,
      "epoch": 1.463874553394204,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047923308602703535,
      "loss": 5.1293,
      "mean_token_accuracy": 0.16625548899173737,
      "num_tokens": 32299656.0,
      "step": 14750
    },
    {
      "entropy": 5.545353412628174,
      "epoch": 1.4643707820563716,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047921821593995684,
      "loss": 5.1061,
      "mean_token_accuracy": 0.16833749115467073,
      "num_tokens": 32309394.0,
      "step": 14755
    },
    {
      "entropy": 5.569381189346314,
      "epoch": 1.464867010718539,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004792033407887066,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1702450767159462,
      "num_tokens": 32319908.0,
      "step": 14760
    },
    {
      "entropy": 5.513531017303467,
      "epoch": 1.4653632393807066,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004791884605736536,
      "loss": 5.0684,
      "mean_token_accuracy": 0.17679776549339293,
      "num_tokens": 32331190.0,
      "step": 14765
    },
    {
      "entropy": 5.548891496658325,
      "epoch": 1.4658594680428743,
      "grad_norm": 0.875,
      "learning_rate": 0.00047917357529516676,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1725134491920471,
      "num_tokens": 32342150.0,
      "step": 14770
    },
    {
      "entropy": 5.571631288528442,
      "epoch": 1.4663556967050417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791586849536152,
      "loss": 5.1775,
      "mean_token_accuracy": 0.17259956598281861,
      "num_tokens": 32353480.0,
      "step": 14775
    },
    {
      "entropy": 5.562407064437866,
      "epoch": 1.4668519253672092,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004791437895493682,
      "loss": 5.154,
      "mean_token_accuracy": 0.17417239248752595,
      "num_tokens": 32366365.0,
      "step": 14780
    },
    {
      "entropy": 5.580090713500977,
      "epoch": 1.4673481540293767,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047912888908279507,
      "loss": 5.1911,
      "mean_token_accuracy": 0.16850529611110687,
      "num_tokens": 32377186.0,
      "step": 14785
    },
    {
      "entropy": 5.54986424446106,
      "epoch": 1.4678443826915442,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004791139835542653,
      "loss": 5.2385,
      "mean_token_accuracy": 0.16543520390987396,
      "num_tokens": 32389207.0,
      "step": 14790
    },
    {
      "entropy": 5.550461387634277,
      "epoch": 1.4683406113537119,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047909907296414864,
      "loss": 5.1743,
      "mean_token_accuracy": 0.17201529294252396,
      "num_tokens": 32400278.0,
      "step": 14795
    },
    {
      "entropy": 5.615238380432129,
      "epoch": 1.4688368400158793,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047908415731281475,
      "loss": 5.192,
      "mean_token_accuracy": 0.1679785892367363,
      "num_tokens": 32412237.0,
      "step": 14800
    },
    {
      "entropy": 5.608131742477417,
      "epoch": 1.4693330686780468,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047906923660063344,
      "loss": 5.1786,
      "mean_token_accuracy": 0.16993998289108275,
      "num_tokens": 32423236.0,
      "step": 14805
    },
    {
      "entropy": 5.571867561340332,
      "epoch": 1.4698292973402145,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000479054310827975,
      "loss": 5.1518,
      "mean_token_accuracy": 0.17088284343481064,
      "num_tokens": 32434296.0,
      "step": 14810
    },
    {
      "entropy": 5.554528379440308,
      "epoch": 1.470325526002382,
      "grad_norm": 1.0,
      "learning_rate": 0.00047903937999520913,
      "loss": 5.1167,
      "mean_token_accuracy": 0.17332854866981506,
      "num_tokens": 32444934.0,
      "step": 14815
    },
    {
      "entropy": 5.486271619796753,
      "epoch": 1.4708217546645495,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004790244441027064,
      "loss": 5.0533,
      "mean_token_accuracy": 0.17526827603578568,
      "num_tokens": 32454157.0,
      "step": 14820
    },
    {
      "entropy": 5.492707777023315,
      "epoch": 1.471317983326717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004790095031508371,
      "loss": 5.1108,
      "mean_token_accuracy": 0.17401860058307647,
      "num_tokens": 32464586.0,
      "step": 14825
    },
    {
      "entropy": 5.537542867660522,
      "epoch": 1.4718142119888844,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047899455713997164,
      "loss": 5.1126,
      "mean_token_accuracy": 0.16536028385162355,
      "num_tokens": 32474121.0,
      "step": 14830
    },
    {
      "entropy": 5.55899829864502,
      "epoch": 1.4723104406510519,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047897960607048084,
      "loss": 5.1805,
      "mean_token_accuracy": 0.16896182149648667,
      "num_tokens": 32485539.0,
      "step": 14835
    },
    {
      "entropy": 5.522295761108398,
      "epoch": 1.4728066693132196,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004789646499427353,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17368097752332687,
      "num_tokens": 32495970.0,
      "step": 14840
    },
    {
      "entropy": 5.5500163555145265,
      "epoch": 1.473302897975387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004789496887571059,
      "loss": 5.0779,
      "mean_token_accuracy": 0.17539246529340743,
      "num_tokens": 32506401.0,
      "step": 14845
    },
    {
      "entropy": 5.5105836391448975,
      "epoch": 1.4737991266375545,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047893472251396373,
      "loss": 5.1063,
      "mean_token_accuracy": 0.18453878462314605,
      "num_tokens": 32518033.0,
      "step": 14850
    },
    {
      "entropy": 5.565513658523559,
      "epoch": 1.4742953552997222,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047891975121368,
      "loss": 5.1828,
      "mean_token_accuracy": 0.16908366978168488,
      "num_tokens": 32529225.0,
      "step": 14855
    },
    {
      "entropy": 5.591162586212159,
      "epoch": 1.4747915839618897,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047890477485662563,
      "loss": 5.1709,
      "mean_token_accuracy": 0.16898744851350783,
      "num_tokens": 32540246.0,
      "step": 14860
    },
    {
      "entropy": 5.532024478912353,
      "epoch": 1.4752878126240572,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047888979344317236,
      "loss": 5.1402,
      "mean_token_accuracy": 0.17460403889417647,
      "num_tokens": 32551001.0,
      "step": 14865
    },
    {
      "entropy": 5.601256418228149,
      "epoch": 1.4757840412862246,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047887480697369155,
      "loss": 5.1225,
      "mean_token_accuracy": 0.1699197769165039,
      "num_tokens": 32561338.0,
      "step": 14870
    },
    {
      "entropy": 5.531030607223511,
      "epoch": 1.4762802699483921,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047885981544855475,
      "loss": 5.1346,
      "mean_token_accuracy": 0.17381287664175032,
      "num_tokens": 32570919.0,
      "step": 14875
    },
    {
      "entropy": 5.526541996002197,
      "epoch": 1.4767764986105598,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004788448188681339,
      "loss": 5.1877,
      "mean_token_accuracy": 0.17006363719701767,
      "num_tokens": 32581361.0,
      "step": 14880
    },
    {
      "entropy": 5.558997058868409,
      "epoch": 1.4772727272727273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004788298172328007,
      "loss": 5.1685,
      "mean_token_accuracy": 0.1697208434343338,
      "num_tokens": 32592533.0,
      "step": 14885
    },
    {
      "entropy": 5.558480978012085,
      "epoch": 1.4777689559348948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047881481054292736,
      "loss": 5.0709,
      "mean_token_accuracy": 0.18095472306013108,
      "num_tokens": 32602207.0,
      "step": 14890
    },
    {
      "entropy": 5.525254726409912,
      "epoch": 1.4782651845970625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047879979879888584,
      "loss": 5.0524,
      "mean_token_accuracy": 0.18161077946424484,
      "num_tokens": 32611288.0,
      "step": 14895
    },
    {
      "entropy": 5.563801431655884,
      "epoch": 1.47876141325923,
      "grad_norm": 1.0,
      "learning_rate": 0.00047878478200104845,
      "loss": 5.1458,
      "mean_token_accuracy": 0.17050823867321013,
      "num_tokens": 32622078.0,
      "step": 14900
    },
    {
      "entropy": 5.585149908065796,
      "epoch": 1.4792576419213974,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004787697601497876,
      "loss": 5.247,
      "mean_token_accuracy": 0.167140956223011,
      "num_tokens": 32634182.0,
      "step": 14905
    },
    {
      "entropy": 5.617268514633179,
      "epoch": 1.479753870583565,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004787547332454758,
      "loss": 5.2211,
      "mean_token_accuracy": 0.167884524166584,
      "num_tokens": 32646245.0,
      "step": 14910
    },
    {
      "entropy": 5.488656234741211,
      "epoch": 1.4802500992457324,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047873970128848567,
      "loss": 5.1001,
      "mean_token_accuracy": 0.16893306374549866,
      "num_tokens": 32657521.0,
      "step": 14915
    },
    {
      "entropy": 5.604245042800903,
      "epoch": 1.4807463279078998,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004787246642791899,
      "loss": 5.1164,
      "mean_token_accuracy": 0.16815983802080153,
      "num_tokens": 32668740.0,
      "step": 14920
    },
    {
      "entropy": 5.631704378128052,
      "epoch": 1.4812425565700675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047870962221796147,
      "loss": 5.1585,
      "mean_token_accuracy": 0.17043596804141997,
      "num_tokens": 32680933.0,
      "step": 14925
    },
    {
      "entropy": 5.513810539245606,
      "epoch": 1.481738785232235,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004786945751051734,
      "loss": 5.0738,
      "mean_token_accuracy": 0.1729463666677475,
      "num_tokens": 32691690.0,
      "step": 14930
    },
    {
      "entropy": 5.488545227050781,
      "epoch": 1.4822350138944025,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004786795229411988,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17888772040605544,
      "num_tokens": 32703500.0,
      "step": 14935
    },
    {
      "entropy": 5.549590921401977,
      "epoch": 1.4827312425565702,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047866446572641084,
      "loss": 5.1561,
      "mean_token_accuracy": 0.1685801327228546,
      "num_tokens": 32714662.0,
      "step": 14940
    },
    {
      "entropy": 5.53198504447937,
      "epoch": 1.4832274712187377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004786494034611831,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17579153180122375,
      "num_tokens": 32725995.0,
      "step": 14945
    },
    {
      "entropy": 5.523895597457885,
      "epoch": 1.4837236998809051,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000478634336145889,
      "loss": 5.1225,
      "mean_token_accuracy": 0.1715308889746666,
      "num_tokens": 32736510.0,
      "step": 14950
    },
    {
      "entropy": 5.554951763153076,
      "epoch": 1.4842199285430726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047861926378090197,
      "loss": 5.1472,
      "mean_token_accuracy": 0.17159119695425035,
      "num_tokens": 32747032.0,
      "step": 14955
    },
    {
      "entropy": 5.532624053955078,
      "epoch": 1.48471615720524,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786041863665961,
      "loss": 5.1006,
      "mean_token_accuracy": 0.17886370420455933,
      "num_tokens": 32757199.0,
      "step": 14960
    },
    {
      "entropy": 5.533308219909668,
      "epoch": 1.4852123858674078,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004785891039033451,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17700377851724625,
      "num_tokens": 32767326.0,
      "step": 14965
    },
    {
      "entropy": 5.637337779998779,
      "epoch": 1.4857086145295753,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000478574016391523,
      "loss": 5.2162,
      "mean_token_accuracy": 0.16348069459199904,
      "num_tokens": 32777621.0,
      "step": 14970
    },
    {
      "entropy": 5.508499479293823,
      "epoch": 1.4862048431917427,
      "grad_norm": 1.078125,
      "learning_rate": 0.000478558923831504,
      "loss": 5.0986,
      "mean_token_accuracy": 0.16993338167667388,
      "num_tokens": 32787938.0,
      "step": 14975
    },
    {
      "entropy": 5.511209392547608,
      "epoch": 1.4867010718539102,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047854382622366235,
      "loss": 5.0231,
      "mean_token_accuracy": 0.18294786810874938,
      "num_tokens": 32798047.0,
      "step": 14980
    },
    {
      "entropy": 5.552734422683716,
      "epoch": 1.487197300516078,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047852872356837236,
      "loss": 5.1516,
      "mean_token_accuracy": 0.17062964886426926,
      "num_tokens": 32809338.0,
      "step": 14985
    },
    {
      "entropy": 5.552220487594605,
      "epoch": 1.4876935291782454,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004785136158660085,
      "loss": 5.093,
      "mean_token_accuracy": 0.17595915943384172,
      "num_tokens": 32820555.0,
      "step": 14990
    },
    {
      "entropy": 5.624473571777344,
      "epoch": 1.4881897578404129,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047849850311694567,
      "loss": 5.1408,
      "mean_token_accuracy": 0.16409851759672164,
      "num_tokens": 32830832.0,
      "step": 14995
    },
    {
      "entropy": 5.496233701705933,
      "epoch": 1.4886859865025803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047848338532155833,
      "loss": 5.0137,
      "mean_token_accuracy": 0.17156196087598802,
      "num_tokens": 32840534.0,
      "step": 15000
    },
    {
      "epoch": 1.4886859865025803,
      "eval_entropy": 5.359660616621152,
      "eval_loss": 5.243100643157959,
      "eval_mean_token_accuracy": 0.17213165217045368,
      "eval_num_tokens": 32840534.0,
      "eval_runtime": 20.3819,
      "eval_samples_per_second": 1651.904,
      "eval_steps_per_second": 206.506,
      "step": 15000
    },
    {
      "entropy": 5.539487743377686,
      "epoch": 1.4891822151647478,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047846826248022155,
      "loss": 5.0951,
      "mean_token_accuracy": 0.17391763180494307,
      "num_tokens": 32851109.0,
      "step": 15005
    },
    {
      "entropy": 5.638070964813233,
      "epoch": 1.4896784438269155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047845313459331034,
      "loss": 5.2242,
      "mean_token_accuracy": 0.1742781102657318,
      "num_tokens": 32862491.0,
      "step": 15010
    },
    {
      "entropy": 5.504267930984497,
      "epoch": 1.490174672489083,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047843800166119974,
      "loss": 5.0275,
      "mean_token_accuracy": 0.17828538417816162,
      "num_tokens": 32873831.0,
      "step": 15015
    },
    {
      "entropy": 5.635790157318115,
      "epoch": 1.4906709011512504,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000478422863684265,
      "loss": 5.1679,
      "mean_token_accuracy": 0.1677897647023201,
      "num_tokens": 32884288.0,
      "step": 15020
    },
    {
      "entropy": 5.531105422973633,
      "epoch": 1.4911671298134181,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047840772066288167,
      "loss": 5.0849,
      "mean_token_accuracy": 0.17495175898075105,
      "num_tokens": 32894867.0,
      "step": 15025
    },
    {
      "entropy": 5.444433689117432,
      "epoch": 1.4916633584755856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004783925725974252,
      "loss": 5.1354,
      "mean_token_accuracy": 0.16987663507461548,
      "num_tokens": 32906627.0,
      "step": 15030
    },
    {
      "entropy": 5.533215618133545,
      "epoch": 1.492159587137753,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004783774194882712,
      "loss": 5.1288,
      "mean_token_accuracy": 0.16910700052976607,
      "num_tokens": 32916245.0,
      "step": 15035
    },
    {
      "entropy": 5.605276679992675,
      "epoch": 1.4926558157999206,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004783622613357954,
      "loss": 5.1937,
      "mean_token_accuracy": 0.17639414370059966,
      "num_tokens": 32927663.0,
      "step": 15040
    },
    {
      "entropy": 5.600903749465942,
      "epoch": 1.493152044462088,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004783470981403737,
      "loss": 5.1749,
      "mean_token_accuracy": 0.16646047681570053,
      "num_tokens": 32939106.0,
      "step": 15045
    },
    {
      "entropy": 5.552848815917969,
      "epoch": 1.4936482731242555,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004783319299023821,
      "loss": 5.1499,
      "mean_token_accuracy": 0.17064561694860458,
      "num_tokens": 32949628.0,
      "step": 15050
    },
    {
      "entropy": 5.535466814041138,
      "epoch": 1.4941445017864232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004783167566221969,
      "loss": 5.1691,
      "mean_token_accuracy": 0.16880062222480774,
      "num_tokens": 32959908.0,
      "step": 15055
    },
    {
      "entropy": 5.527226305007934,
      "epoch": 1.4946407304485907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047830157830019423,
      "loss": 5.0725,
      "mean_token_accuracy": 0.17808809876441956,
      "num_tokens": 32970312.0,
      "step": 15060
    },
    {
      "entropy": 5.542770576477051,
      "epoch": 1.4951369591107582,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004782863949367505,
      "loss": 5.1193,
      "mean_token_accuracy": 0.17275449335575105,
      "num_tokens": 32982876.0,
      "step": 15065
    },
    {
      "entropy": 5.556779241561889,
      "epoch": 1.4956331877729259,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004782712065322421,
      "loss": 5.1403,
      "mean_token_accuracy": 0.16716893166303634,
      "num_tokens": 32994223.0,
      "step": 15070
    },
    {
      "entropy": 5.630604553222656,
      "epoch": 1.4961294164350933,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004782560130870459,
      "loss": 5.215,
      "mean_token_accuracy": 0.1750601202249527,
      "num_tokens": 33006282.0,
      "step": 15075
    },
    {
      "entropy": 5.582544612884521,
      "epoch": 1.4966256450972608,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047824081460153846,
      "loss": 5.1935,
      "mean_token_accuracy": 0.16784726828336716,
      "num_tokens": 33016823.0,
      "step": 15080
    },
    {
      "entropy": 5.54985671043396,
      "epoch": 1.4971218737594283,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004782256110760969,
      "loss": 5.1453,
      "mean_token_accuracy": 0.17013245820999146,
      "num_tokens": 33027963.0,
      "step": 15085
    },
    {
      "entropy": 5.612236595153808,
      "epoch": 1.4976181024215958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000478210402511098,
      "loss": 5.1972,
      "mean_token_accuracy": 0.1648408755660057,
      "num_tokens": 33038711.0,
      "step": 15090
    },
    {
      "entropy": 5.571948480606079,
      "epoch": 1.4981143310837635,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004781951889069189,
      "loss": 5.1133,
      "mean_token_accuracy": 0.17684465646743774,
      "num_tokens": 33048847.0,
      "step": 15095
    },
    {
      "entropy": 5.570899724960327,
      "epoch": 1.498610559745931,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047817997026393714,
      "loss": 5.1516,
      "mean_token_accuracy": 0.17015185058116913,
      "num_tokens": 33059775.0,
      "step": 15100
    },
    {
      "entropy": 5.648701524734497,
      "epoch": 1.4991067884080984,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047816474658252985,
      "loss": 5.2523,
      "mean_token_accuracy": 0.17028822451829911,
      "num_tokens": 33071543.0,
      "step": 15105
    },
    {
      "entropy": 5.624151515960693,
      "epoch": 1.499603017070266,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047814951786307467,
      "loss": 5.1984,
      "mean_token_accuracy": 0.16685270071029662,
      "num_tokens": 33082487.0,
      "step": 15110
    },
    {
      "entropy": 5.591991233825683,
      "epoch": 1.5000992457324336,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004781342841059492,
      "loss": 5.1725,
      "mean_token_accuracy": 0.1663709968328476,
      "num_tokens": 33093381.0,
      "step": 15115
    },
    {
      "entropy": 5.530705165863037,
      "epoch": 1.500595474394601,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004781190453115311,
      "loss": 5.0753,
      "mean_token_accuracy": 0.1733100160956383,
      "num_tokens": 33104122.0,
      "step": 15120
    },
    {
      "entropy": 5.611812400817871,
      "epoch": 1.5010917030567685,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047810380148019847,
      "loss": 5.2428,
      "mean_token_accuracy": 0.16964877098798753,
      "num_tokens": 33115541.0,
      "step": 15125
    },
    {
      "entropy": 5.458960723876953,
      "epoch": 1.501587931718936,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047808855261232915,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1735677018761635,
      "num_tokens": 33125002.0,
      "step": 15130
    },
    {
      "entropy": 5.537723350524902,
      "epoch": 1.5020841603811035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047807329870830134,
      "loss": 5.2028,
      "mean_token_accuracy": 0.17011199444532393,
      "num_tokens": 33135733.0,
      "step": 15135
    },
    {
      "entropy": 5.571906185150146,
      "epoch": 1.5025803890432712,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004780580397684933,
      "loss": 5.117,
      "mean_token_accuracy": 0.1676930695772171,
      "num_tokens": 33146514.0,
      "step": 15140
    },
    {
      "entropy": 5.600480270385742,
      "epoch": 1.5030766177054387,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004780427757932835,
      "loss": 5.105,
      "mean_token_accuracy": 0.16542457193136215,
      "num_tokens": 33157070.0,
      "step": 15145
    },
    {
      "entropy": 5.516840028762817,
      "epoch": 1.5035728463676064,
      "grad_norm": 0.875,
      "learning_rate": 0.0004780275067830503,
      "loss": 5.1164,
      "mean_token_accuracy": 0.1748972937464714,
      "num_tokens": 33170311.0,
      "step": 15150
    },
    {
      "entropy": 5.564585161209107,
      "epoch": 1.5040690750297738,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004780122327381725,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1757461592555046,
      "num_tokens": 33180510.0,
      "step": 15155
    },
    {
      "entropy": 5.53900671005249,
      "epoch": 1.5045653036919413,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047799695365902865,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17404764145612717,
      "num_tokens": 33192265.0,
      "step": 15160
    },
    {
      "entropy": 5.571393442153931,
      "epoch": 1.5050615323541088,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004779816695459979,
      "loss": 5.1216,
      "mean_token_accuracy": 0.17036762088537216,
      "num_tokens": 33204080.0,
      "step": 15165
    },
    {
      "entropy": 5.548530197143554,
      "epoch": 1.5055577610162763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004779663803994591,
      "loss": 5.1663,
      "mean_token_accuracy": 0.17031425833702088,
      "num_tokens": 33216090.0,
      "step": 15170
    },
    {
      "entropy": 5.609906148910523,
      "epoch": 1.5060539896784437,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004779510862197914,
      "loss": 5.2084,
      "mean_token_accuracy": 0.16210805475711823,
      "num_tokens": 33227465.0,
      "step": 15175
    },
    {
      "entropy": 5.6083275318145756,
      "epoch": 1.5065502183406112,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004779357870073741,
      "loss": 5.1718,
      "mean_token_accuracy": 0.16870828717947006,
      "num_tokens": 33237510.0,
      "step": 15180
    },
    {
      "entropy": 5.624159431457519,
      "epoch": 1.507046447002779,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047792048276258653,
      "loss": 5.2197,
      "mean_token_accuracy": 0.1676163673400879,
      "num_tokens": 33247349.0,
      "step": 15185
    },
    {
      "entropy": 5.557723951339722,
      "epoch": 1.5075426756649464,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004779051734858083,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17551186680793762,
      "num_tokens": 33258357.0,
      "step": 15190
    },
    {
      "entropy": 5.572274541854858,
      "epoch": 1.508038904327114,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047788985917741894,
      "loss": 5.2137,
      "mean_token_accuracy": 0.16664404273033143,
      "num_tokens": 33269355.0,
      "step": 15195
    },
    {
      "entropy": 5.563484382629395,
      "epoch": 1.5085351329892815,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047787453983779826,
      "loss": 5.0333,
      "mean_token_accuracy": 0.17926741391420364,
      "num_tokens": 33279129.0,
      "step": 15200
    },
    {
      "entropy": 5.562342882156372,
      "epoch": 1.509031361651449,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004778592154673262,
      "loss": 5.0859,
      "mean_token_accuracy": 0.16859497576951982,
      "num_tokens": 33290681.0,
      "step": 15205
    },
    {
      "entropy": 5.564073467254639,
      "epoch": 1.5095275903136165,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047784388606638263,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1736873671412468,
      "num_tokens": 33300738.0,
      "step": 15210
    },
    {
      "entropy": 5.600442504882812,
      "epoch": 1.510023818975784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004778285516353478,
      "loss": 5.1953,
      "mean_token_accuracy": 0.16898488700389863,
      "num_tokens": 33310829.0,
      "step": 15215
    },
    {
      "entropy": 5.575578832626343,
      "epoch": 1.5105200476379514,
      "grad_norm": 1.0,
      "learning_rate": 0.00047781321217460205,
      "loss": 5.1679,
      "mean_token_accuracy": 0.17518054991960524,
      "num_tokens": 33320814.0,
      "step": 15220
    },
    {
      "entropy": 5.510858488082886,
      "epoch": 1.5110162763001191,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047779786768452554,
      "loss": 5.072,
      "mean_token_accuracy": 0.17451430708169938,
      "num_tokens": 33330105.0,
      "step": 15225
    },
    {
      "entropy": 5.498918151855468,
      "epoch": 1.5115125049622866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047778251816549895,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17905570864677428,
      "num_tokens": 33340986.0,
      "step": 15230
    },
    {
      "entropy": 5.539443445205689,
      "epoch": 1.512008733624454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047776716361790284,
      "loss": 5.1897,
      "mean_token_accuracy": 0.16624922156333924,
      "num_tokens": 33352096.0,
      "step": 15235
    },
    {
      "entropy": 5.589443302154541,
      "epoch": 1.5125049622866218,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047775180404211804,
      "loss": 5.1604,
      "mean_token_accuracy": 0.16917768865823746,
      "num_tokens": 33362581.0,
      "step": 15240
    },
    {
      "entropy": 5.576366710662842,
      "epoch": 1.5130011909487893,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004777364394385254,
      "loss": 5.1808,
      "mean_token_accuracy": 0.1776186242699623,
      "num_tokens": 33374020.0,
      "step": 15245
    },
    {
      "entropy": 5.5252461433410645,
      "epoch": 1.5134974196109567,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047772106980750593,
      "loss": 5.1469,
      "mean_token_accuracy": 0.17434298247098923,
      "num_tokens": 33384755.0,
      "step": 15250
    },
    {
      "entropy": 5.593572807312012,
      "epoch": 1.5139936482731242,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004777056951494407,
      "loss": 5.1384,
      "mean_token_accuracy": 0.17855725884437562,
      "num_tokens": 33395837.0,
      "step": 15255
    },
    {
      "entropy": 5.56649260520935,
      "epoch": 1.5144898769352917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477690315464711,
      "loss": 5.1667,
      "mean_token_accuracy": 0.1731877878308296,
      "num_tokens": 33408083.0,
      "step": 15260
    },
    {
      "entropy": 5.5642131805419925,
      "epoch": 1.5149861055974592,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047767493075369827,
      "loss": 5.2229,
      "mean_token_accuracy": 0.17151160389184952,
      "num_tokens": 33419433.0,
      "step": 15265
    },
    {
      "entropy": 5.606595659255982,
      "epoch": 1.5154823342596269,
      "grad_norm": 0.859375,
      "learning_rate": 0.000477659541016784,
      "loss": 5.2087,
      "mean_token_accuracy": 0.17334197759628295,
      "num_tokens": 33431920.0,
      "step": 15270
    },
    {
      "entropy": 5.595513105392456,
      "epoch": 1.5159785629217943,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047764414625434983,
      "loss": 5.1484,
      "mean_token_accuracy": 0.16829569190740584,
      "num_tokens": 33443428.0,
      "step": 15275
    },
    {
      "entropy": 5.541310834884643,
      "epoch": 1.516474791583962,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004776287464667774,
      "loss": 5.0919,
      "mean_token_accuracy": 0.16591509133577348,
      "num_tokens": 33455324.0,
      "step": 15280
    },
    {
      "entropy": 5.56178126335144,
      "epoch": 1.5169710202461295,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047761334165444876,
      "loss": 5.1641,
      "mean_token_accuracy": 0.17028047293424606,
      "num_tokens": 33468640.0,
      "step": 15285
    },
    {
      "entropy": 5.533207750320434,
      "epoch": 1.517467248908297,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047759793181774586,
      "loss": 5.123,
      "mean_token_accuracy": 0.1753290191292763,
      "num_tokens": 33479681.0,
      "step": 15290
    },
    {
      "entropy": 5.605574417114258,
      "epoch": 1.5179634775704645,
      "grad_norm": 1.0,
      "learning_rate": 0.00047758251695705073,
      "loss": 5.1755,
      "mean_token_accuracy": 0.1707095757126808,
      "num_tokens": 33490797.0,
      "step": 15295
    },
    {
      "entropy": 5.56215353012085,
      "epoch": 1.518459706232632,
      "grad_norm": 0.875,
      "learning_rate": 0.00047756709707274576,
      "loss": 5.2099,
      "mean_token_accuracy": 0.17242370694875717,
      "num_tokens": 33501649.0,
      "step": 15300
    },
    {
      "entropy": 5.523342418670654,
      "epoch": 1.5189559348947994,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004775516721652133,
      "loss": 5.0404,
      "mean_token_accuracy": 0.18114760667085647,
      "num_tokens": 33512169.0,
      "step": 15305
    },
    {
      "entropy": 5.541622447967529,
      "epoch": 1.519452163556967,
      "grad_norm": 1.0,
      "learning_rate": 0.0004775362422348358,
      "loss": 5.1388,
      "mean_token_accuracy": 0.16830806732177733,
      "num_tokens": 33522866.0,
      "step": 15310
    },
    {
      "entropy": 5.488192987442017,
      "epoch": 1.5199483922191346,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047752080728199586,
      "loss": 5.0556,
      "mean_token_accuracy": 0.17523340731859208,
      "num_tokens": 33533671.0,
      "step": 15315
    },
    {
      "entropy": 5.545552682876587,
      "epoch": 1.520444620881302,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004775053673070764,
      "loss": 5.1081,
      "mean_token_accuracy": 0.178180268406868,
      "num_tokens": 33544475.0,
      "step": 15320
    },
    {
      "entropy": 5.538996934890747,
      "epoch": 1.5209408495434698,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004774899223104601,
      "loss": 5.2261,
      "mean_token_accuracy": 0.17271717637777328,
      "num_tokens": 33555691.0,
      "step": 15325
    },
    {
      "entropy": 5.574954700469971,
      "epoch": 1.5214370782056372,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004774744722925301,
      "loss": 5.1347,
      "mean_token_accuracy": 0.17057869732379913,
      "num_tokens": 33566980.0,
      "step": 15330
    },
    {
      "entropy": 5.522434282302856,
      "epoch": 1.5219333068678047,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004774590172536695,
      "loss": 5.0795,
      "mean_token_accuracy": 0.17030835896730423,
      "num_tokens": 33577947.0,
      "step": 15335
    },
    {
      "entropy": 5.544117403030396,
      "epoch": 1.5224295355299722,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047744355719426146,
      "loss": 5.1564,
      "mean_token_accuracy": 0.17432398498058319,
      "num_tokens": 33589703.0,
      "step": 15340
    },
    {
      "entropy": 5.527563667297363,
      "epoch": 1.5229257641921397,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047742809211468953,
      "loss": 5.1323,
      "mean_token_accuracy": 0.1735093340277672,
      "num_tokens": 33599418.0,
      "step": 15345
    },
    {
      "entropy": 5.589821434020996,
      "epoch": 1.5234219928543071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000477412622015337,
      "loss": 5.077,
      "mean_token_accuracy": 0.1740545004606247,
      "num_tokens": 33609914.0,
      "step": 15350
    },
    {
      "entropy": 5.539106941223144,
      "epoch": 1.5239182215164748,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004773971468965876,
      "loss": 5.1786,
      "mean_token_accuracy": 0.16492396742105483,
      "num_tokens": 33621134.0,
      "step": 15355
    },
    {
      "entropy": 5.596384334564209,
      "epoch": 1.5244144501786423,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047738166675882516,
      "loss": 5.2023,
      "mean_token_accuracy": 0.16908951252698898,
      "num_tokens": 33631665.0,
      "step": 15360
    },
    {
      "entropy": 5.600810575485229,
      "epoch": 1.52491067884081,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004773661816024335,
      "loss": 5.1613,
      "mean_token_accuracy": 0.1672820121049881,
      "num_tokens": 33643161.0,
      "step": 15365
    },
    {
      "entropy": 5.564737796783447,
      "epoch": 1.5254069075029775,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004773506914277965,
      "loss": 5.1433,
      "mean_token_accuracy": 0.17112142443656922,
      "num_tokens": 33654893.0,
      "step": 15370
    },
    {
      "entropy": 5.538595485687256,
      "epoch": 1.525903136165145,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047733519623529853,
      "loss": 5.1418,
      "mean_token_accuracy": 0.16837041825056076,
      "num_tokens": 33664688.0,
      "step": 15375
    },
    {
      "entropy": 5.5296409130096436,
      "epoch": 1.5263993648273124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004773196960253236,
      "loss": 5.1393,
      "mean_token_accuracy": 0.17312368899583816,
      "num_tokens": 33676084.0,
      "step": 15380
    },
    {
      "entropy": 5.562348222732544,
      "epoch": 1.52689559348948,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004773041907982561,
      "loss": 5.1609,
      "mean_token_accuracy": 0.16654255241155624,
      "num_tokens": 33686946.0,
      "step": 15385
    },
    {
      "entropy": 5.491490125656128,
      "epoch": 1.5273918221516474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047728868055448064,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17732964903116227,
      "num_tokens": 33697994.0,
      "step": 15390
    },
    {
      "entropy": 5.529824256896973,
      "epoch": 1.5278880508138148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004772731652943818,
      "loss": 5.1582,
      "mean_token_accuracy": 0.16738653182983398,
      "num_tokens": 33709618.0,
      "step": 15395
    },
    {
      "entropy": 5.563512945175171,
      "epoch": 1.5283842794759825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004772576450183444,
      "loss": 5.177,
      "mean_token_accuracy": 0.17754658460617065,
      "num_tokens": 33720282.0,
      "step": 15400
    },
    {
      "entropy": 5.526491117477417,
      "epoch": 1.52888050813815,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047724211972675315,
      "loss": 5.1405,
      "mean_token_accuracy": 0.167572021484375,
      "num_tokens": 33730581.0,
      "step": 15405
    },
    {
      "entropy": 5.596597337722779,
      "epoch": 1.5293767368003177,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004772265894199931,
      "loss": 5.1869,
      "mean_token_accuracy": 0.1716473802924156,
      "num_tokens": 33741732.0,
      "step": 15410
    },
    {
      "entropy": 5.637492275238037,
      "epoch": 1.5298729654624852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004772110540984494,
      "loss": 5.2191,
      "mean_token_accuracy": 0.17000600695610046,
      "num_tokens": 33753043.0,
      "step": 15415
    },
    {
      "entropy": 5.548772716522217,
      "epoch": 1.5303691941246527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004771955137625073,
      "loss": 5.084,
      "mean_token_accuracy": 0.17809344083070755,
      "num_tokens": 33763974.0,
      "step": 15420
    },
    {
      "entropy": 5.556105995178223,
      "epoch": 1.5308654227868201,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047717996841255224,
      "loss": 5.0331,
      "mean_token_accuracy": 0.1804227888584137,
      "num_tokens": 33775247.0,
      "step": 15425
    },
    {
      "entropy": 5.593108129501343,
      "epoch": 1.5313616514489876,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047716441804896955,
      "loss": 5.1856,
      "mean_token_accuracy": 0.1761174127459526,
      "num_tokens": 33786843.0,
      "step": 15430
    },
    {
      "entropy": 5.5571410179138185,
      "epoch": 1.531857880111155,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004771488626721449,
      "loss": 5.0875,
      "mean_token_accuracy": 0.17593777775764466,
      "num_tokens": 33797974.0,
      "step": 15435
    },
    {
      "entropy": 5.5414644241333,
      "epoch": 1.5323541087733228,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047713330228246406,
      "loss": 5.1347,
      "mean_token_accuracy": 0.1721976801753044,
      "num_tokens": 33807965.0,
      "step": 15440
    },
    {
      "entropy": 5.523689031600952,
      "epoch": 1.5328503374354903,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004771177368803129,
      "loss": 5.127,
      "mean_token_accuracy": 0.17405371367931366,
      "num_tokens": 33819835.0,
      "step": 15445
    },
    {
      "entropy": 5.541213274002075,
      "epoch": 1.533346566097658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047710216646607733,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17970117032527924,
      "num_tokens": 33830041.0,
      "step": 15450
    },
    {
      "entropy": 5.614241409301758,
      "epoch": 1.5338427947598254,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047708659104014363,
      "loss": 5.1412,
      "mean_token_accuracy": 0.1661602735519409,
      "num_tokens": 33841681.0,
      "step": 15455
    },
    {
      "entropy": 5.5539332866668705,
      "epoch": 1.534339023421993,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004770710106028978,
      "loss": 5.1039,
      "mean_token_accuracy": 0.16841839998960495,
      "num_tokens": 33852411.0,
      "step": 15460
    },
    {
      "entropy": 5.543445062637329,
      "epoch": 1.5348352520841604,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047705542515472647,
      "loss": 5.1155,
      "mean_token_accuracy": 0.1743115022778511,
      "num_tokens": 33862130.0,
      "step": 15465
    },
    {
      "entropy": 5.519922828674316,
      "epoch": 1.5353314807463279,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004770398346960159,
      "loss": 5.1463,
      "mean_token_accuracy": 0.17489728480577468,
      "num_tokens": 33872308.0,
      "step": 15470
    },
    {
      "entropy": 5.578985548019409,
      "epoch": 1.5358277094084953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047702423922715284,
      "loss": 5.0983,
      "mean_token_accuracy": 0.1728236570954323,
      "num_tokens": 33882704.0,
      "step": 15475
    },
    {
      "entropy": 5.598556232452393,
      "epoch": 1.5363239380706628,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004770086387485239,
      "loss": 5.1571,
      "mean_token_accuracy": 0.16862391233444213,
      "num_tokens": 33893796.0,
      "step": 15480
    },
    {
      "entropy": 5.504836654663086,
      "epoch": 1.5368201667328305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000476993033260516,
      "loss": 5.0383,
      "mean_token_accuracy": 0.18265025317668915,
      "num_tokens": 33903723.0,
      "step": 15485
    },
    {
      "entropy": 5.4927764415740965,
      "epoch": 1.537316395394998,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004769774227635162,
      "loss": 5.0342,
      "mean_token_accuracy": 0.17889236211776732,
      "num_tokens": 33914295.0,
      "step": 15490
    },
    {
      "entropy": 5.704986572265625,
      "epoch": 1.5378126240571657,
      "grad_norm": 1.0,
      "learning_rate": 0.0004769618072579115,
      "loss": 5.2655,
      "mean_token_accuracy": 0.1684098646044731,
      "num_tokens": 33925414.0,
      "step": 15495
    },
    {
      "entropy": 5.5883502006530765,
      "epoch": 1.5383088527193332,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047694618674408917,
      "loss": 5.1233,
      "mean_token_accuracy": 0.1656289204955101,
      "num_tokens": 33936878.0,
      "step": 15500
    },
    {
      "entropy": 5.54160475730896,
      "epoch": 1.5388050813815006,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047693056122243655,
      "loss": 5.0973,
      "mean_token_accuracy": 0.16982535123825074,
      "num_tokens": 33947684.0,
      "step": 15505
    },
    {
      "entropy": 5.583298206329346,
      "epoch": 1.539301310043668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047691493069334124,
      "loss": 5.161,
      "mean_token_accuracy": 0.1708986520767212,
      "num_tokens": 33958102.0,
      "step": 15510
    },
    {
      "entropy": 5.619075345993042,
      "epoch": 1.5397975387058356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047689929515719066,
      "loss": 5.1536,
      "mean_token_accuracy": 0.16536005586385727,
      "num_tokens": 33968606.0,
      "step": 15515
    },
    {
      "entropy": 5.558409929275513,
      "epoch": 1.540293767368003,
      "grad_norm": 1.0,
      "learning_rate": 0.00047688365461437263,
      "loss": 5.0676,
      "mean_token_accuracy": 0.1815166488289833,
      "num_tokens": 33978859.0,
      "step": 15520
    },
    {
      "entropy": 5.521059370040893,
      "epoch": 1.5407899960301707,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047686800906527493,
      "loss": 5.1623,
      "mean_token_accuracy": 0.1706143543124199,
      "num_tokens": 33991056.0,
      "step": 15525
    },
    {
      "entropy": 5.544672012329102,
      "epoch": 1.5412862246923382,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004768523585102856,
      "loss": 5.1314,
      "mean_token_accuracy": 0.17441264241933824,
      "num_tokens": 34001257.0,
      "step": 15530
    },
    {
      "entropy": 5.5343945026397705,
      "epoch": 1.541782453354506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004768367029497928,
      "loss": 5.1295,
      "mean_token_accuracy": 0.17110612690448762,
      "num_tokens": 34012881.0,
      "step": 15535
    },
    {
      "entropy": 5.56253981590271,
      "epoch": 1.5422786820166734,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047682104238418464,
      "loss": 5.1113,
      "mean_token_accuracy": 0.1682277426123619,
      "num_tokens": 34023503.0,
      "step": 15540
    },
    {
      "entropy": 5.663846158981324,
      "epoch": 1.5427749106788409,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047680537681384956,
      "loss": 5.1882,
      "mean_token_accuracy": 0.17299121618270874,
      "num_tokens": 34033137.0,
      "step": 15545
    },
    {
      "entropy": 5.569497346878052,
      "epoch": 1.5432711393410083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004767897062391759,
      "loss": 5.0935,
      "mean_token_accuracy": 0.17460624277591705,
      "num_tokens": 34044584.0,
      "step": 15550
    },
    {
      "entropy": 5.591175937652588,
      "epoch": 1.5437673680031758,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004767740306605524,
      "loss": 5.1105,
      "mean_token_accuracy": 0.169645032286644,
      "num_tokens": 34055575.0,
      "step": 15555
    },
    {
      "entropy": 5.571935796737671,
      "epoch": 1.5442635966653433,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004767583500783677,
      "loss": 5.1752,
      "mean_token_accuracy": 0.17342199832201005,
      "num_tokens": 34067036.0,
      "step": 15560
    },
    {
      "entropy": 5.566770076751709,
      "epoch": 1.5447598253275108,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047674266449301074,
      "loss": 5.1448,
      "mean_token_accuracy": 0.1708250805735588,
      "num_tokens": 34078705.0,
      "step": 15565
    },
    {
      "entropy": 5.595689821243286,
      "epoch": 1.5452560539896785,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047672697390487034,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1671139970421791,
      "num_tokens": 34089437.0,
      "step": 15570
    },
    {
      "entropy": 5.5844707012176515,
      "epoch": 1.545752282651846,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004767112783143357,
      "loss": 5.167,
      "mean_token_accuracy": 0.17696029245853423,
      "num_tokens": 34101714.0,
      "step": 15575
    },
    {
      "entropy": 5.578916740417481,
      "epoch": 1.5462485113140136,
      "grad_norm": 0.96875,
      "learning_rate": 0.000476695577721796,
      "loss": 5.1632,
      "mean_token_accuracy": 0.16738015785813332,
      "num_tokens": 34112596.0,
      "step": 15580
    },
    {
      "entropy": 5.555521631240845,
      "epoch": 1.5467447399761811,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047667987212764065,
      "loss": 5.042,
      "mean_token_accuracy": 0.1844472572207451,
      "num_tokens": 34122777.0,
      "step": 15585
    },
    {
      "entropy": 5.5732354640960695,
      "epoch": 1.5472409686383486,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047666416153225894,
      "loss": 5.1767,
      "mean_token_accuracy": 0.1730051353573799,
      "num_tokens": 34133887.0,
      "step": 15590
    },
    {
      "entropy": 5.588015079498291,
      "epoch": 1.547737197300516,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004766484459360406,
      "loss": 5.1745,
      "mean_token_accuracy": 0.16691353172063828,
      "num_tokens": 34144494.0,
      "step": 15595
    },
    {
      "entropy": 5.612547874450684,
      "epoch": 1.5482334259626835,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047663272533937537,
      "loss": 5.1779,
      "mean_token_accuracy": 0.17513032108545304,
      "num_tokens": 34156604.0,
      "step": 15600
    },
    {
      "entropy": 5.530529022216797,
      "epoch": 1.548729654624851,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047661699974265294,
      "loss": 5.1351,
      "mean_token_accuracy": 0.17220962792634964,
      "num_tokens": 34167342.0,
      "step": 15605
    },
    {
      "entropy": 5.5928185939788815,
      "epoch": 1.5492258832870187,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047660126914626344,
      "loss": 5.0524,
      "mean_token_accuracy": 0.17704794853925704,
      "num_tokens": 34178611.0,
      "step": 15610
    },
    {
      "entropy": 5.576434659957886,
      "epoch": 1.5497221119491862,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004765855335505969,
      "loss": 5.0858,
      "mean_token_accuracy": 0.17098576873540877,
      "num_tokens": 34189658.0,
      "step": 15615
    },
    {
      "entropy": 5.497120666503906,
      "epoch": 1.5502183406113537,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004765697929560434,
      "loss": 5.0541,
      "mean_token_accuracy": 0.17879538238048553,
      "num_tokens": 34200080.0,
      "step": 15620
    },
    {
      "entropy": 5.565816259384155,
      "epoch": 1.5507145692735214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047655404736299354,
      "loss": 5.1211,
      "mean_token_accuracy": 0.18178338706493377,
      "num_tokens": 34210184.0,
      "step": 15625
    },
    {
      "entropy": 5.60120530128479,
      "epoch": 1.5512107979356888,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047653829677183745,
      "loss": 5.1479,
      "mean_token_accuracy": 0.17209903746843339,
      "num_tokens": 34220448.0,
      "step": 15630
    },
    {
      "entropy": 5.51785101890564,
      "epoch": 1.5517070265978563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047652254118296596,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1676121324300766,
      "num_tokens": 34231595.0,
      "step": 15635
    },
    {
      "entropy": 5.530422115325928,
      "epoch": 1.5522032552600238,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004765067805967697,
      "loss": 5.1449,
      "mean_token_accuracy": 0.16904819160699844,
      "num_tokens": 34244066.0,
      "step": 15640
    },
    {
      "entropy": 5.548707151412964,
      "epoch": 1.5526994839221913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004764910150136395,
      "loss": 5.1454,
      "mean_token_accuracy": 0.169217786192894,
      "num_tokens": 34255505.0,
      "step": 15645
    },
    {
      "entropy": 5.56300311088562,
      "epoch": 1.5531957125843587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047647524443396625,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17469275891780853,
      "num_tokens": 34265386.0,
      "step": 15650
    },
    {
      "entropy": 5.533375453948975,
      "epoch": 1.5536919412465264,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047645946885814105,
      "loss": 5.1101,
      "mean_token_accuracy": 0.17387116253376006,
      "num_tokens": 34278317.0,
      "step": 15655
    },
    {
      "entropy": 5.614212608337402,
      "epoch": 1.554188169908694,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004764436882865552,
      "loss": 5.2131,
      "mean_token_accuracy": 0.1626477211713791,
      "num_tokens": 34290814.0,
      "step": 15660
    },
    {
      "entropy": 5.602675247192383,
      "epoch": 1.5546843985708616,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004764279027195999,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16987041234970093,
      "num_tokens": 34302819.0,
      "step": 15665
    },
    {
      "entropy": 5.523565578460693,
      "epoch": 1.555180627233029,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047641211215766673,
      "loss": 5.1136,
      "mean_token_accuracy": 0.1813419744372368,
      "num_tokens": 34314055.0,
      "step": 15670
    },
    {
      "entropy": 5.506222724914551,
      "epoch": 1.5556768558951966,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004763963166011471,
      "loss": 5.0714,
      "mean_token_accuracy": 0.18090740740299224,
      "num_tokens": 34324826.0,
      "step": 15675
    },
    {
      "entropy": 5.565904188156128,
      "epoch": 1.556173084557364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004763805160504328,
      "loss": 5.1233,
      "mean_token_accuracy": 0.1817057728767395,
      "num_tokens": 34334540.0,
      "step": 15680
    },
    {
      "entropy": 5.559889268875122,
      "epoch": 1.5566693132195315,
      "grad_norm": 1.0,
      "learning_rate": 0.0004763647105059157,
      "loss": 5.1087,
      "mean_token_accuracy": 0.1761892706155777,
      "num_tokens": 34344501.0,
      "step": 15685
    },
    {
      "entropy": 5.5242908000946045,
      "epoch": 1.557165541881699,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004763488999679877,
      "loss": 5.074,
      "mean_token_accuracy": 0.17178085446357727,
      "num_tokens": 34355207.0,
      "step": 15690
    },
    {
      "entropy": 5.554287385940552,
      "epoch": 1.5576617705438665,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047633308443704076,
      "loss": 5.1364,
      "mean_token_accuracy": 0.17130820602178573,
      "num_tokens": 34366839.0,
      "step": 15695
    },
    {
      "entropy": 5.63304762840271,
      "epoch": 1.5581579992060341,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047631726391346717,
      "loss": 5.2703,
      "mean_token_accuracy": 0.17287765592336654,
      "num_tokens": 34377863.0,
      "step": 15700
    },
    {
      "entropy": 5.69172101020813,
      "epoch": 1.5586542278682016,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004763014383976593,
      "loss": 5.2609,
      "mean_token_accuracy": 0.1592468038201332,
      "num_tokens": 34390490.0,
      "step": 15705
    },
    {
      "entropy": 5.575370264053345,
      "epoch": 1.5591504565303693,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047628560789000945,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17584925442934035,
      "num_tokens": 34401979.0,
      "step": 15710
    },
    {
      "entropy": 5.5386114597320555,
      "epoch": 1.5596466851925368,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047626977239091027,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1667221337556839,
      "num_tokens": 34413470.0,
      "step": 15715
    },
    {
      "entropy": 5.653692436218262,
      "epoch": 1.5601429138547043,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004762539319007545,
      "loss": 5.1889,
      "mean_token_accuracy": 0.16975097507238388,
      "num_tokens": 34424805.0,
      "step": 15720
    },
    {
      "entropy": 5.546775245666504,
      "epoch": 1.5606391425168717,
      "grad_norm": 1.125,
      "learning_rate": 0.0004762380864199348,
      "loss": 5.0896,
      "mean_token_accuracy": 0.18097617030143737,
      "num_tokens": 34433994.0,
      "step": 15725
    },
    {
      "entropy": 5.559839820861816,
      "epoch": 1.5611353711790392,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004762222359488443,
      "loss": 5.1253,
      "mean_token_accuracy": 0.16789624840021133,
      "num_tokens": 34444230.0,
      "step": 15730
    },
    {
      "entropy": 5.5227149486541744,
      "epoch": 1.5616315998412067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004762063804878758,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17765397429466248,
      "num_tokens": 34453424.0,
      "step": 15735
    },
    {
      "entropy": 5.573944091796875,
      "epoch": 1.5621278285033744,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047619052003742277,
      "loss": 5.1265,
      "mean_token_accuracy": 0.17211733013391495,
      "num_tokens": 34464283.0,
      "step": 15740
    },
    {
      "entropy": 5.52869725227356,
      "epoch": 1.5626240571655419,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004761746545978782,
      "loss": 5.0234,
      "mean_token_accuracy": 0.17962296903133393,
      "num_tokens": 34475397.0,
      "step": 15745
    },
    {
      "entropy": 5.529394960403442,
      "epoch": 1.5631202858277096,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004761587841696358,
      "loss": 5.1244,
      "mean_token_accuracy": 0.1780758187174797,
      "num_tokens": 34486534.0,
      "step": 15750
    },
    {
      "entropy": 5.529142189025879,
      "epoch": 1.563616514489877,
      "grad_norm": 0.875,
      "learning_rate": 0.000476142908753089,
      "loss": 5.1836,
      "mean_token_accuracy": 0.16724986135959624,
      "num_tokens": 34498245.0,
      "step": 15755
    },
    {
      "entropy": 5.600046014785766,
      "epoch": 1.5641127431520445,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004761270283486314,
      "loss": 5.2013,
      "mean_token_accuracy": 0.1674613505601883,
      "num_tokens": 34509502.0,
      "step": 15760
    },
    {
      "entropy": 5.566359758377075,
      "epoch": 1.564608971814212,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047611114295665697,
      "loss": 5.0213,
      "mean_token_accuracy": 0.18188208788633348,
      "num_tokens": 34519754.0,
      "step": 15765
    },
    {
      "entropy": 5.574743938446045,
      "epoch": 1.5651052004763795,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047609525257755967,
      "loss": 5.111,
      "mean_token_accuracy": 0.17353034764528275,
      "num_tokens": 34530952.0,
      "step": 15770
    },
    {
      "entropy": 5.490397691726685,
      "epoch": 1.565601429138547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004760793572117333,
      "loss": 5.062,
      "mean_token_accuracy": 0.16983846127986907,
      "num_tokens": 34540780.0,
      "step": 15775
    },
    {
      "entropy": 5.518644094467163,
      "epoch": 1.5660976578007144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004760634568595721,
      "loss": 5.1598,
      "mean_token_accuracy": 0.17214315533638,
      "num_tokens": 34550828.0,
      "step": 15780
    },
    {
      "entropy": 5.552333068847656,
      "epoch": 1.5665938864628821,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004760475515214704,
      "loss": 5.0314,
      "mean_token_accuracy": 0.17744303196668626,
      "num_tokens": 34561842.0,
      "step": 15785
    },
    {
      "entropy": 5.54453330039978,
      "epoch": 1.5670901151250496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004760316411978228,
      "loss": 5.1187,
      "mean_token_accuracy": 0.17400076985359192,
      "num_tokens": 34573894.0,
      "step": 15790
    },
    {
      "entropy": 5.490247011184692,
      "epoch": 1.5675863437872173,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047601572588902363,
      "loss": 5.1282,
      "mean_token_accuracy": 0.1667056828737259,
      "num_tokens": 34583003.0,
      "step": 15795
    },
    {
      "entropy": 5.493989038467407,
      "epoch": 1.5680825724493848,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047599980559546757,
      "loss": 4.9966,
      "mean_token_accuracy": 0.17987972497940063,
      "num_tokens": 34594912.0,
      "step": 15800
    },
    {
      "entropy": 5.577528715133667,
      "epoch": 1.5685788011115522,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004759838803175495,
      "loss": 5.1499,
      "mean_token_accuracy": 0.16984748095273972,
      "num_tokens": 34606638.0,
      "step": 15805
    },
    {
      "entropy": 5.53093113899231,
      "epoch": 1.5690750297737197,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047596795005566423,
      "loss": 5.0722,
      "mean_token_accuracy": 0.17618204206228255,
      "num_tokens": 34617438.0,
      "step": 15810
    },
    {
      "entropy": 5.565863561630249,
      "epoch": 1.5695712584358872,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047595201481020685,
      "loss": 5.1632,
      "mean_token_accuracy": 0.16739971786737443,
      "num_tokens": 34627857.0,
      "step": 15815
    },
    {
      "entropy": 5.50247278213501,
      "epoch": 1.5700674870980547,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047593607458157257,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1758079394698143,
      "num_tokens": 34638243.0,
      "step": 15820
    },
    {
      "entropy": 5.560802268981933,
      "epoch": 1.5705637157602224,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047592012937015647,
      "loss": 5.0903,
      "mean_token_accuracy": 0.17577720135450364,
      "num_tokens": 34648355.0,
      "step": 15825
    },
    {
      "entropy": 5.513834619522095,
      "epoch": 1.5710599444223898,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047590417917635423,
      "loss": 4.9988,
      "mean_token_accuracy": 0.1786194249987602,
      "num_tokens": 34658644.0,
      "step": 15830
    },
    {
      "entropy": 5.524498319625854,
      "epoch": 1.5715561730845575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004758882240005613,
      "loss": 5.1228,
      "mean_token_accuracy": 0.17384404689073563,
      "num_tokens": 34669892.0,
      "step": 15835
    },
    {
      "entropy": 5.658471393585205,
      "epoch": 1.572052401746725,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047587226384317323,
      "loss": 5.1354,
      "mean_token_accuracy": 0.1698479250073433,
      "num_tokens": 34681197.0,
      "step": 15840
    },
    {
      "entropy": 5.584916019439698,
      "epoch": 1.5725486304088925,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047585629870458584,
      "loss": 5.1531,
      "mean_token_accuracy": 0.1677242785692215,
      "num_tokens": 34692070.0,
      "step": 15845
    },
    {
      "entropy": 5.571759557723999,
      "epoch": 1.57304485907106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004758403285851951,
      "loss": 5.1777,
      "mean_token_accuracy": 0.16891357749700547,
      "num_tokens": 34702667.0,
      "step": 15850
    },
    {
      "entropy": 5.517935562133789,
      "epoch": 1.5735410877332274,
      "grad_norm": 1.0,
      "learning_rate": 0.00047582435348539695,
      "loss": 5.0425,
      "mean_token_accuracy": 0.17908268570899963,
      "num_tokens": 34713936.0,
      "step": 15855
    },
    {
      "entropy": 5.634885311126709,
      "epoch": 1.574037316395395,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004758083734055875,
      "loss": 5.1285,
      "mean_token_accuracy": 0.176302070915699,
      "num_tokens": 34724195.0,
      "step": 15860
    },
    {
      "entropy": 5.596161937713623,
      "epoch": 1.5745335450575624,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004757923883461632,
      "loss": 5.1364,
      "mean_token_accuracy": 0.16749471724033355,
      "num_tokens": 34735932.0,
      "step": 15865
    },
    {
      "entropy": 5.640626335144043,
      "epoch": 1.57502977371973,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047577639830752034,
      "loss": 5.2302,
      "mean_token_accuracy": 0.16650352030992507,
      "num_tokens": 34747451.0,
      "step": 15870
    },
    {
      "entropy": 5.620966529846191,
      "epoch": 1.5755260023818976,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047576040329005536,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17572931796312333,
      "num_tokens": 34757591.0,
      "step": 15875
    },
    {
      "entropy": 5.560580253601074,
      "epoch": 1.5760222310440652,
      "grad_norm": 1.078125,
      "learning_rate": 0.000475744403294165,
      "loss": 5.1271,
      "mean_token_accuracy": 0.17774790227413179,
      "num_tokens": 34768420.0,
      "step": 15880
    },
    {
      "entropy": 5.59808874130249,
      "epoch": 1.5765184597062327,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047572839832024606,
      "loss": 5.1729,
      "mean_token_accuracy": 0.17383802980184554,
      "num_tokens": 34778984.0,
      "step": 15885
    },
    {
      "entropy": 5.581893587112427,
      "epoch": 1.5770146883684002,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047571238836869537,
      "loss": 5.1966,
      "mean_token_accuracy": 0.1637273132801056,
      "num_tokens": 34790913.0,
      "step": 15890
    },
    {
      "entropy": 5.529088544845581,
      "epoch": 1.5775109170305677,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004756963734399099,
      "loss": 5.1018,
      "mean_token_accuracy": 0.1729653596878052,
      "num_tokens": 34801873.0,
      "step": 15895
    },
    {
      "entropy": 5.669301080703735,
      "epoch": 1.5780071456927351,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047568035353428683,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1615908920764923,
      "num_tokens": 34812871.0,
      "step": 15900
    },
    {
      "entropy": 5.580141878128051,
      "epoch": 1.5785033743549026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047566432865222343,
      "loss": 5.1531,
      "mean_token_accuracy": 0.17512816935777664,
      "num_tokens": 34824405.0,
      "step": 15905
    },
    {
      "entropy": 5.578042602539062,
      "epoch": 1.5789996030170703,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047564829879411706,
      "loss": 5.166,
      "mean_token_accuracy": 0.1710125610232353,
      "num_tokens": 34834048.0,
      "step": 15910
    },
    {
      "entropy": 5.599609279632569,
      "epoch": 1.5794958316792378,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004756322639603653,
      "loss": 5.1379,
      "mean_token_accuracy": 0.16838570535182953,
      "num_tokens": 34844915.0,
      "step": 15915
    },
    {
      "entropy": 5.539455270767212,
      "epoch": 1.5799920603414053,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047561622415136565,
      "loss": 5.0912,
      "mean_token_accuracy": 0.1705571800470352,
      "num_tokens": 34855282.0,
      "step": 15920
    },
    {
      "entropy": 5.5386110782623295,
      "epoch": 1.580488289003573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047560017936751597,
      "loss": 5.0979,
      "mean_token_accuracy": 0.1759043574333191,
      "num_tokens": 34865541.0,
      "step": 15925
    },
    {
      "entropy": 5.5632775783538815,
      "epoch": 1.5809845176657404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755841296092141,
      "loss": 5.141,
      "mean_token_accuracy": 0.16728954017162323,
      "num_tokens": 34876326.0,
      "step": 15930
    },
    {
      "entropy": 5.580970525741577,
      "epoch": 1.581480746327908,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047556807487685797,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16906953305006028,
      "num_tokens": 34887319.0,
      "step": 15935
    },
    {
      "entropy": 5.663427925109863,
      "epoch": 1.5819769749900754,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004755520151708458,
      "loss": 5.2323,
      "mean_token_accuracy": 0.17245990931987762,
      "num_tokens": 34900511.0,
      "step": 15940
    },
    {
      "entropy": 5.604037427902222,
      "epoch": 1.5824732036522429,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004755359504915758,
      "loss": 5.0935,
      "mean_token_accuracy": 0.17019872218370438,
      "num_tokens": 34910037.0,
      "step": 15945
    },
    {
      "entropy": 5.501194334030151,
      "epoch": 1.5829694323144103,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004755198808394464,
      "loss": 5.0764,
      "mean_token_accuracy": 0.17678191959857942,
      "num_tokens": 34921618.0,
      "step": 15950
    },
    {
      "entropy": 5.574261665344238,
      "epoch": 1.583465660976578,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000475503806214856,
      "loss": 5.1805,
      "mean_token_accuracy": 0.1716103196144104,
      "num_tokens": 34932010.0,
      "step": 15955
    },
    {
      "entropy": 5.5322043895721436,
      "epoch": 1.5839618896387455,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004754877266182033,
      "loss": 5.0878,
      "mean_token_accuracy": 0.17728703022003173,
      "num_tokens": 34942252.0,
      "step": 15960
    },
    {
      "entropy": 5.614795684814453,
      "epoch": 1.5844581183009132,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000475471642049887,
      "loss": 5.1588,
      "mean_token_accuracy": 0.17035353034734727,
      "num_tokens": 34953343.0,
      "step": 15965
    },
    {
      "entropy": 5.632703065872192,
      "epoch": 1.5849543469630807,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004754555525103058,
      "loss": 5.0881,
      "mean_token_accuracy": 0.17763637602329255,
      "num_tokens": 34964537.0,
      "step": 15970
    },
    {
      "entropy": 5.63137674331665,
      "epoch": 1.5854505756252482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475439457999859,
      "loss": 5.169,
      "mean_token_accuracy": 0.1682259753346443,
      "num_tokens": 34973655.0,
      "step": 15975
    },
    {
      "entropy": 5.572925281524658,
      "epoch": 1.5859468042874156,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047542335851894553,
      "loss": 5.1773,
      "mean_token_accuracy": 0.1734218955039978,
      "num_tokens": 34984826.0,
      "step": 15980
    },
    {
      "entropy": 5.570227003097534,
      "epoch": 1.586443032949583,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047540725406796454,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1758432224392891,
      "num_tokens": 34995492.0,
      "step": 15985
    },
    {
      "entropy": 5.536518621444702,
      "epoch": 1.5869392616117506,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004753911446473156,
      "loss": 5.0853,
      "mean_token_accuracy": 0.17811484038829803,
      "num_tokens": 35007617.0,
      "step": 15990
    },
    {
      "entropy": 5.665560007095337,
      "epoch": 1.587435490273918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000475375030257398,
      "loss": 5.232,
      "mean_token_accuracy": 0.16265056878328324,
      "num_tokens": 35018022.0,
      "step": 15995
    },
    {
      "entropy": 5.5606041431427,
      "epoch": 1.5879317189360858,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004753589108986114,
      "loss": 5.1768,
      "mean_token_accuracy": 0.17253759354352952,
      "num_tokens": 35028725.0,
      "step": 16000
    },
    {
      "entropy": 5.637623977661133,
      "epoch": 1.5884279475982532,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047534278657135565,
      "loss": 5.1763,
      "mean_token_accuracy": 0.17744076997041702,
      "num_tokens": 35039067.0,
      "step": 16005
    },
    {
      "entropy": 5.602351188659668,
      "epoch": 1.588924176260421,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004753266572760304,
      "loss": 5.1041,
      "mean_token_accuracy": 0.1732746258378029,
      "num_tokens": 35051292.0,
      "step": 16010
    },
    {
      "entropy": 5.591950368881226,
      "epoch": 1.5894204049225884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047531052301303577,
      "loss": 5.1299,
      "mean_token_accuracy": 0.16840226501226424,
      "num_tokens": 35062046.0,
      "step": 16015
    },
    {
      "entropy": 5.583362865447998,
      "epoch": 1.5899166335847559,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004752943837827718,
      "loss": 5.1791,
      "mean_token_accuracy": 0.17230916917324066,
      "num_tokens": 35072275.0,
      "step": 16020
    },
    {
      "entropy": 5.523928642272949,
      "epoch": 1.5904128622469234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004752782395856387,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17364226281642914,
      "num_tokens": 35083173.0,
      "step": 16025
    },
    {
      "entropy": 5.570323324203491,
      "epoch": 1.5909090909090908,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004752620904220368,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17332767993211745,
      "num_tokens": 35094323.0,
      "step": 16030
    },
    {
      "entropy": 5.52860016822815,
      "epoch": 1.5914053195712583,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004752459362923666,
      "loss": 5.1047,
      "mean_token_accuracy": 0.16899746358394624,
      "num_tokens": 35105216.0,
      "step": 16035
    },
    {
      "entropy": 5.567777156829834,
      "epoch": 1.591901548233426,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004752297771970287,
      "loss": 5.1322,
      "mean_token_accuracy": 0.17398248314857484,
      "num_tokens": 35116283.0,
      "step": 16040
    },
    {
      "entropy": 5.6096912860870365,
      "epoch": 1.5923977768955935,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047521361313642373,
      "loss": 5.1306,
      "mean_token_accuracy": 0.17369395941495896,
      "num_tokens": 35126821.0,
      "step": 16045
    },
    {
      "entropy": 5.5625251770019535,
      "epoch": 1.5928940055577612,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004751974441109526,
      "loss": 5.1102,
      "mean_token_accuracy": 0.17086736261844634,
      "num_tokens": 35137860.0,
      "step": 16050
    },
    {
      "entropy": 5.56798825263977,
      "epoch": 1.5933902342199286,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047518127012101624,
      "loss": 5.0962,
      "mean_token_accuracy": 0.176138237118721,
      "num_tokens": 35149103.0,
      "step": 16055
    },
    {
      "entropy": 5.643968772888184,
      "epoch": 1.5938864628820961,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004751650911670157,
      "loss": 5.1581,
      "mean_token_accuracy": 0.16796195656061172,
      "num_tokens": 35158981.0,
      "step": 16060
    },
    {
      "entropy": 5.635502290725708,
      "epoch": 1.5943826915442636,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047514890724935236,
      "loss": 5.143,
      "mean_token_accuracy": 0.17501790523529054,
      "num_tokens": 35171189.0,
      "step": 16065
    },
    {
      "entropy": 5.5821291446685795,
      "epoch": 1.594878920206431,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004751327183684272,
      "loss": 5.1256,
      "mean_token_accuracy": 0.16781093776226044,
      "num_tokens": 35185034.0,
      "step": 16070
    },
    {
      "entropy": 5.615073919296265,
      "epoch": 1.5953751488685985,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047511652452464196,
      "loss": 5.2363,
      "mean_token_accuracy": 0.16342036724090575,
      "num_tokens": 35196106.0,
      "step": 16075
    },
    {
      "entropy": 5.554545497894287,
      "epoch": 1.595871377530766,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047510032571839816,
      "loss": 5.0854,
      "mean_token_accuracy": 0.176429907977581,
      "num_tokens": 35206736.0,
      "step": 16080
    },
    {
      "entropy": 5.527156639099121,
      "epoch": 1.5963676061929337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004750841219500975,
      "loss": 5.0719,
      "mean_token_accuracy": 0.17986322790384293,
      "num_tokens": 35217279.0,
      "step": 16085
    },
    {
      "entropy": 5.5531919479370115,
      "epoch": 1.5968638348551012,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004750679132201416,
      "loss": 5.0584,
      "mean_token_accuracy": 0.16933271437883377,
      "num_tokens": 35227443.0,
      "step": 16090
    },
    {
      "entropy": 5.546277236938477,
      "epoch": 1.597360063517269,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047505169952893276,
      "loss": 5.1534,
      "mean_token_accuracy": 0.16395957320928572,
      "num_tokens": 35238107.0,
      "step": 16095
    },
    {
      "entropy": 5.553424978256226,
      "epoch": 1.5978562921794364,
      "grad_norm": 1.0,
      "learning_rate": 0.00047503548087687267,
      "loss": 5.1489,
      "mean_token_accuracy": 0.17183574438095092,
      "num_tokens": 35247960.0,
      "step": 16100
    },
    {
      "entropy": 5.626528882980347,
      "epoch": 1.5983525208416038,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047501925726436376,
      "loss": 5.2364,
      "mean_token_accuracy": 0.1654413089156151,
      "num_tokens": 35259560.0,
      "step": 16105
    },
    {
      "entropy": 5.575746488571167,
      "epoch": 1.5988487495037713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004750030286918082,
      "loss": 5.0874,
      "mean_token_accuracy": 0.17274938449263572,
      "num_tokens": 35270363.0,
      "step": 16110
    },
    {
      "entropy": 5.624682760238647,
      "epoch": 1.5993449781659388,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004749867951596086,
      "loss": 5.1367,
      "mean_token_accuracy": 0.17503493875265122,
      "num_tokens": 35281282.0,
      "step": 16115
    },
    {
      "entropy": 5.6115968227386475,
      "epoch": 1.5998412068281063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047497055666816736,
      "loss": 5.1472,
      "mean_token_accuracy": 0.17958660572767257,
      "num_tokens": 35291916.0,
      "step": 16120
    },
    {
      "entropy": 5.580826234817505,
      "epoch": 1.600337435490274,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004749543132178873,
      "loss": 5.1622,
      "mean_token_accuracy": 0.16669413149356843,
      "num_tokens": 35302493.0,
      "step": 16125
    },
    {
      "entropy": 5.57397141456604,
      "epoch": 1.6008336641524414,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000474938064809171,
      "loss": 5.1057,
      "mean_token_accuracy": 0.17403149753808975,
      "num_tokens": 35313364.0,
      "step": 16130
    },
    {
      "entropy": 5.556273984909057,
      "epoch": 1.6013298928146091,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004749218114424216,
      "loss": 5.0698,
      "mean_token_accuracy": 0.17274512201547623,
      "num_tokens": 35324072.0,
      "step": 16135
    },
    {
      "entropy": 5.555614757537842,
      "epoch": 1.6018261214767766,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047490555311804206,
      "loss": 5.1524,
      "mean_token_accuracy": 0.17045714557170868,
      "num_tokens": 35335290.0,
      "step": 16140
    },
    {
      "entropy": 5.536255025863648,
      "epoch": 1.602322350138944,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047488928983643565,
      "loss": 5.1505,
      "mean_token_accuracy": 0.17436153292655945,
      "num_tokens": 35347189.0,
      "step": 16145
    },
    {
      "entropy": 5.583349514007568,
      "epoch": 1.6028185788011116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047487302159800545,
      "loss": 5.158,
      "mean_token_accuracy": 0.17244450896978378,
      "num_tokens": 35360095.0,
      "step": 16150
    },
    {
      "entropy": 5.581042337417602,
      "epoch": 1.603314807463279,
      "grad_norm": 1.015625,
      "learning_rate": 0.000474856748403155,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17593439817428588,
      "num_tokens": 35370075.0,
      "step": 16155
    },
    {
      "entropy": 5.534018850326538,
      "epoch": 1.6038110361254465,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004748404702522879,
      "loss": 5.0682,
      "mean_token_accuracy": 0.1788652628660202,
      "num_tokens": 35381647.0,
      "step": 16160
    },
    {
      "entropy": 5.572353219985962,
      "epoch": 1.604307264787614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047482418714580774,
      "loss": 5.0599,
      "mean_token_accuracy": 0.18118226379156113,
      "num_tokens": 35391629.0,
      "step": 16165
    },
    {
      "entropy": 5.55161862373352,
      "epoch": 1.6048034934497817,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748078990841183,
      "loss": 5.1076,
      "mean_token_accuracy": 0.17219713926315308,
      "num_tokens": 35401385.0,
      "step": 16170
    },
    {
      "entropy": 5.541615724563599,
      "epoch": 1.6052997221119492,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004747916060676235,
      "loss": 5.0752,
      "mean_token_accuracy": 0.17676389813423157,
      "num_tokens": 35414289.0,
      "step": 16175
    },
    {
      "entropy": 5.509043025970459,
      "epoch": 1.6057959507741169,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047477530809672747,
      "loss": 5.0796,
      "mean_token_accuracy": 0.172505983710289,
      "num_tokens": 35424397.0,
      "step": 16180
    },
    {
      "entropy": 5.6356611251831055,
      "epoch": 1.6062921794362843,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004747590051718342,
      "loss": 5.1744,
      "mean_token_accuracy": 0.17094643115997316,
      "num_tokens": 35435897.0,
      "step": 16185
    },
    {
      "entropy": 5.572555446624756,
      "epoch": 1.6067884080984518,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047474269729334806,
      "loss": 5.1117,
      "mean_token_accuracy": 0.17324193716049194,
      "num_tokens": 35446794.0,
      "step": 16190
    },
    {
      "entropy": 5.526512336730957,
      "epoch": 1.6072846367606193,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047472638446167346,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17387555688619613,
      "num_tokens": 35456854.0,
      "step": 16195
    },
    {
      "entropy": 5.532343435287475,
      "epoch": 1.6077808654227868,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047471006667721484,
      "loss": 5.1296,
      "mean_token_accuracy": 0.1723683401942253,
      "num_tokens": 35467924.0,
      "step": 16200
    },
    {
      "entropy": 5.5058375835418705,
      "epoch": 1.6082770940849542,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000474693743940377,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1795547768473625,
      "num_tokens": 35479290.0,
      "step": 16205
    },
    {
      "entropy": 5.599468183517456,
      "epoch": 1.608773322747122,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047467741625156453,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17631965726614,
      "num_tokens": 35489477.0,
      "step": 16210
    },
    {
      "entropy": 5.656747817993164,
      "epoch": 1.6092695514092894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047466108361118235,
      "loss": 5.2642,
      "mean_token_accuracy": 0.16146941781044005,
      "num_tokens": 35501531.0,
      "step": 16215
    },
    {
      "entropy": 5.6939257144927975,
      "epoch": 1.6097657800714569,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047464474601963564,
      "loss": 5.1802,
      "mean_token_accuracy": 0.16374532133340836,
      "num_tokens": 35513325.0,
      "step": 16220
    },
    {
      "entropy": 5.592284107208252,
      "epoch": 1.6102620087336246,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004746284034773294,
      "loss": 5.1091,
      "mean_token_accuracy": 0.17297792285680771,
      "num_tokens": 35524683.0,
      "step": 16225
    },
    {
      "entropy": 5.459389066696167,
      "epoch": 1.610758237395792,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004746120559846689,
      "loss": 5.0344,
      "mean_token_accuracy": 0.18217969536781312,
      "num_tokens": 35534179.0,
      "step": 16230
    },
    {
      "entropy": 5.51647834777832,
      "epoch": 1.6112544660579595,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047459570354205954,
      "loss": 5.0921,
      "mean_token_accuracy": 0.17718457281589509,
      "num_tokens": 35544894.0,
      "step": 16235
    },
    {
      "entropy": 5.5421209812164305,
      "epoch": 1.611750694720127,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047457934614990684,
      "loss": 5.1076,
      "mean_token_accuracy": 0.1775764063000679,
      "num_tokens": 35556915.0,
      "step": 16240
    },
    {
      "entropy": 5.5625669956207275,
      "epoch": 1.6122469233822945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745629838086164,
      "loss": 5.1295,
      "mean_token_accuracy": 0.17911574691534043,
      "num_tokens": 35568037.0,
      "step": 16245
    },
    {
      "entropy": 5.626315593719482,
      "epoch": 1.612743152044462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047454661651859394,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16666755229234695,
      "num_tokens": 35578969.0,
      "step": 16250
    },
    {
      "entropy": 5.582198667526245,
      "epoch": 1.6132393807066296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004745302442802454,
      "loss": 5.0583,
      "mean_token_accuracy": 0.17727223932743072,
      "num_tokens": 35588637.0,
      "step": 16255
    },
    {
      "entropy": 5.543774366378784,
      "epoch": 1.6137356093687971,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004745138670939768,
      "loss": 5.1714,
      "mean_token_accuracy": 0.17284202426671982,
      "num_tokens": 35600865.0,
      "step": 16260
    },
    {
      "entropy": 5.5923093318939205,
      "epoch": 1.6142318380309648,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004744974849601941,
      "loss": 5.1676,
      "mean_token_accuracy": 0.16652936190366746,
      "num_tokens": 35612440.0,
      "step": 16265
    },
    {
      "entropy": 5.623034811019897,
      "epoch": 1.6147280666931323,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004744810978793038,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17386336475610734,
      "num_tokens": 35623143.0,
      "step": 16270
    },
    {
      "entropy": 5.606679105758667,
      "epoch": 1.6152242953552998,
      "grad_norm": 0.875,
      "learning_rate": 0.00047446470585171206,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17446010708808898,
      "num_tokens": 35636245.0,
      "step": 16275
    },
    {
      "entropy": 5.604040622711182,
      "epoch": 1.6157205240174672,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047444830887782537,
      "loss": 5.1724,
      "mean_token_accuracy": 0.17306607514619826,
      "num_tokens": 35646954.0,
      "step": 16280
    },
    {
      "entropy": 5.555650949478149,
      "epoch": 1.6162167526796347,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004744319069580504,
      "loss": 5.1039,
      "mean_token_accuracy": 0.18221383839845656,
      "num_tokens": 35657822.0,
      "step": 16285
    },
    {
      "entropy": 5.56492805480957,
      "epoch": 1.6167129813418022,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047441550009279395,
      "loss": 5.1358,
      "mean_token_accuracy": 0.17558741122484206,
      "num_tokens": 35669306.0,
      "step": 16290
    },
    {
      "entropy": 5.54651255607605,
      "epoch": 1.6172092100039697,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004743990882824627,
      "loss": 5.1483,
      "mean_token_accuracy": 0.1735636457800865,
      "num_tokens": 35681701.0,
      "step": 16295
    },
    {
      "entropy": 5.6021161556243895,
      "epoch": 1.6177054386661374,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047438267152746376,
      "loss": 5.1072,
      "mean_token_accuracy": 0.1743927091360092,
      "num_tokens": 35692561.0,
      "step": 16300
    },
    {
      "entropy": 5.629041910171509,
      "epoch": 1.6182016673283048,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047436624982820423,
      "loss": 5.1839,
      "mean_token_accuracy": 0.16592070013284682,
      "num_tokens": 35703097.0,
      "step": 16305
    },
    {
      "entropy": 5.572886419296265,
      "epoch": 1.6186978959904725,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004743498231850913,
      "loss": 5.0895,
      "mean_token_accuracy": 0.18045682460069656,
      "num_tokens": 35713010.0,
      "step": 16310
    },
    {
      "entropy": 5.622765064239502,
      "epoch": 1.61919412465264,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004743333915985323,
      "loss": 5.2397,
      "mean_token_accuracy": 0.16977233588695526,
      "num_tokens": 35724240.0,
      "step": 16315
    },
    {
      "entropy": 5.6329587459564205,
      "epoch": 1.6196903533148075,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047431695506893476,
      "loss": 5.1935,
      "mean_token_accuracy": 0.17089924961328506,
      "num_tokens": 35735022.0,
      "step": 16320
    },
    {
      "entropy": 5.648626184463501,
      "epoch": 1.620186581976975,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047430051359670616,
      "loss": 5.2607,
      "mean_token_accuracy": 0.16656603664159775,
      "num_tokens": 35745994.0,
      "step": 16325
    },
    {
      "entropy": 5.635349321365356,
      "epoch": 1.6206828106391424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004742840671822542,
      "loss": 5.1747,
      "mean_token_accuracy": 0.1700363501906395,
      "num_tokens": 35756611.0,
      "step": 16330
    },
    {
      "entropy": 5.565572595596313,
      "epoch": 1.62117903930131,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000474267615825987,
      "loss": 5.0875,
      "mean_token_accuracy": 0.17681076675653457,
      "num_tokens": 35767155.0,
      "step": 16335
    },
    {
      "entropy": 5.531439876556396,
      "epoch": 1.6216752679634776,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047425115952831213,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1808216616511345,
      "num_tokens": 35777763.0,
      "step": 16340
    },
    {
      "entropy": 5.460783815383911,
      "epoch": 1.622171496625645,
      "grad_norm": 0.953125,
      "learning_rate": 0.000474234698289638,
      "loss": 5.0359,
      "mean_token_accuracy": 0.17567408829927444,
      "num_tokens": 35787730.0,
      "step": 16345
    },
    {
      "entropy": 5.59519362449646,
      "epoch": 1.6226677252878128,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004742182321103725,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17258143275976182,
      "num_tokens": 35798959.0,
      "step": 16350
    },
    {
      "entropy": 5.577343940734863,
      "epoch": 1.6231639539499803,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004742017609909242,
      "loss": 4.9978,
      "mean_token_accuracy": 0.18074162304401398,
      "num_tokens": 35808426.0,
      "step": 16355
    },
    {
      "entropy": 5.4927576065063475,
      "epoch": 1.6236601826121477,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047418528493170157,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18068450093269348,
      "num_tokens": 35818798.0,
      "step": 16360
    },
    {
      "entropy": 5.612867832183838,
      "epoch": 1.6241564112743152,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004741688039331129,
      "loss": 5.1968,
      "mean_token_accuracy": 0.17021633982658385,
      "num_tokens": 35830429.0,
      "step": 16365
    },
    {
      "entropy": 5.623977994918823,
      "epoch": 1.6246526399364827,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047415231799556725,
      "loss": 5.116,
      "mean_token_accuracy": 0.1702694982290268,
      "num_tokens": 35840834.0,
      "step": 16370
    },
    {
      "entropy": 5.55428900718689,
      "epoch": 1.6251488685986502,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047413582711947305,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17516877502202988,
      "num_tokens": 35850889.0,
      "step": 16375
    },
    {
      "entropy": 5.470477104187012,
      "epoch": 1.6256450972608176,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047411933130523954,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1820181518793106,
      "num_tokens": 35862491.0,
      "step": 16380
    },
    {
      "entropy": 5.588985395431519,
      "epoch": 1.6261413259229853,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004741028305532757,
      "loss": 5.1455,
      "mean_token_accuracy": 0.17194889932870866,
      "num_tokens": 35873567.0,
      "step": 16385
    },
    {
      "entropy": 5.5500946044921875,
      "epoch": 1.6266375545851528,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004740863248639907,
      "loss": 5.0728,
      "mean_token_accuracy": 0.17689508944749832,
      "num_tokens": 35884970.0,
      "step": 16390
    },
    {
      "entropy": 5.643901062011719,
      "epoch": 1.6271337832473205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047406981423779376,
      "loss": 5.157,
      "mean_token_accuracy": 0.17451009452342986,
      "num_tokens": 35895352.0,
      "step": 16395
    },
    {
      "entropy": 5.629940223693848,
      "epoch": 1.627630011909488,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047405329867509447,
      "loss": 5.1876,
      "mean_token_accuracy": 0.16624921858310698,
      "num_tokens": 35908224.0,
      "step": 16400
    },
    {
      "entropy": 5.50414252281189,
      "epoch": 1.6281262405716554,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047403677817630226,
      "loss": 4.9989,
      "mean_token_accuracy": 0.17750672549009322,
      "num_tokens": 35918484.0,
      "step": 16405
    },
    {
      "entropy": 5.517140007019043,
      "epoch": 1.628622469233823,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004740202527418269,
      "loss": 5.0554,
      "mean_token_accuracy": 0.17610628306865692,
      "num_tokens": 35929682.0,
      "step": 16410
    },
    {
      "entropy": 5.579895067214966,
      "epoch": 1.6291186978959904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000474003722372078,
      "loss": 5.1097,
      "mean_token_accuracy": 0.1774996355175972,
      "num_tokens": 35939692.0,
      "step": 16415
    },
    {
      "entropy": 5.607784938812256,
      "epoch": 1.6296149265581579,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047398718706746556,
      "loss": 5.1584,
      "mean_token_accuracy": 0.17234123647212982,
      "num_tokens": 35950871.0,
      "step": 16420
    },
    {
      "entropy": 5.597295951843262,
      "epoch": 1.6301111552203256,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004739706468283998,
      "loss": 5.1936,
      "mean_token_accuracy": 0.16855676472187042,
      "num_tokens": 35962292.0,
      "step": 16425
    },
    {
      "entropy": 5.565132093429566,
      "epoch": 1.630607383882493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047395410165529075,
      "loss": 5.0743,
      "mean_token_accuracy": 0.1796128824353218,
      "num_tokens": 35973838.0,
      "step": 16430
    },
    {
      "entropy": 5.558991241455078,
      "epoch": 1.6311036125446607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047393755154854855,
      "loss": 5.0976,
      "mean_token_accuracy": 0.17434774786233903,
      "num_tokens": 35984309.0,
      "step": 16435
    },
    {
      "entropy": 5.533134317398071,
      "epoch": 1.6315998412068282,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004739209965085839,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17168208658695222,
      "num_tokens": 35994200.0,
      "step": 16440
    },
    {
      "entropy": 5.644713592529297,
      "epoch": 1.6320960698689957,
      "grad_norm": 1.078125,
      "learning_rate": 0.000473904436535807,
      "loss": 5.2274,
      "mean_token_accuracy": 0.16801096349954606,
      "num_tokens": 36004839.0,
      "step": 16445
    },
    {
      "entropy": 5.528043985366821,
      "epoch": 1.6325922985311632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004738878716306288,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17001353949308395,
      "num_tokens": 36015422.0,
      "step": 16450
    },
    {
      "entropy": 5.640442943572998,
      "epoch": 1.6330885271933306,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047387130179345996,
      "loss": 5.2048,
      "mean_token_accuracy": 0.16631716191768647,
      "num_tokens": 36027453.0,
      "step": 16455
    },
    {
      "entropy": 5.60477123260498,
      "epoch": 1.6335847558554981,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047385472702471133,
      "loss": 5.2281,
      "mean_token_accuracy": 0.17367670238018035,
      "num_tokens": 36039671.0,
      "step": 16460
    },
    {
      "entropy": 5.586149883270264,
      "epoch": 1.6340809845176656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000473838147324794,
      "loss": 5.1593,
      "mean_token_accuracy": 0.17037093192338942,
      "num_tokens": 36053006.0,
      "step": 16465
    },
    {
      "entropy": 5.59634370803833,
      "epoch": 1.6345772131798333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000473821562694119,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17335128486156465,
      "num_tokens": 36063764.0,
      "step": 16470
    },
    {
      "entropy": 5.550228023529053,
      "epoch": 1.6350734418420008,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738049731330977,
      "loss": 5.0754,
      "mean_token_accuracy": 0.17379796355962754,
      "num_tokens": 36073449.0,
      "step": 16475
    },
    {
      "entropy": 5.595335245132446,
      "epoch": 1.6355696705041685,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004737883786421414,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1728828951716423,
      "num_tokens": 36084912.0,
      "step": 16480
    },
    {
      "entropy": 5.484679937362671,
      "epoch": 1.636065899166336,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047377177922166175,
      "loss": 5.0385,
      "mean_token_accuracy": 0.1770404428243637,
      "num_tokens": 36095031.0,
      "step": 16485
    },
    {
      "entropy": 5.554383611679077,
      "epoch": 1.6365621278285034,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047375517487207024,
      "loss": 5.1153,
      "mean_token_accuracy": 0.17857690900564194,
      "num_tokens": 36106689.0,
      "step": 16490
    },
    {
      "entropy": 5.553959465026855,
      "epoch": 1.6370583564906709,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047373856559377876,
      "loss": 5.0096,
      "mean_token_accuracy": 0.17579258233308792,
      "num_tokens": 36118043.0,
      "step": 16495
    },
    {
      "entropy": 5.55500340461731,
      "epoch": 1.6375545851528384,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004737219513871989,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17625396996736525,
      "num_tokens": 36128155.0,
      "step": 16500
    },
    {
      "entropy": 5.594151496887207,
      "epoch": 1.6380508138150058,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047370533225274305,
      "loss": 5.1217,
      "mean_token_accuracy": 0.1780672252178192,
      "num_tokens": 36139377.0,
      "step": 16505
    },
    {
      "entropy": 5.533447742462158,
      "epoch": 1.6385470424771735,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004736887081908231,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17905569225549697,
      "num_tokens": 36149575.0,
      "step": 16510
    },
    {
      "entropy": 5.5296073913574215,
      "epoch": 1.639043271139341,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004736720792018513,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17149801701307296,
      "num_tokens": 36161707.0,
      "step": 16515
    },
    {
      "entropy": 5.541195917129516,
      "epoch": 1.6395394998015085,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004736554452862399,
      "loss": 5.1105,
      "mean_token_accuracy": 0.17418026328086852,
      "num_tokens": 36173095.0,
      "step": 16520
    },
    {
      "entropy": 5.6272642612457275,
      "epoch": 1.6400357284636762,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047363880644440173,
      "loss": 5.1102,
      "mean_token_accuracy": 0.17875873893499375,
      "num_tokens": 36184895.0,
      "step": 16525
    },
    {
      "entropy": 5.512523937225342,
      "epoch": 1.6405319571258437,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047362216267674906,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17265758514404297,
      "num_tokens": 36195277.0,
      "step": 16530
    },
    {
      "entropy": 5.546697044372559,
      "epoch": 1.6410281857880111,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004736055139836948,
      "loss": 5.122,
      "mean_token_accuracy": 0.1697360932826996,
      "num_tokens": 36207555.0,
      "step": 16535
    },
    {
      "entropy": 5.533960628509521,
      "epoch": 1.6415244144501786,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047358886036565176,
      "loss": 5.0433,
      "mean_token_accuracy": 0.18365218192338945,
      "num_tokens": 36217632.0,
      "step": 16540
    },
    {
      "entropy": 5.586761045455932,
      "epoch": 1.642020643112346,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047357220182303285,
      "loss": 5.1164,
      "mean_token_accuracy": 0.17051098644733428,
      "num_tokens": 36227368.0,
      "step": 16545
    },
    {
      "entropy": 5.586897945404052,
      "epoch": 1.6425168717745136,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004735555383562513,
      "loss": 5.1791,
      "mean_token_accuracy": 0.1705404996871948,
      "num_tokens": 36237932.0,
      "step": 16550
    },
    {
      "entropy": 5.579228401184082,
      "epoch": 1.6430131004366813,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047353886996572014,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17887124121189119,
      "num_tokens": 36249365.0,
      "step": 16555
    },
    {
      "entropy": 5.6265472888946535,
      "epoch": 1.6435093290988487,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047352219665185293,
      "loss": 5.1369,
      "mean_token_accuracy": 0.16682802587747575,
      "num_tokens": 36259254.0,
      "step": 16560
    },
    {
      "entropy": 5.543618631362915,
      "epoch": 1.6440055577610164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004735055184150629,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17507602870464326,
      "num_tokens": 36269249.0,
      "step": 16565
    },
    {
      "entropy": 5.602028656005859,
      "epoch": 1.644501786423184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047348883525576385,
      "loss": 5.1684,
      "mean_token_accuracy": 0.170199853181839,
      "num_tokens": 36280275.0,
      "step": 16570
    },
    {
      "entropy": 5.69281907081604,
      "epoch": 1.6449980150853514,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004734721471743694,
      "loss": 5.2131,
      "mean_token_accuracy": 0.1721682772040367,
      "num_tokens": 36292750.0,
      "step": 16575
    },
    {
      "entropy": 5.5812554359436035,
      "epoch": 1.6454942437475188,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004734554541712933,
      "loss": 5.128,
      "mean_token_accuracy": 0.16716381162405014,
      "num_tokens": 36303148.0,
      "step": 16580
    },
    {
      "entropy": 5.565799760818481,
      "epoch": 1.6459904724096863,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004734387562469496,
      "loss": 5.1362,
      "mean_token_accuracy": 0.1725514754652977,
      "num_tokens": 36314939.0,
      "step": 16585
    },
    {
      "entropy": 5.601307487487793,
      "epoch": 1.6464867010718538,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047342205340175235,
      "loss": 5.2029,
      "mean_token_accuracy": 0.17023302465677262,
      "num_tokens": 36327036.0,
      "step": 16590
    },
    {
      "entropy": 5.545242404937744,
      "epoch": 1.6469829297340213,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047340534563611574,
      "loss": 5.0423,
      "mean_token_accuracy": 0.17782391756772994,
      "num_tokens": 36336844.0,
      "step": 16595
    },
    {
      "entropy": 5.550745058059692,
      "epoch": 1.647479158396189,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047338863295045414,
      "loss": 5.0578,
      "mean_token_accuracy": 0.17292785346508027,
      "num_tokens": 36347538.0,
      "step": 16600
    },
    {
      "entropy": 5.500924587249756,
      "epoch": 1.6479753870583564,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047337191534518186,
      "loss": 5.0465,
      "mean_token_accuracy": 0.17671357244253158,
      "num_tokens": 36358221.0,
      "step": 16605
    },
    {
      "entropy": 5.588835430145264,
      "epoch": 1.6484716157205241,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004733551928207135,
      "loss": 5.0758,
      "mean_token_accuracy": 0.17644811123609544,
      "num_tokens": 36368756.0,
      "step": 16610
    },
    {
      "entropy": 5.560000324249268,
      "epoch": 1.6489678443826916,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004733384653774639,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17677556723356247,
      "num_tokens": 36379201.0,
      "step": 16615
    },
    {
      "entropy": 5.569677352905273,
      "epoch": 1.649464073044859,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047332173301584764,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17665245682001113,
      "num_tokens": 36389375.0,
      "step": 16620
    },
    {
      "entropy": 5.587802410125732,
      "epoch": 1.6499603017070266,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047330499573627975,
      "loss": 5.1538,
      "mean_token_accuracy": 0.16708672493696214,
      "num_tokens": 36400064.0,
      "step": 16625
    },
    {
      "entropy": 5.587437152862549,
      "epoch": 1.650456530369194,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047328825353917526,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1845708653330803,
      "num_tokens": 36410165.0,
      "step": 16630
    },
    {
      "entropy": 5.578358745574951,
      "epoch": 1.6509527590313615,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047327150642494944,
      "loss": 5.1246,
      "mean_token_accuracy": 0.1669653832912445,
      "num_tokens": 36421600.0,
      "step": 16635
    },
    {
      "entropy": 5.5945902347564695,
      "epoch": 1.6514489876935292,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004732547543940174,
      "loss": 5.139,
      "mean_token_accuracy": 0.17186530083417892,
      "num_tokens": 36433439.0,
      "step": 16640
    },
    {
      "entropy": 5.552958011627197,
      "epoch": 1.6519452163556967,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004732379974467947,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17120419591665267,
      "num_tokens": 36444001.0,
      "step": 16645
    },
    {
      "entropy": 5.598403549194336,
      "epoch": 1.6524414450178644,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004732212355836968,
      "loss": 5.1712,
      "mean_token_accuracy": 0.16561972051858903,
      "num_tokens": 36454911.0,
      "step": 16650
    },
    {
      "entropy": 5.5529550075531,
      "epoch": 1.6529376736800319,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004732044688051395,
      "loss": 5.114,
      "mean_token_accuracy": 0.17295146286487578,
      "num_tokens": 36465537.0,
      "step": 16655
    },
    {
      "entropy": 5.592068862915039,
      "epoch": 1.6534339023421993,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047318769711153833,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17134109288454055,
      "num_tokens": 36476487.0,
      "step": 16660
    },
    {
      "entropy": 5.59166522026062,
      "epoch": 1.6539301310043668,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004731709205033094,
      "loss": 5.1682,
      "mean_token_accuracy": 0.17180057168006896,
      "num_tokens": 36486447.0,
      "step": 16665
    },
    {
      "entropy": 5.550963354110718,
      "epoch": 1.6544263596665343,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047315413898086863,
      "loss": 5.1594,
      "mean_token_accuracy": 0.17448064535856248,
      "num_tokens": 36497704.0,
      "step": 16670
    },
    {
      "entropy": 5.653050184249878,
      "epoch": 1.6549225883287018,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004731373525446322,
      "loss": 5.1641,
      "mean_token_accuracy": 0.1662793904542923,
      "num_tokens": 36508369.0,
      "step": 16675
    },
    {
      "entropy": 5.685352468490601,
      "epoch": 1.6554188169908692,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004731205611950164,
      "loss": 5.1857,
      "mean_token_accuracy": 0.16735468357801436,
      "num_tokens": 36519766.0,
      "step": 16680
    },
    {
      "entropy": 5.564163064956665,
      "epoch": 1.655915045653037,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047310376493243754,
      "loss": 5.107,
      "mean_token_accuracy": 0.17748954743146897,
      "num_tokens": 36531022.0,
      "step": 16685
    },
    {
      "entropy": 5.559226369857788,
      "epoch": 1.6564112743152044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004730869637573122,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17598967999219894,
      "num_tokens": 36542325.0,
      "step": 16690
    },
    {
      "entropy": 5.621406698226929,
      "epoch": 1.656907502977372,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047307015767005705,
      "loss": 5.1535,
      "mean_token_accuracy": 0.17138733118772506,
      "num_tokens": 36553671.0,
      "step": 16695
    },
    {
      "entropy": 5.630747699737549,
      "epoch": 1.6574037316395396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004730533466710888,
      "loss": 5.1008,
      "mean_token_accuracy": 0.1697777032852173,
      "num_tokens": 36564835.0,
      "step": 16700
    },
    {
      "entropy": 5.5523875713348385,
      "epoch": 1.657899960301707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004730365307608242,
      "loss": 5.064,
      "mean_token_accuracy": 0.1758968099951744,
      "num_tokens": 36575793.0,
      "step": 16705
    },
    {
      "entropy": 5.59416766166687,
      "epoch": 1.6583961889638745,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047301970993968054,
      "loss": 5.1619,
      "mean_token_accuracy": 0.1749774694442749,
      "num_tokens": 36587097.0,
      "step": 16710
    },
    {
      "entropy": 5.552208995819091,
      "epoch": 1.658892417626042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047300288420807463,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17299058884382248,
      "num_tokens": 36598609.0,
      "step": 16715
    },
    {
      "entropy": 5.591580390930176,
      "epoch": 1.6593886462882095,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047298605356642387,
      "loss": 5.0909,
      "mean_token_accuracy": 0.1709992527961731,
      "num_tokens": 36609476.0,
      "step": 16720
    },
    {
      "entropy": 5.580354928970337,
      "epoch": 1.6598848749503772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047296921801514565,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17704582810401917,
      "num_tokens": 36620160.0,
      "step": 16725
    },
    {
      "entropy": 5.560772275924682,
      "epoch": 1.6603811036125447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004729523775546574,
      "loss": 5.0364,
      "mean_token_accuracy": 0.17715616673231124,
      "num_tokens": 36630742.0,
      "step": 16730
    },
    {
      "entropy": 5.551728963851929,
      "epoch": 1.6608773322747123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004729355321853767,
      "loss": 5.0485,
      "mean_token_accuracy": 0.18107294142246247,
      "num_tokens": 36640980.0,
      "step": 16735
    },
    {
      "entropy": 5.572649431228638,
      "epoch": 1.6613735609368798,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004729186819077214,
      "loss": 5.1222,
      "mean_token_accuracy": 0.17493440359830856,
      "num_tokens": 36651703.0,
      "step": 16740
    },
    {
      "entropy": 5.544734954833984,
      "epoch": 1.6618697895990473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004729018267221092,
      "loss": 5.0705,
      "mean_token_accuracy": 0.18352148532867432,
      "num_tokens": 36662362.0,
      "step": 16745
    },
    {
      "entropy": 5.559345722198486,
      "epoch": 1.6623660182612148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004728849666289581,
      "loss": 5.0243,
      "mean_token_accuracy": 0.1798667535185814,
      "num_tokens": 36671775.0,
      "step": 16750
    },
    {
      "entropy": 5.514663314819336,
      "epoch": 1.6628622469233822,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047286810162868625,
      "loss": 5.0907,
      "mean_token_accuracy": 0.17507643699645997,
      "num_tokens": 36683183.0,
      "step": 16755
    },
    {
      "entropy": 5.577113008499145,
      "epoch": 1.6633584755855497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004728512317217119,
      "loss": 5.0786,
      "mean_token_accuracy": 0.1744253545999527,
      "num_tokens": 36694106.0,
      "step": 16760
    },
    {
      "entropy": 5.558796072006226,
      "epoch": 1.6638547042477172,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047283435690845336,
      "loss": 5.0878,
      "mean_token_accuracy": 0.18129507601261138,
      "num_tokens": 36704607.0,
      "step": 16765
    },
    {
      "entropy": 5.574303150177002,
      "epoch": 1.664350932909885,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047281747718932906,
      "loss": 5.0802,
      "mean_token_accuracy": 0.17347227782011032,
      "num_tokens": 36715330.0,
      "step": 16770
    },
    {
      "entropy": 5.567510461807251,
      "epoch": 1.6648471615720524,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004728005925647575,
      "loss": 5.1407,
      "mean_token_accuracy": 0.17008682638406752,
      "num_tokens": 36727238.0,
      "step": 16775
    },
    {
      "entropy": 5.569040441513062,
      "epoch": 1.66534339023422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047278370303515744,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17286260575056075,
      "num_tokens": 36737355.0,
      "step": 16780
    },
    {
      "entropy": 5.511401844024658,
      "epoch": 1.6658396188963875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004727668086009478,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18323722034692763,
      "num_tokens": 36748331.0,
      "step": 16785
    },
    {
      "entropy": 5.535847234725952,
      "epoch": 1.666335847558555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047274990926254753,
      "loss": 5.0687,
      "mean_token_accuracy": 0.17829849123954772,
      "num_tokens": 36758999.0,
      "step": 16790
    },
    {
      "entropy": 5.529321622848511,
      "epoch": 1.6668320762207225,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004727330050203755,
      "loss": 5.0316,
      "mean_token_accuracy": 0.18225447237491607,
      "num_tokens": 36771646.0,
      "step": 16795
    },
    {
      "entropy": 5.494856834411621,
      "epoch": 1.66732830488289,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004727160958748511,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17669889479875564,
      "num_tokens": 36781670.0,
      "step": 16800
    },
    {
      "entropy": 5.540293407440186,
      "epoch": 1.6678245335450574,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047269918182639355,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17170077562332153,
      "num_tokens": 36792664.0,
      "step": 16805
    },
    {
      "entropy": 5.515321540832519,
      "epoch": 1.6683207622072251,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004726822628754222,
      "loss": 5.0207,
      "mean_token_accuracy": 0.18309032618999482,
      "num_tokens": 36803551.0,
      "step": 16810
    },
    {
      "entropy": 5.510124826431275,
      "epoch": 1.6688169908693926,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047266533902235683,
      "loss": 5.0971,
      "mean_token_accuracy": 0.17418603003025054,
      "num_tokens": 36818202.0,
      "step": 16815
    },
    {
      "entropy": 5.592900896072388,
      "epoch": 1.66931321953156,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004726484102676169,
      "loss": 5.1368,
      "mean_token_accuracy": 0.17444814294576644,
      "num_tokens": 36829861.0,
      "step": 16820
    },
    {
      "entropy": 5.612976884841919,
      "epoch": 1.6698094481937278,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004726314766116223,
      "loss": 5.1218,
      "mean_token_accuracy": 0.1774122893810272,
      "num_tokens": 36841115.0,
      "step": 16825
    },
    {
      "entropy": 5.5587742805480955,
      "epoch": 1.6703056768558953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000472614538054793,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17235881537199021,
      "num_tokens": 36851176.0,
      "step": 16830
    },
    {
      "entropy": 5.527522468566895,
      "epoch": 1.6708019055180627,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047259759459754893,
      "loss": 5.0807,
      "mean_token_accuracy": 0.17136659026145934,
      "num_tokens": 36861408.0,
      "step": 16835
    },
    {
      "entropy": 5.549653816223144,
      "epoch": 1.6712981341802302,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004725806462403103,
      "loss": 5.0331,
      "mean_token_accuracy": 0.18007284551858901,
      "num_tokens": 36871244.0,
      "step": 16840
    },
    {
      "entropy": 5.578930282592774,
      "epoch": 1.6717943628423977,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047256369298349736,
      "loss": 5.1354,
      "mean_token_accuracy": 0.1650789648294449,
      "num_tokens": 36882955.0,
      "step": 16845
    },
    {
      "entropy": 5.563187932968139,
      "epoch": 1.6722905915045652,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047254673482753065,
      "loss": 5.0439,
      "mean_token_accuracy": 0.1804004654288292,
      "num_tokens": 36893118.0,
      "step": 16850
    },
    {
      "entropy": 5.613183879852295,
      "epoch": 1.6727868201667329,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047252977177283045,
      "loss": 5.242,
      "mean_token_accuracy": 0.17157860547304155,
      "num_tokens": 36903484.0,
      "step": 16855
    },
    {
      "entropy": 5.539911937713623,
      "epoch": 1.6732830488289003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047251280381981764,
      "loss": 5.056,
      "mean_token_accuracy": 0.1793104290962219,
      "num_tokens": 36913439.0,
      "step": 16860
    },
    {
      "entropy": 5.608049058914185,
      "epoch": 1.673779277491068,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004724958309689129,
      "loss": 5.1588,
      "mean_token_accuracy": 0.17076359689235687,
      "num_tokens": 36925977.0,
      "step": 16865
    },
    {
      "entropy": 5.57463231086731,
      "epoch": 1.6742755061532355,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047247885322053705,
      "loss": 5.083,
      "mean_token_accuracy": 0.1750575751066208,
      "num_tokens": 36937318.0,
      "step": 16870
    },
    {
      "entropy": 5.607946348190308,
      "epoch": 1.674771734815403,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047246187057511127,
      "loss": 5.1179,
      "mean_token_accuracy": 0.17237639427185059,
      "num_tokens": 36948199.0,
      "step": 16875
    },
    {
      "entropy": 5.570090913772583,
      "epoch": 1.6752679634775705,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004724448830330565,
      "loss": 5.0981,
      "mean_token_accuracy": 0.1698041409254074,
      "num_tokens": 36959813.0,
      "step": 16880
    },
    {
      "entropy": 5.565642881393432,
      "epoch": 1.675764192139738,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047242789059479413,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1777702435851097,
      "num_tokens": 36970548.0,
      "step": 16885
    },
    {
      "entropy": 5.553309583663941,
      "epoch": 1.6762604208019054,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047241089326074547,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1687386855483055,
      "num_tokens": 36983072.0,
      "step": 16890
    },
    {
      "entropy": 5.572058916091919,
      "epoch": 1.6767566494640729,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004723938910313321,
      "loss": 5.1321,
      "mean_token_accuracy": 0.1745602622628212,
      "num_tokens": 36993684.0,
      "step": 16895
    },
    {
      "entropy": 5.5663012027740475,
      "epoch": 1.6772528781262406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047237688390697555,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18135023415088652,
      "num_tokens": 37004776.0,
      "step": 16900
    },
    {
      "entropy": 5.590132761001587,
      "epoch": 1.677749106788408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004723598718880975,
      "loss": 5.0989,
      "mean_token_accuracy": 0.1799945965409279,
      "num_tokens": 37014669.0,
      "step": 16905
    },
    {
      "entropy": 5.540630149841308,
      "epoch": 1.6782453354505757,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047234285497512005,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1800594672560692,
      "num_tokens": 37025645.0,
      "step": 16910
    },
    {
      "entropy": 5.546304082870483,
      "epoch": 1.6787415641127432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047232583316846494,
      "loss": 5.136,
      "mean_token_accuracy": 0.16929339468479157,
      "num_tokens": 37037065.0,
      "step": 16915
    },
    {
      "entropy": 5.5801807880401615,
      "epoch": 1.6792377927749107,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004723088064685544,
      "loss": 5.1098,
      "mean_token_accuracy": 0.17183686494827272,
      "num_tokens": 37047957.0,
      "step": 16920
    },
    {
      "entropy": 5.612519645690918,
      "epoch": 1.6797340214370782,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047229177487581055,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17929104566574097,
      "num_tokens": 37059798.0,
      "step": 16925
    },
    {
      "entropy": 5.629334545135498,
      "epoch": 1.6802302500992456,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047227473839065587,
      "loss": 5.1763,
      "mean_token_accuracy": 0.16922227889299393,
      "num_tokens": 37071229.0,
      "step": 16930
    },
    {
      "entropy": 5.561092281341553,
      "epoch": 1.6807264787614131,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047225769701351284,
      "loss": 5.1469,
      "mean_token_accuracy": 0.17251265943050384,
      "num_tokens": 37082191.0,
      "step": 16935
    },
    {
      "entropy": 5.524011278152466,
      "epoch": 1.6812227074235808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004722406507448038,
      "loss": 5.0339,
      "mean_token_accuracy": 0.18338896185159684,
      "num_tokens": 37092718.0,
      "step": 16940
    },
    {
      "entropy": 5.509593391418457,
      "epoch": 1.6817189360857483,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047222359958495175,
      "loss": 5.0902,
      "mean_token_accuracy": 0.17513153403997422,
      "num_tokens": 37104287.0,
      "step": 16945
    },
    {
      "entropy": 5.562680625915528,
      "epoch": 1.682215164747916,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004722065435343794,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1752450779080391,
      "num_tokens": 37114994.0,
      "step": 16950
    },
    {
      "entropy": 5.592937278747558,
      "epoch": 1.6827113934100835,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004721894825935097,
      "loss": 5.0233,
      "mean_token_accuracy": 0.17567918300628663,
      "num_tokens": 37125258.0,
      "step": 16955
    },
    {
      "entropy": 5.604757308959961,
      "epoch": 1.683207622072251,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004721724167627657,
      "loss": 5.1372,
      "mean_token_accuracy": 0.1759994223713875,
      "num_tokens": 37136055.0,
      "step": 16960
    },
    {
      "entropy": 5.593980312347412,
      "epoch": 1.6837038507344184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004721553460425708,
      "loss": 5.083,
      "mean_token_accuracy": 0.1741182029247284,
      "num_tokens": 37147312.0,
      "step": 16965
    },
    {
      "entropy": 5.6231142520904545,
      "epoch": 1.684200079396586,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000472138270433348,
      "loss": 5.1217,
      "mean_token_accuracy": 0.17713548392057418,
      "num_tokens": 37157940.0,
      "step": 16970
    },
    {
      "entropy": 5.595066547393799,
      "epoch": 1.6846963080587534,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047212118993552097,
      "loss": 5.1524,
      "mean_token_accuracy": 0.16667592376470566,
      "num_tokens": 37169863.0,
      "step": 16975
    },
    {
      "entropy": 5.624895906448364,
      "epoch": 1.6851925367209208,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004721041045495131,
      "loss": 5.1625,
      "mean_token_accuracy": 0.1668195053935051,
      "num_tokens": 37181840.0,
      "step": 16980
    },
    {
      "entropy": 5.57620849609375,
      "epoch": 1.6856887653830885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047208701427574817,
      "loss": 5.0475,
      "mean_token_accuracy": 0.18145481646060943,
      "num_tokens": 37191015.0,
      "step": 16985
    },
    {
      "entropy": 5.546496152877808,
      "epoch": 1.686184994045256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047206991911465,
      "loss": 5.0032,
      "mean_token_accuracy": 0.18017019480466842,
      "num_tokens": 37201829.0,
      "step": 16990
    },
    {
      "entropy": 5.508873796463012,
      "epoch": 1.6866812227074237,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004720528190666425,
      "loss": 5.0669,
      "mean_token_accuracy": 0.16670711934566498,
      "num_tokens": 37213870.0,
      "step": 16995
    },
    {
      "entropy": 5.577471017837524,
      "epoch": 1.6871774513695912,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047203571413214967,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17076486945152283,
      "num_tokens": 37225232.0,
      "step": 17000
    },
    {
      "entropy": 5.631160545349121,
      "epoch": 1.6876736800317587,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004720186043115957,
      "loss": 5.2212,
      "mean_token_accuracy": 0.1647544875741005,
      "num_tokens": 37236329.0,
      "step": 17005
    },
    {
      "entropy": 5.660006427764893,
      "epoch": 1.6881699086939261,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004720014896054048,
      "loss": 5.1982,
      "mean_token_accuracy": 0.17104246467351913,
      "num_tokens": 37247402.0,
      "step": 17010
    },
    {
      "entropy": 5.60420560836792,
      "epoch": 1.6886661373560936,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004719843700140017,
      "loss": 5.133,
      "mean_token_accuracy": 0.16391166150569916,
      "num_tokens": 37258629.0,
      "step": 17015
    },
    {
      "entropy": 5.574168920516968,
      "epoch": 1.689162366018261,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004719672455378105,
      "loss": 5.1126,
      "mean_token_accuracy": 0.177654992043972,
      "num_tokens": 37270154.0,
      "step": 17020
    },
    {
      "entropy": 5.575639152526856,
      "epoch": 1.6896585946804288,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000471950116177256,
      "loss": 5.1286,
      "mean_token_accuracy": 0.17549273520708084,
      "num_tokens": 37280801.0,
      "step": 17025
    },
    {
      "entropy": 5.594921112060547,
      "epoch": 1.6901548233425963,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004719329819327631,
      "loss": 5.1521,
      "mean_token_accuracy": 0.17314542829990387,
      "num_tokens": 37293057.0,
      "step": 17030
    },
    {
      "entropy": 5.60534143447876,
      "epoch": 1.690651052004764,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004719158428047566,
      "loss": 5.1497,
      "mean_token_accuracy": 0.17571026533842088,
      "num_tokens": 37305784.0,
      "step": 17035
    },
    {
      "entropy": 5.570098876953125,
      "epoch": 1.6911472806669314,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004718986987936615,
      "loss": 5.131,
      "mean_token_accuracy": 0.1682923123240471,
      "num_tokens": 37317067.0,
      "step": 17040
    },
    {
      "entropy": 5.632052421569824,
      "epoch": 1.691643509329099,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047188154989990296,
      "loss": 5.1314,
      "mean_token_accuracy": 0.17268729954957962,
      "num_tokens": 37327568.0,
      "step": 17045
    },
    {
      "entropy": 5.617778635025024,
      "epoch": 1.6921397379912664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004718643961239062,
      "loss": 5.1628,
      "mean_token_accuracy": 0.17904388457536696,
      "num_tokens": 37339235.0,
      "step": 17050
    },
    {
      "entropy": 5.562147855758667,
      "epoch": 1.6926359666534339,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047184723746609666,
      "loss": 5.037,
      "mean_token_accuracy": 0.1736544519662857,
      "num_tokens": 37348943.0,
      "step": 17055
    },
    {
      "entropy": 5.549044942855835,
      "epoch": 1.6931321953156013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047183007392689977,
      "loss": 5.111,
      "mean_token_accuracy": 0.17313006073236464,
      "num_tokens": 37359792.0,
      "step": 17060
    },
    {
      "entropy": 5.588190650939941,
      "epoch": 1.6936284239777688,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047181290550674115,
      "loss": 5.1413,
      "mean_token_accuracy": 0.17622416019439696,
      "num_tokens": 37371905.0,
      "step": 17065
    },
    {
      "entropy": 5.600088405609131,
      "epoch": 1.6941246526399365,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004717957322060466,
      "loss": 5.1473,
      "mean_token_accuracy": 0.17087748795747756,
      "num_tokens": 37383262.0,
      "step": 17070
    },
    {
      "entropy": 5.602162885665893,
      "epoch": 1.694620881302104,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004717785540252419,
      "loss": 5.1319,
      "mean_token_accuracy": 0.17709943354129792,
      "num_tokens": 37394666.0,
      "step": 17075
    },
    {
      "entropy": 5.659142255783081,
      "epoch": 1.6951171099642717,
      "grad_norm": 0.875,
      "learning_rate": 0.0004717613709647532,
      "loss": 5.1914,
      "mean_token_accuracy": 0.16900621801614762,
      "num_tokens": 37405860.0,
      "step": 17080
    },
    {
      "entropy": 5.579074287414551,
      "epoch": 1.6956133386264391,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047174418302500635,
      "loss": 5.0594,
      "mean_token_accuracy": 0.18032332211732865,
      "num_tokens": 37416596.0,
      "step": 17085
    },
    {
      "entropy": 5.5833251953125,
      "epoch": 1.6961095672886066,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004717269902064278,
      "loss": 5.1577,
      "mean_token_accuracy": 0.16413833796977997,
      "num_tokens": 37427293.0,
      "step": 17090
    },
    {
      "entropy": 5.527523994445801,
      "epoch": 1.696605795950774,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004717097925094438,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18386131823062896,
      "num_tokens": 37438753.0,
      "step": 17095
    },
    {
      "entropy": 5.603373050689697,
      "epoch": 1.6971020246129416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004716925899344808,
      "loss": 5.074,
      "mean_token_accuracy": 0.1744974821805954,
      "num_tokens": 37449917.0,
      "step": 17100
    },
    {
      "entropy": 5.5579112529754635,
      "epoch": 1.697598253275109,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004716753824819654,
      "loss": 4.997,
      "mean_token_accuracy": 0.18201542943716048,
      "num_tokens": 37460248.0,
      "step": 17105
    },
    {
      "entropy": 5.525997829437256,
      "epoch": 1.6980944819372767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004716581701523243,
      "loss": 5.0642,
      "mean_token_accuracy": 0.1739104837179184,
      "num_tokens": 37470404.0,
      "step": 17110
    },
    {
      "entropy": 5.565442848205566,
      "epoch": 1.6985907105994442,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047164095294598433,
      "loss": 5.1561,
      "mean_token_accuracy": 0.17142199501395225,
      "num_tokens": 37481764.0,
      "step": 17115
    },
    {
      "entropy": 5.6081537246704105,
      "epoch": 1.699086939261612,
      "grad_norm": 1.0,
      "learning_rate": 0.00047162373086337243,
      "loss": 5.0923,
      "mean_token_accuracy": 0.17721329033374786,
      "num_tokens": 37492927.0,
      "step": 17120
    },
    {
      "entropy": 5.692150115966797,
      "epoch": 1.6995831679237794,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047160650390491574,
      "loss": 5.22,
      "mean_token_accuracy": 0.17272175550460817,
      "num_tokens": 37503101.0,
      "step": 17125
    },
    {
      "entropy": 5.654058027267456,
      "epoch": 1.7000793965859469,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004715892720710414,
      "loss": 5.1822,
      "mean_token_accuracy": 0.17079950869083405,
      "num_tokens": 37514265.0,
      "step": 17130
    },
    {
      "entropy": 5.565940189361572,
      "epoch": 1.7005756252481143,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004715720353621767,
      "loss": 5.1175,
      "mean_token_accuracy": 0.16673235446214676,
      "num_tokens": 37524564.0,
      "step": 17135
    },
    {
      "entropy": 5.571040487289428,
      "epoch": 1.7010718539102818,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047155479377874917,
      "loss": 5.1806,
      "mean_token_accuracy": 0.17092925012111665,
      "num_tokens": 37536876.0,
      "step": 17140
    },
    {
      "entropy": 5.586730003356934,
      "epoch": 1.7015680825724493,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047153754732118616,
      "loss": 5.0863,
      "mean_token_accuracy": 0.17365335077047347,
      "num_tokens": 37547868.0,
      "step": 17145
    },
    {
      "entropy": 5.598105525970459,
      "epoch": 1.7020643112346168,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004715202959899156,
      "loss": 5.1172,
      "mean_token_accuracy": 0.17416927367448806,
      "num_tokens": 37558487.0,
      "step": 17150
    },
    {
      "entropy": 5.628242301940918,
      "epoch": 1.7025605398967845,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004715030397853651,
      "loss": 5.0713,
      "mean_token_accuracy": 0.1771671876311302,
      "num_tokens": 37568226.0,
      "step": 17155
    },
    {
      "entropy": 5.635484504699707,
      "epoch": 1.703056768558952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047148577870796264,
      "loss": 5.2009,
      "mean_token_accuracy": 0.16865737736225128,
      "num_tokens": 37579571.0,
      "step": 17160
    },
    {
      "entropy": 5.612712669372558,
      "epoch": 1.7035529972211196,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004714685127581363,
      "loss": 5.1373,
      "mean_token_accuracy": 0.16941481083631516,
      "num_tokens": 37590532.0,
      "step": 17165
    },
    {
      "entropy": 5.636055612564087,
      "epoch": 1.704049225883287,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047145124193631415,
      "loss": 5.1633,
      "mean_token_accuracy": 0.17199888527393342,
      "num_tokens": 37601946.0,
      "step": 17170
    },
    {
      "entropy": 5.495044708251953,
      "epoch": 1.7045454545454546,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047143396624292454,
      "loss": 4.9888,
      "mean_token_accuracy": 0.18531711548566818,
      "num_tokens": 37613944.0,
      "step": 17175
    },
    {
      "entropy": 5.519129896163941,
      "epoch": 1.705041683207622,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047141668567839584,
      "loss": 5.0008,
      "mean_token_accuracy": 0.1826828345656395,
      "num_tokens": 37625308.0,
      "step": 17180
    },
    {
      "entropy": 5.531841135025024,
      "epoch": 1.7055379118697895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004713994002431566,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17116490751504898,
      "num_tokens": 37635570.0,
      "step": 17185
    },
    {
      "entropy": 5.609490919113159,
      "epoch": 1.706034140531957,
      "grad_norm": 1.0,
      "learning_rate": 0.00047138210993763534,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17178215831518173,
      "num_tokens": 37646279.0,
      "step": 17190
    },
    {
      "entropy": 5.606864929199219,
      "epoch": 1.7065303691941247,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000471364814762261,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17847078889608384,
      "num_tokens": 37657182.0,
      "step": 17195
    },
    {
      "entropy": 5.552755832672119,
      "epoch": 1.7070265978562922,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004713475147174624,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17388345152139664,
      "num_tokens": 37668996.0,
      "step": 17200
    },
    {
      "entropy": 5.533081674575806,
      "epoch": 1.7075228265184597,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004713302098036685,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17864137440919875,
      "num_tokens": 37679501.0,
      "step": 17205
    },
    {
      "entropy": 5.564648580551148,
      "epoch": 1.7080190551806274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004713129000213085,
      "loss": 5.1756,
      "mean_token_accuracy": 0.1696133181452751,
      "num_tokens": 37691173.0,
      "step": 17210
    },
    {
      "entropy": 5.617045307159424,
      "epoch": 1.7085152838427948,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004712955853708115,
      "loss": 5.1342,
      "mean_token_accuracy": 0.17831853181123733,
      "num_tokens": 37701875.0,
      "step": 17215
    },
    {
      "entropy": 5.6091728687286375,
      "epoch": 1.7090115125049623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000471278265852607,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17487506866455077,
      "num_tokens": 37711003.0,
      "step": 17220
    },
    {
      "entropy": 5.635334873199463,
      "epoch": 1.7095077411671298,
      "grad_norm": 1.0,
      "learning_rate": 0.0004712609414671245,
      "loss": 5.1603,
      "mean_token_accuracy": 0.17145764976739883,
      "num_tokens": 37721335.0,
      "step": 17225
    },
    {
      "entropy": 5.537389612197876,
      "epoch": 1.7100039698292973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004712436122147935,
      "loss": 5.046,
      "mean_token_accuracy": 0.18321012258529662,
      "num_tokens": 37731664.0,
      "step": 17230
    },
    {
      "entropy": 5.6496116638183596,
      "epoch": 1.7105001984914647,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004712262780960438,
      "loss": 5.2526,
      "mean_token_accuracy": 0.17173455357551576,
      "num_tokens": 37741619.0,
      "step": 17235
    },
    {
      "entropy": 5.612613487243652,
      "epoch": 1.7109964271536324,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047120893911130525,
      "loss": 5.1316,
      "mean_token_accuracy": 0.17573576420545578,
      "num_tokens": 37752387.0,
      "step": 17240
    },
    {
      "entropy": 5.540917444229126,
      "epoch": 1.7114926558158,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047119159526100773,
      "loss": 5.0271,
      "mean_token_accuracy": 0.17445333451032638,
      "num_tokens": 37763459.0,
      "step": 17245
    },
    {
      "entropy": 5.579714393615722,
      "epoch": 1.7119888844779676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047117424654558146,
      "loss": 5.0745,
      "mean_token_accuracy": 0.17875669598579408,
      "num_tokens": 37773848.0,
      "step": 17250
    },
    {
      "entropy": 5.5767515182495115,
      "epoch": 1.712485113140135,
      "grad_norm": 1.0,
      "learning_rate": 0.0004711568929654566,
      "loss": 5.1499,
      "mean_token_accuracy": 0.17460047602653503,
      "num_tokens": 37784074.0,
      "step": 17255
    },
    {
      "entropy": 5.580115222930909,
      "epoch": 1.7129813418023025,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004711395345210635,
      "loss": 5.1533,
      "mean_token_accuracy": 0.17468897104263306,
      "num_tokens": 37794926.0,
      "step": 17260
    },
    {
      "entropy": 5.5565746307373045,
      "epoch": 1.71347757046447,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004711221712128325,
      "loss": 5.0914,
      "mean_token_accuracy": 0.17849989533424376,
      "num_tokens": 37804926.0,
      "step": 17265
    },
    {
      "entropy": 5.613753747940064,
      "epoch": 1.7139737991266375,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004711048030411943,
      "loss": 5.2022,
      "mean_token_accuracy": 0.17084813714027405,
      "num_tokens": 37815241.0,
      "step": 17270
    },
    {
      "entropy": 5.592760038375855,
      "epoch": 1.714470027788805,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047108743000657963,
      "loss": 5.091,
      "mean_token_accuracy": 0.1727021485567093,
      "num_tokens": 37825047.0,
      "step": 17275
    },
    {
      "entropy": 5.607930612564087,
      "epoch": 1.7149662564509724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004710700521094192,
      "loss": 5.1387,
      "mean_token_accuracy": 0.17856706827878951,
      "num_tokens": 37836201.0,
      "step": 17280
    },
    {
      "entropy": 5.561310052871704,
      "epoch": 1.7154624851131401,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047105266935014385,
      "loss": 5.0971,
      "mean_token_accuracy": 0.1717386245727539,
      "num_tokens": 37848242.0,
      "step": 17285
    },
    {
      "entropy": 5.523507404327392,
      "epoch": 1.7159587137753076,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047103528172918487,
      "loss": 5.0828,
      "mean_token_accuracy": 0.18439775109291076,
      "num_tokens": 37859956.0,
      "step": 17290
    },
    {
      "entropy": 5.578397130966186,
      "epoch": 1.7164549424374753,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004710178892469733,
      "loss": 5.1983,
      "mean_token_accuracy": 0.17168510407209397,
      "num_tokens": 37870054.0,
      "step": 17295
    },
    {
      "entropy": 5.5676618099212645,
      "epoch": 1.7169511710996428,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047100049190394044,
      "loss": 5.0722,
      "mean_token_accuracy": 0.1721964731812477,
      "num_tokens": 37879871.0,
      "step": 17300
    },
    {
      "entropy": 5.60065598487854,
      "epoch": 1.7174473997618103,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004709830897005178,
      "loss": 5.185,
      "mean_token_accuracy": 0.17179017513990402,
      "num_tokens": 37892642.0,
      "step": 17305
    },
    {
      "entropy": 5.576100301742554,
      "epoch": 1.7179436284239777,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004709656826371367,
      "loss": 5.1103,
      "mean_token_accuracy": 0.17501808553934098,
      "num_tokens": 37904371.0,
      "step": 17310
    },
    {
      "entropy": 5.662422800064087,
      "epoch": 1.7184398570861452,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004709482707142291,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1699893832206726,
      "num_tokens": 37915538.0,
      "step": 17315
    },
    {
      "entropy": 5.544990015029907,
      "epoch": 1.7189360857483127,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047093085393222646,
      "loss": 5.0478,
      "mean_token_accuracy": 0.1764781355857849,
      "num_tokens": 37926254.0,
      "step": 17320
    },
    {
      "entropy": 5.550130367279053,
      "epoch": 1.7194323144104804,
      "grad_norm": 0.875,
      "learning_rate": 0.00047091343229156094,
      "loss": 5.1315,
      "mean_token_accuracy": 0.17655944526195527,
      "num_tokens": 37939242.0,
      "step": 17325
    },
    {
      "entropy": 5.546997690200806,
      "epoch": 1.7199285430726479,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004708960057926644,
      "loss": 5.0695,
      "mean_token_accuracy": 0.17920308858156203,
      "num_tokens": 37948970.0,
      "step": 17330
    },
    {
      "entropy": 5.638411951065064,
      "epoch": 1.7204247717348156,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047087857443596905,
      "loss": 5.289,
      "mean_token_accuracy": 0.1584939703345299,
      "num_tokens": 37960290.0,
      "step": 17335
    },
    {
      "entropy": 5.654365062713623,
      "epoch": 1.720921000396983,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047086113822190716,
      "loss": 5.137,
      "mean_token_accuracy": 0.1748058944940567,
      "num_tokens": 37971004.0,
      "step": 17340
    },
    {
      "entropy": 5.670430135726929,
      "epoch": 1.7214172290591505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047084369715091103,
      "loss": 5.1543,
      "mean_token_accuracy": 0.17077600210905075,
      "num_tokens": 37982455.0,
      "step": 17345
    },
    {
      "entropy": 5.588890123367309,
      "epoch": 1.721913457721318,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004708262512234133,
      "loss": 5.075,
      "mean_token_accuracy": 0.17568221241235732,
      "num_tokens": 37993623.0,
      "step": 17350
    },
    {
      "entropy": 5.538778352737427,
      "epoch": 1.7224096863834855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004708088004398464,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17752674520015715,
      "num_tokens": 38002636.0,
      "step": 17355
    },
    {
      "entropy": 5.573910331726074,
      "epoch": 1.722905915045653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004707913448006433,
      "loss": 5.1571,
      "mean_token_accuracy": 0.17350490987300873,
      "num_tokens": 38014469.0,
      "step": 17360
    },
    {
      "entropy": 5.58060393333435,
      "epoch": 1.7234021437078204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004707738843062367,
      "loss": 5.0374,
      "mean_token_accuracy": 0.1792618826031685,
      "num_tokens": 38024558.0,
      "step": 17365
    },
    {
      "entropy": 5.538830709457398,
      "epoch": 1.723898372369988,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004707564189570596,
      "loss": 5.1013,
      "mean_token_accuracy": 0.17598968148231506,
      "num_tokens": 38034872.0,
      "step": 17370
    },
    {
      "entropy": 5.5948343753814695,
      "epoch": 1.7243946010321556,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004707389487535451,
      "loss": 5.099,
      "mean_token_accuracy": 0.1771678149700165,
      "num_tokens": 38044947.0,
      "step": 17375
    },
    {
      "entropy": 5.666007661819458,
      "epoch": 1.7248908296943233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004707214736961265,
      "loss": 5.2588,
      "mean_token_accuracy": 0.1687694564461708,
      "num_tokens": 38055349.0,
      "step": 17380
    },
    {
      "entropy": 5.643075799942016,
      "epoch": 1.7253870583564908,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004707039937852371,
      "loss": 5.1654,
      "mean_token_accuracy": 0.17223668992519378,
      "num_tokens": 38066337.0,
      "step": 17385
    },
    {
      "entropy": 5.5671392440795895,
      "epoch": 1.7258832870186582,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047068650902131034,
      "loss": 5.1845,
      "mean_token_accuracy": 0.17117101699113846,
      "num_tokens": 38078402.0,
      "step": 17390
    },
    {
      "entropy": 5.559861421585083,
      "epoch": 1.7263795156808257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004706690194047798,
      "loss": 5.0213,
      "mean_token_accuracy": 0.18059539794921875,
      "num_tokens": 38089781.0,
      "step": 17395
    },
    {
      "entropy": 5.584499740600586,
      "epoch": 1.7268757443429932,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004706515249360792,
      "loss": 5.0113,
      "mean_token_accuracy": 0.1747628256678581,
      "num_tokens": 38100496.0,
      "step": 17400
    },
    {
      "entropy": 5.594361639022827,
      "epoch": 1.7273719730051607,
      "grad_norm": 1.0,
      "learning_rate": 0.0004706340256156425,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1781156301498413,
      "num_tokens": 38110471.0,
      "step": 17405
    },
    {
      "entropy": 5.614654922485352,
      "epoch": 1.7278682016673284,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004706165214439034,
      "loss": 5.0987,
      "mean_token_accuracy": 0.17517899572849274,
      "num_tokens": 38120511.0,
      "step": 17410
    },
    {
      "entropy": 5.604961967468261,
      "epoch": 1.7283644303294958,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004705990124212961,
      "loss": 5.0833,
      "mean_token_accuracy": 0.17269066721200943,
      "num_tokens": 38130501.0,
      "step": 17415
    },
    {
      "entropy": 5.547133350372315,
      "epoch": 1.7288606589916635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004705814985482548,
      "loss": 5.097,
      "mean_token_accuracy": 0.1728976845741272,
      "num_tokens": 38139634.0,
      "step": 17420
    },
    {
      "entropy": 5.611413526535034,
      "epoch": 1.729356887653831,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047056397982521366,
      "loss": 5.1259,
      "mean_token_accuracy": 0.16574106216430665,
      "num_tokens": 38151216.0,
      "step": 17425
    },
    {
      "entropy": 5.562301349639893,
      "epoch": 1.7298531163159985,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004705464562526074,
      "loss": 5.0167,
      "mean_token_accuracy": 0.17705488204956055,
      "num_tokens": 38162014.0,
      "step": 17430
    },
    {
      "entropy": 5.598207426071167,
      "epoch": 1.730349344978166,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004705289278308703,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17308807969093323,
      "num_tokens": 38174928.0,
      "step": 17435
    },
    {
      "entropy": 5.634600114822388,
      "epoch": 1.7308455736403334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047051139456043707,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1643284797668457,
      "num_tokens": 38185818.0,
      "step": 17440
    },
    {
      "entropy": 5.60419807434082,
      "epoch": 1.731341802302501,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004704938564417427,
      "loss": 5.1514,
      "mean_token_accuracy": 0.17000762820243837,
      "num_tokens": 38197595.0,
      "step": 17445
    },
    {
      "entropy": 5.616260576248169,
      "epoch": 1.7318380309646684,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047047631347522175,
      "loss": 5.1455,
      "mean_token_accuracy": 0.16809532195329666,
      "num_tokens": 38209874.0,
      "step": 17450
    },
    {
      "entropy": 5.6067877292633055,
      "epoch": 1.732334259626836,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047045876566130957,
      "loss": 5.129,
      "mean_token_accuracy": 0.1662623755633831,
      "num_tokens": 38221054.0,
      "step": 17455
    },
    {
      "entropy": 5.642846393585205,
      "epoch": 1.7328304882890035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004704412130004411,
      "loss": 5.136,
      "mean_token_accuracy": 0.171579971909523,
      "num_tokens": 38230766.0,
      "step": 17460
    },
    {
      "entropy": 5.563117408752442,
      "epoch": 1.7333267169511712,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047042365549305174,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1763167828321457,
      "num_tokens": 38242925.0,
      "step": 17465
    },
    {
      "entropy": 5.544460964202881,
      "epoch": 1.7338229456133387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004704060931395768,
      "loss": 5.11,
      "mean_token_accuracy": 0.18025638908147812,
      "num_tokens": 38254228.0,
      "step": 17470
    },
    {
      "entropy": 5.556156158447266,
      "epoch": 1.7343191742755062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047038852594045186,
      "loss": 5.076,
      "mean_token_accuracy": 0.1753190502524376,
      "num_tokens": 38265515.0,
      "step": 17475
    },
    {
      "entropy": 5.589363956451416,
      "epoch": 1.7348154029376737,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047037095389611245,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17652652710676192,
      "num_tokens": 38277232.0,
      "step": 17480
    },
    {
      "entropy": 5.566019439697266,
      "epoch": 1.7353116315998411,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004703533770069944,
      "loss": 5.1058,
      "mean_token_accuracy": 0.17330703884363174,
      "num_tokens": 38288038.0,
      "step": 17485
    },
    {
      "entropy": 5.617719602584839,
      "epoch": 1.7358078602620086,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047033579527353355,
      "loss": 5.1505,
      "mean_token_accuracy": 0.17399047166109086,
      "num_tokens": 38299053.0,
      "step": 17490
    },
    {
      "entropy": 5.663906192779541,
      "epoch": 1.7363040889241763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004703182086961659,
      "loss": 5.1907,
      "mean_token_accuracy": 0.1732718676328659,
      "num_tokens": 38311029.0,
      "step": 17495
    },
    {
      "entropy": 5.7145429134368895,
      "epoch": 1.7368003175863438,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004703006172753274,
      "loss": 5.2752,
      "mean_token_accuracy": 0.17340957671403884,
      "num_tokens": 38321525.0,
      "step": 17500
    },
    {
      "entropy": 5.573951959609985,
      "epoch": 1.7372965462485113,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047028302101145467,
      "loss": 5.0553,
      "mean_token_accuracy": 0.17492837309837342,
      "num_tokens": 38331035.0,
      "step": 17505
    },
    {
      "entropy": 5.595009422302246,
      "epoch": 1.737792774910679,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004702654199049836,
      "loss": 5.1675,
      "mean_token_accuracy": 0.17433960735797882,
      "num_tokens": 38342046.0,
      "step": 17510
    },
    {
      "entropy": 5.582774353027344,
      "epoch": 1.7382890035728464,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000470247813956351,
      "loss": 5.0774,
      "mean_token_accuracy": 0.17988122552633284,
      "num_tokens": 38352614.0,
      "step": 17515
    },
    {
      "entropy": 5.617387580871582,
      "epoch": 1.738785232235014,
      "grad_norm": 1.125,
      "learning_rate": 0.00047023020316599334,
      "loss": 5.1684,
      "mean_token_accuracy": 0.17212685346603393,
      "num_tokens": 38364379.0,
      "step": 17520
    },
    {
      "entropy": 5.60879282951355,
      "epoch": 1.7392814608971814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047021258753434725,
      "loss": 5.201,
      "mean_token_accuracy": 0.16685703545808792,
      "num_tokens": 38373823.0,
      "step": 17525
    },
    {
      "entropy": 5.6208442687988285,
      "epoch": 1.7397776895593489,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047019496706184966,
      "loss": 5.1166,
      "mean_token_accuracy": 0.17232927083969116,
      "num_tokens": 38384238.0,
      "step": 17530
    },
    {
      "entropy": 5.591731929779053,
      "epoch": 1.7402739182215163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004701773417489375,
      "loss": 5.0823,
      "mean_token_accuracy": 0.17775111347436906,
      "num_tokens": 38394980.0,
      "step": 17535
    },
    {
      "entropy": 5.509933757781982,
      "epoch": 1.740770146883684,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047015971159604787,
      "loss": 5.1692,
      "mean_token_accuracy": 0.1678324744105339,
      "num_tokens": 38406722.0,
      "step": 17540
    },
    {
      "entropy": 5.567563724517822,
      "epoch": 1.7412663755458515,
      "grad_norm": 1.09375,
      "learning_rate": 0.000470142076603618,
      "loss": 5.0626,
      "mean_token_accuracy": 0.17587853521108626,
      "num_tokens": 38416204.0,
      "step": 17545
    },
    {
      "entropy": 5.574887895584107,
      "epoch": 1.7417626042080192,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047012443677208496,
      "loss": 5.0995,
      "mean_token_accuracy": 0.18302798867225648,
      "num_tokens": 38426166.0,
      "step": 17550
    },
    {
      "entropy": 5.582759666442871,
      "epoch": 1.7422588328701867,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004701067921018864,
      "loss": 5.1192,
      "mean_token_accuracy": 0.17819337248802186,
      "num_tokens": 38439195.0,
      "step": 17555
    },
    {
      "entropy": 5.578144884109497,
      "epoch": 1.7427550615323542,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047008914259345986,
      "loss": 5.1039,
      "mean_token_accuracy": 0.1754085436463356,
      "num_tokens": 38449276.0,
      "step": 17560
    },
    {
      "entropy": 5.645083093643189,
      "epoch": 1.7432512901945216,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047007148824724295,
      "loss": 5.1701,
      "mean_token_accuracy": 0.16923018395900727,
      "num_tokens": 38460961.0,
      "step": 17565
    },
    {
      "entropy": 5.612904500961304,
      "epoch": 1.743747518856689,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004700538290636734,
      "loss": 5.0481,
      "mean_token_accuracy": 0.17873757928609849,
      "num_tokens": 38470961.0,
      "step": 17570
    },
    {
      "entropy": 5.593082761764526,
      "epoch": 1.7442437475188566,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047003616504318925,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17648623287677764,
      "num_tokens": 38481233.0,
      "step": 17575
    },
    {
      "entropy": 5.566795015335083,
      "epoch": 1.744739976181024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047001849618622844,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1733985349535942,
      "num_tokens": 38492260.0,
      "step": 17580
    },
    {
      "entropy": 5.5929083824157715,
      "epoch": 1.7452362048431918,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047000082249322916,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17804200053215027,
      "num_tokens": 38503492.0,
      "step": 17585
    },
    {
      "entropy": 5.559781837463379,
      "epoch": 1.7457324335053592,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004699831439646297,
      "loss": 5.0816,
      "mean_token_accuracy": 0.1797600731253624,
      "num_tokens": 38514680.0,
      "step": 17590
    },
    {
      "entropy": 5.5316162109375,
      "epoch": 1.746228662167527,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004699654606008684,
      "loss": 5.016,
      "mean_token_accuracy": 0.17625977247953414,
      "num_tokens": 38525948.0,
      "step": 17595
    },
    {
      "entropy": 5.669258308410645,
      "epoch": 1.7467248908296944,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004699477724023838,
      "loss": 5.1508,
      "mean_token_accuracy": 0.17345528900623322,
      "num_tokens": 38535955.0,
      "step": 17600
    },
    {
      "entropy": 5.597427892684936,
      "epoch": 1.7472211194918619,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004699300793696145,
      "loss": 5.1346,
      "mean_token_accuracy": 0.17525719553232194,
      "num_tokens": 38548420.0,
      "step": 17605
    },
    {
      "entropy": 5.701650905609131,
      "epoch": 1.7477173481540293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004699123815029993,
      "loss": 5.2714,
      "mean_token_accuracy": 0.16538833528757096,
      "num_tokens": 38559220.0,
      "step": 17610
    },
    {
      "entropy": 5.629284286499024,
      "epoch": 1.7482135768161968,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046989467880297704,
      "loss": 5.1793,
      "mean_token_accuracy": 0.1668732598423958,
      "num_tokens": 38570045.0,
      "step": 17615
    },
    {
      "entropy": 5.643173122406006,
      "epoch": 1.7487098054783643,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046987697126998673,
      "loss": 5.1149,
      "mean_token_accuracy": 0.1639850027859211,
      "num_tokens": 38582191.0,
      "step": 17620
    },
    {
      "entropy": 5.625611925125122,
      "epoch": 1.749206034140532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004698592589044674,
      "loss": 5.0772,
      "mean_token_accuracy": 0.17316126972436904,
      "num_tokens": 38592545.0,
      "step": 17625
    },
    {
      "entropy": 5.5574225902557375,
      "epoch": 1.7497022628026995,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046984154170685833,
      "loss": 5.0873,
      "mean_token_accuracy": 0.1733898177742958,
      "num_tokens": 38604024.0,
      "step": 17630
    },
    {
      "entropy": 5.5736610889434814,
      "epoch": 1.7501984914648672,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004698238196775989,
      "loss": 5.0864,
      "mean_token_accuracy": 0.17741398960351945,
      "num_tokens": 38614961.0,
      "step": 17635
    },
    {
      "entropy": 5.585589790344239,
      "epoch": 1.7506947201270346,
      "grad_norm": 1.0,
      "learning_rate": 0.00046980609281712857,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17657016217708588,
      "num_tokens": 38624706.0,
      "step": 17640
    },
    {
      "entropy": 5.64966082572937,
      "epoch": 1.7511909487892021,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004697883611258868,
      "loss": 5.0845,
      "mean_token_accuracy": 0.18258240967988967,
      "num_tokens": 38636006.0,
      "step": 17645
    },
    {
      "entropy": 5.632290887832641,
      "epoch": 1.7516871774513696,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004697706246043136,
      "loss": 5.1586,
      "mean_token_accuracy": 0.1777812123298645,
      "num_tokens": 38647826.0,
      "step": 17650
    },
    {
      "entropy": 5.567259216308594,
      "epoch": 1.752183406113537,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046975288325284845,
      "loss": 5.1072,
      "mean_token_accuracy": 0.16548297107219695,
      "num_tokens": 38659463.0,
      "step": 17655
    },
    {
      "entropy": 5.599087524414062,
      "epoch": 1.7526796347757045,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004697351370719315,
      "loss": 5.1877,
      "mean_token_accuracy": 0.170122866332531,
      "num_tokens": 38670593.0,
      "step": 17660
    },
    {
      "entropy": 5.563382339477539,
      "epoch": 1.753175863437872,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004697173860620027,
      "loss": 5.0669,
      "mean_token_accuracy": 0.18225714713335037,
      "num_tokens": 38681560.0,
      "step": 17665
    },
    {
      "entropy": 5.5612139225006105,
      "epoch": 1.7536720921000397,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004696996302235024,
      "loss": 5.009,
      "mean_token_accuracy": 0.17627447545528413,
      "num_tokens": 38693624.0,
      "step": 17670
    },
    {
      "entropy": 5.595105266571045,
      "epoch": 1.7541683207622072,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046968186955687074,
      "loss": 5.1093,
      "mean_token_accuracy": 0.17213891744613646,
      "num_tokens": 38704347.0,
      "step": 17675
    },
    {
      "entropy": 5.618455505371093,
      "epoch": 1.7546645494243749,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004696641040625483,
      "loss": 5.1184,
      "mean_token_accuracy": 0.1730091765522957,
      "num_tokens": 38715578.0,
      "step": 17680
    },
    {
      "entropy": 5.551497268676758,
      "epoch": 1.7551607780865424,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046964633374097544,
      "loss": 5.1349,
      "mean_token_accuracy": 0.17087809592485428,
      "num_tokens": 38729539.0,
      "step": 17685
    },
    {
      "entropy": 5.582420873641968,
      "epoch": 1.7556570067487098,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046962855859259306,
      "loss": 5.1536,
      "mean_token_accuracy": 0.17616589218378068,
      "num_tokens": 38742031.0,
      "step": 17690
    },
    {
      "entropy": 5.58052806854248,
      "epoch": 1.7561532354108773,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046961077861784167,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1801002323627472,
      "num_tokens": 38754046.0,
      "step": 17695
    },
    {
      "entropy": 5.529612874984741,
      "epoch": 1.7566494640730448,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004695929938171624,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17679598033428193,
      "num_tokens": 38764678.0,
      "step": 17700
    },
    {
      "entropy": 5.625293207168579,
      "epoch": 1.7571456927352123,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004695752041909962,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17781684696674346,
      "num_tokens": 38775954.0,
      "step": 17705
    },
    {
      "entropy": 5.570199489593506,
      "epoch": 1.75764192139738,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004695574097397842,
      "loss": 5.0609,
      "mean_token_accuracy": 0.17742561548948288,
      "num_tokens": 38785856.0,
      "step": 17710
    },
    {
      "entropy": 5.52103705406189,
      "epoch": 1.7581381500595474,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046953961046396767,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17601171582937242,
      "num_tokens": 38796811.0,
      "step": 17715
    },
    {
      "entropy": 5.58359808921814,
      "epoch": 1.7586343787217151,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046952180636398793,
      "loss": 5.1835,
      "mean_token_accuracy": 0.16861285716295243,
      "num_tokens": 38808536.0,
      "step": 17720
    },
    {
      "entropy": 5.586890077590942,
      "epoch": 1.7591306073838826,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004695039974402866,
      "loss": 5.0708,
      "mean_token_accuracy": 0.17509666532278062,
      "num_tokens": 38819828.0,
      "step": 17725
    },
    {
      "entropy": 5.599941301345825,
      "epoch": 1.75962683604605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046948618369330533,
      "loss": 5.1236,
      "mean_token_accuracy": 0.17667941153049468,
      "num_tokens": 38830975.0,
      "step": 17730
    },
    {
      "entropy": 5.549032926559448,
      "epoch": 1.7601230647082176,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046946836512348566,
      "loss": 5.0533,
      "mean_token_accuracy": 0.17648427337408065,
      "num_tokens": 38841201.0,
      "step": 17735
    },
    {
      "entropy": 5.5622413635253904,
      "epoch": 1.760619293370385,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046945054173126965,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17469056248664855,
      "num_tokens": 38851570.0,
      "step": 17740
    },
    {
      "entropy": 5.627494764328003,
      "epoch": 1.7611155220325525,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046943271351709923,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1741650730371475,
      "num_tokens": 38862869.0,
      "step": 17745
    },
    {
      "entropy": 5.575467252731324,
      "epoch": 1.76161175069472,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004694148804814164,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17400117814540864,
      "num_tokens": 38874162.0,
      "step": 17750
    },
    {
      "entropy": 5.495044898986817,
      "epoch": 1.7621079793568877,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004693970426246635,
      "loss": 4.9901,
      "mean_token_accuracy": 0.17943394929170609,
      "num_tokens": 38884480.0,
      "step": 17755
    },
    {
      "entropy": 5.612374973297119,
      "epoch": 1.7626042080190552,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046937919994728287,
      "loss": 5.1731,
      "mean_token_accuracy": 0.16984302997589112,
      "num_tokens": 38895623.0,
      "step": 17760
    },
    {
      "entropy": 5.559154033660889,
      "epoch": 1.7631004366812228,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004693613524497168,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1789169505238533,
      "num_tokens": 38907286.0,
      "step": 17765
    },
    {
      "entropy": 5.514584159851074,
      "epoch": 1.7635966653433903,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046934350013240814,
      "loss": 5.0214,
      "mean_token_accuracy": 0.17955413609743118,
      "num_tokens": 38919342.0,
      "step": 17770
    },
    {
      "entropy": 5.533892488479614,
      "epoch": 1.7640928940055578,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004693256429957994,
      "loss": 5.0716,
      "mean_token_accuracy": 0.17566768527030946,
      "num_tokens": 38930289.0,
      "step": 17775
    },
    {
      "entropy": 5.534735536575317,
      "epoch": 1.7645891226677253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004693077810403335,
      "loss": 5.045,
      "mean_token_accuracy": 0.17625189423561097,
      "num_tokens": 38941053.0,
      "step": 17780
    },
    {
      "entropy": 5.581464672088623,
      "epoch": 1.7650853513298927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046928991426645325,
      "loss": 5.069,
      "mean_token_accuracy": 0.17737413793802262,
      "num_tokens": 38952474.0,
      "step": 17785
    },
    {
      "entropy": 5.533790969848633,
      "epoch": 1.7655815799920602,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046927204267460177,
      "loss": 5.0486,
      "mean_token_accuracy": 0.18329508900642394,
      "num_tokens": 38963049.0,
      "step": 17790
    },
    {
      "entropy": 5.716771411895752,
      "epoch": 1.766077808654228,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046925416626522235,
      "loss": 5.2635,
      "mean_token_accuracy": 0.1665349066257477,
      "num_tokens": 38975290.0,
      "step": 17795
    },
    {
      "entropy": 5.616739225387573,
      "epoch": 1.7665740373163954,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004692362850387581,
      "loss": 5.1022,
      "mean_token_accuracy": 0.17290093898773193,
      "num_tokens": 38985825.0,
      "step": 17800
    },
    {
      "entropy": 5.6407557964324955,
      "epoch": 1.7670702659785629,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046921839899565263,
      "loss": 5.1699,
      "mean_token_accuracy": 0.168709696829319,
      "num_tokens": 38998119.0,
      "step": 17805
    },
    {
      "entropy": 5.61807951927185,
      "epoch": 1.7675664946407306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004692005081363493,
      "loss": 5.144,
      "mean_token_accuracy": 0.1771093487739563,
      "num_tokens": 39009204.0,
      "step": 17810
    },
    {
      "entropy": 5.514318513870239,
      "epoch": 1.768062723302898,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046918261246129176,
      "loss": 5.0404,
      "mean_token_accuracy": 0.17685907632112502,
      "num_tokens": 39020138.0,
      "step": 17815
    },
    {
      "entropy": 5.575516843795777,
      "epoch": 1.7685589519650655,
      "grad_norm": 0.859375,
      "learning_rate": 0.000469164711970924,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1740242972970009,
      "num_tokens": 39031590.0,
      "step": 17820
    },
    {
      "entropy": 5.608274602890015,
      "epoch": 1.769055180627233,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004691468066656896,
      "loss": 5.1854,
      "mean_token_accuracy": 0.1712047502398491,
      "num_tokens": 39042361.0,
      "step": 17825
    },
    {
      "entropy": 5.633759069442749,
      "epoch": 1.7695514092894005,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046912889654603286,
      "loss": 5.0671,
      "mean_token_accuracy": 0.17738888263702393,
      "num_tokens": 39053560.0,
      "step": 17830
    },
    {
      "entropy": 5.545594072341919,
      "epoch": 1.770047637951568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004691109816123977,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1821296364068985,
      "num_tokens": 39064172.0,
      "step": 17835
    },
    {
      "entropy": 5.635965633392334,
      "epoch": 1.7705438666137356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690930618652286,
      "loss": 5.1253,
      "mean_token_accuracy": 0.17344083935022353,
      "num_tokens": 39075860.0,
      "step": 17840
    },
    {
      "entropy": 5.5872447967529295,
      "epoch": 1.7710400952759031,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046907513730496975,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17146498560905457,
      "num_tokens": 39087148.0,
      "step": 17845
    },
    {
      "entropy": 5.5724844455719,
      "epoch": 1.7715363239380708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004690572079320657,
      "loss": 5.1397,
      "mean_token_accuracy": 0.1711200773715973,
      "num_tokens": 39097860.0,
      "step": 17850
    },
    {
      "entropy": 5.558564233779907,
      "epoch": 1.7720325526002383,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046903927374696106,
      "loss": 5.0186,
      "mean_token_accuracy": 0.17607691287994384,
      "num_tokens": 39108475.0,
      "step": 17855
    },
    {
      "entropy": 5.491811800003052,
      "epoch": 1.7725287812624058,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004690213347501005,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17628397643566132,
      "num_tokens": 39119167.0,
      "step": 17860
    },
    {
      "entropy": 5.643675994873047,
      "epoch": 1.7730250099245732,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004690033909419289,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17190787345170974,
      "num_tokens": 39130387.0,
      "step": 17865
    },
    {
      "entropy": 5.556180000305176,
      "epoch": 1.7735212385867407,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046898544232289126,
      "loss": 4.978,
      "mean_token_accuracy": 0.18084857165813445,
      "num_tokens": 39140673.0,
      "step": 17870
    },
    {
      "entropy": 5.573554134368896,
      "epoch": 1.7740174672489082,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046896748889343267,
      "loss": 5.058,
      "mean_token_accuracy": 0.17551981657743454,
      "num_tokens": 39151808.0,
      "step": 17875
    },
    {
      "entropy": 5.577996683120728,
      "epoch": 1.7745136959110757,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046894953065399836,
      "loss": 5.123,
      "mean_token_accuracy": 0.17654286921024323,
      "num_tokens": 39163432.0,
      "step": 17880
    },
    {
      "entropy": 5.552130794525146,
      "epoch": 1.7750099245732434,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046893156760503357,
      "loss": 5.1131,
      "mean_token_accuracy": 0.17143853902816772,
      "num_tokens": 39173802.0,
      "step": 17885
    },
    {
      "entropy": 5.578811264038086,
      "epoch": 1.7755061532354108,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004689135997469838,
      "loss": 5.13,
      "mean_token_accuracy": 0.1768750935792923,
      "num_tokens": 39184475.0,
      "step": 17890
    },
    {
      "entropy": 5.621394300460816,
      "epoch": 1.7760023818975785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004688956270802945,
      "loss": 5.0864,
      "mean_token_accuracy": 0.17681587040424346,
      "num_tokens": 39195607.0,
      "step": 17895
    },
    {
      "entropy": 5.631475782394409,
      "epoch": 1.776498610559746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004688776496054117,
      "loss": 5.1436,
      "mean_token_accuracy": 0.16998857408761978,
      "num_tokens": 39205721.0,
      "step": 17900
    },
    {
      "entropy": 5.646664714813232,
      "epoch": 1.7769948392219135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046885966732278077,
      "loss": 5.115,
      "mean_token_accuracy": 0.17327467501163482,
      "num_tokens": 39216352.0,
      "step": 17905
    },
    {
      "entropy": 5.535884714126587,
      "epoch": 1.777491067884081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046884168023284787,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18244670629501342,
      "num_tokens": 39225872.0,
      "step": 17910
    },
    {
      "entropy": 5.471955871582031,
      "epoch": 1.7779872965462484,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046882368833605894,
      "loss": 5.0227,
      "mean_token_accuracy": 0.18395255357027054,
      "num_tokens": 39237751.0,
      "step": 17915
    },
    {
      "entropy": 5.541976547241211,
      "epoch": 1.778483525208416,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046880569163286023,
      "loss": 5.0622,
      "mean_token_accuracy": 0.17768409848213196,
      "num_tokens": 39247854.0,
      "step": 17920
    },
    {
      "entropy": 5.577293539047242,
      "epoch": 1.7789797538705836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046878769012369807,
      "loss": 5.1368,
      "mean_token_accuracy": 0.17796580642461776,
      "num_tokens": 39258220.0,
      "step": 17925
    },
    {
      "entropy": 5.588753175735474,
      "epoch": 1.779475982532751,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046876968380901864,
      "loss": 5.0754,
      "mean_token_accuracy": 0.1745940387248993,
      "num_tokens": 39269303.0,
      "step": 17930
    },
    {
      "entropy": 5.574344491958618,
      "epoch": 1.7799722111949188,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046875167268926865,
      "loss": 5.0857,
      "mean_token_accuracy": 0.16981194466352462,
      "num_tokens": 39280048.0,
      "step": 17935
    },
    {
      "entropy": 5.569331789016724,
      "epoch": 1.7804684398570862,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004687336567648947,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18401660323143004,
      "num_tokens": 39291174.0,
      "step": 17940
    },
    {
      "entropy": 5.562114667892456,
      "epoch": 1.7809646685192537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046871563603634347,
      "loss": 5.0501,
      "mean_token_accuracy": 0.17736072838306427,
      "num_tokens": 39301332.0,
      "step": 17945
    },
    {
      "entropy": 5.581981754302978,
      "epoch": 1.7814608971814212,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004686976105040619,
      "loss": 5.1091,
      "mean_token_accuracy": 0.17728225886821747,
      "num_tokens": 39311935.0,
      "step": 17950
    },
    {
      "entropy": 5.604891014099121,
      "epoch": 1.7819571258435887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004686795801684969,
      "loss": 5.1603,
      "mean_token_accuracy": 0.1636400431394577,
      "num_tokens": 39321973.0,
      "step": 17955
    },
    {
      "entropy": 5.582123756408691,
      "epoch": 1.7824533545057561,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004686615450300958,
      "loss": 5.1274,
      "mean_token_accuracy": 0.1764960542321205,
      "num_tokens": 39332384.0,
      "step": 17960
    },
    {
      "entropy": 5.546534442901612,
      "epoch": 1.7829495831679236,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004686435050893056,
      "loss": 5.0369,
      "mean_token_accuracy": 0.17568558305501938,
      "num_tokens": 39342844.0,
      "step": 17965
    },
    {
      "entropy": 5.577013301849365,
      "epoch": 1.7834458118300913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046862546034657376,
      "loss": 5.1458,
      "mean_token_accuracy": 0.17673320323228836,
      "num_tokens": 39354045.0,
      "step": 17970
    },
    {
      "entropy": 5.5804211616516115,
      "epoch": 1.7839420404922588,
      "grad_norm": 1.0,
      "learning_rate": 0.0004686074108023477,
      "loss": 5.1108,
      "mean_token_accuracy": 0.17045334130525588,
      "num_tokens": 39364529.0,
      "step": 17975
    },
    {
      "entropy": 5.479108428955078,
      "epoch": 1.7844382691544265,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046858935645707506,
      "loss": 5.0357,
      "mean_token_accuracy": 0.1885897621512413,
      "num_tokens": 39375671.0,
      "step": 17980
    },
    {
      "entropy": 5.56713490486145,
      "epoch": 1.784934497816594,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004685712973112035,
      "loss": 5.0867,
      "mean_token_accuracy": 0.1776898443698883,
      "num_tokens": 39387546.0,
      "step": 17985
    },
    {
      "entropy": 5.649010038375854,
      "epoch": 1.7854307264787614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004685532333651809,
      "loss": 5.061,
      "mean_token_accuracy": 0.18097138702869414,
      "num_tokens": 39397726.0,
      "step": 17990
    },
    {
      "entropy": 5.596323394775391,
      "epoch": 1.785926955140929,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046853516461945523,
      "loss": 5.0938,
      "mean_token_accuracy": 0.170897077023983,
      "num_tokens": 39409869.0,
      "step": 17995
    },
    {
      "entropy": 5.497243213653564,
      "epoch": 1.7864231838030964,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046851709107447446,
      "loss": 5.0689,
      "mean_token_accuracy": 0.18236904144287108,
      "num_tokens": 39419843.0,
      "step": 18000
    },
    {
      "epoch": 1.7864231838030964,
      "eval_entropy": 5.328605877189065,
      "eval_loss": 5.206164360046387,
      "eval_mean_token_accuracy": 0.1757542160453607,
      "eval_num_tokens": 39419843.0,
      "eval_runtime": 20.2282,
      "eval_samples_per_second": 1664.456,
      "eval_steps_per_second": 208.075,
      "step": 18000
    },
    {
      "entropy": 5.552086353302002,
      "epoch": 1.7869194124652639,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004684990127306868,
      "loss": 5.0832,
      "mean_token_accuracy": 0.17747580111026764,
      "num_tokens": 39430307.0,
      "step": 18005
    },
    {
      "entropy": 5.653769683837891,
      "epoch": 1.7874156411274316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046848092958854065,
      "loss": 5.2159,
      "mean_token_accuracy": 0.16519177705049515,
      "num_tokens": 39442134.0,
      "step": 18010
    },
    {
      "entropy": 5.591872930526733,
      "epoch": 1.787911869789599,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004684628416484843,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1830161303281784,
      "num_tokens": 39451945.0,
      "step": 18015
    },
    {
      "entropy": 5.597269582748413,
      "epoch": 1.7884080984517667,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004684447489109663,
      "loss": 5.1075,
      "mean_token_accuracy": 0.1749233990907669,
      "num_tokens": 39463045.0,
      "step": 18020
    },
    {
      "entropy": 5.592468118667602,
      "epoch": 1.7889043271139342,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046842665137643546,
      "loss": 5.0935,
      "mean_token_accuracy": 0.1834035888314247,
      "num_tokens": 39474523.0,
      "step": 18025
    },
    {
      "entropy": 5.580668354034424,
      "epoch": 1.7894005557761017,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046840854904534046,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18044792711734772,
      "num_tokens": 39486005.0,
      "step": 18030
    },
    {
      "entropy": 5.584986925125122,
      "epoch": 1.7898967844382692,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004683904419181302,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1760050669312477,
      "num_tokens": 39497418.0,
      "step": 18035
    },
    {
      "entropy": 5.5154577732086185,
      "epoch": 1.7903930131004366,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046837232999525366,
      "loss": 5.0448,
      "mean_token_accuracy": 0.17674286365509034,
      "num_tokens": 39507965.0,
      "step": 18040
    },
    {
      "entropy": 5.61885437965393,
      "epoch": 1.7908892417626041,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046835421327716,
      "loss": 5.1185,
      "mean_token_accuracy": 0.17623037993907928,
      "num_tokens": 39518537.0,
      "step": 18045
    },
    {
      "entropy": 5.544904947280884,
      "epoch": 1.7913854704247716,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004683360917642986,
      "loss": 5.0836,
      "mean_token_accuracy": 0.17268504202365875,
      "num_tokens": 39530364.0,
      "step": 18050
    },
    {
      "entropy": 5.620218276977539,
      "epoch": 1.7918816990869393,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046831796545711865,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1656372681260109,
      "num_tokens": 39541336.0,
      "step": 18055
    },
    {
      "entropy": 5.590735530853271,
      "epoch": 1.7923779277491068,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046829983435606973,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17740633934736252,
      "num_tokens": 39552556.0,
      "step": 18060
    },
    {
      "entropy": 5.574355554580689,
      "epoch": 1.7928741564112745,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004682816984616014,
      "loss": 5.0526,
      "mean_token_accuracy": 0.1769322782754898,
      "num_tokens": 39565036.0,
      "step": 18065
    },
    {
      "entropy": 5.6203293800354,
      "epoch": 1.793370385073442,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046826355777416346,
      "loss": 5.1211,
      "mean_token_accuracy": 0.1725521892309189,
      "num_tokens": 39575542.0,
      "step": 18070
    },
    {
      "entropy": 5.582083034515381,
      "epoch": 1.7938666137356094,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046824541229420583,
      "loss": 5.1272,
      "mean_token_accuracy": 0.17396182268857957,
      "num_tokens": 39586099.0,
      "step": 18075
    },
    {
      "entropy": 5.5307928085327145,
      "epoch": 1.7943628423977769,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046822726202217825,
      "loss": 5.069,
      "mean_token_accuracy": 0.18288020938634872,
      "num_tokens": 39597142.0,
      "step": 18080
    },
    {
      "entropy": 5.561190843582153,
      "epoch": 1.7948590710599444,
      "grad_norm": 1.03125,
      "learning_rate": 0.000468209106958531,
      "loss": 5.1024,
      "mean_token_accuracy": 0.1702594518661499,
      "num_tokens": 39608906.0,
      "step": 18085
    },
    {
      "entropy": 5.587188386917115,
      "epoch": 1.7953552997221118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004681909471037142,
      "loss": 5.1154,
      "mean_token_accuracy": 0.1767791584134102,
      "num_tokens": 39620269.0,
      "step": 18090
    },
    {
      "entropy": 5.693841218948364,
      "epoch": 1.7958515283842795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046817278245817815,
      "loss": 5.2334,
      "mean_token_accuracy": 0.16894394755363465,
      "num_tokens": 39630742.0,
      "step": 18095
    },
    {
      "entropy": 5.575520658493042,
      "epoch": 1.796347757046447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004681546130223735,
      "loss": 5.1468,
      "mean_token_accuracy": 0.171039155125618,
      "num_tokens": 39640867.0,
      "step": 18100
    },
    {
      "entropy": 5.518161249160767,
      "epoch": 1.7968439857086145,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004681364387967505,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18037778437137603,
      "num_tokens": 39651883.0,
      "step": 18105
    },
    {
      "entropy": 5.553920412063599,
      "epoch": 1.7973402143707822,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004681182597817601,
      "loss": 5.0698,
      "mean_token_accuracy": 0.180068938434124,
      "num_tokens": 39662224.0,
      "step": 18110
    },
    {
      "entropy": 5.609895467758179,
      "epoch": 1.7978364430329496,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004681000759778529,
      "loss": 5.1772,
      "mean_token_accuracy": 0.1681657388806343,
      "num_tokens": 39673847.0,
      "step": 18115
    },
    {
      "entropy": 5.6433045864105225,
      "epoch": 1.7983326716951171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004680818873854799,
      "loss": 5.0448,
      "mean_token_accuracy": 0.1846429705619812,
      "num_tokens": 39684721.0,
      "step": 18120
    },
    {
      "entropy": 5.64415807723999,
      "epoch": 1.7988289003572846,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004680636940050922,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17596376836299896,
      "num_tokens": 39695779.0,
      "step": 18125
    },
    {
      "entropy": 5.550685453414917,
      "epoch": 1.799325129019452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000468045495837141,
      "loss": 5.1045,
      "mean_token_accuracy": 0.17055448442697524,
      "num_tokens": 39706635.0,
      "step": 18130
    },
    {
      "entropy": 5.534542655944824,
      "epoch": 1.7998213576816195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046802729288207735,
      "loss": 5.0481,
      "mean_token_accuracy": 0.18137500137090684,
      "num_tokens": 39716934.0,
      "step": 18135
    },
    {
      "entropy": 5.567750120162964,
      "epoch": 1.8003175863437872,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046800908514035285,
      "loss": 5.0649,
      "mean_token_accuracy": 0.18352790772914887,
      "num_tokens": 39728232.0,
      "step": 18140
    },
    {
      "entropy": 5.583821439743042,
      "epoch": 1.8008138150059547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046799087261241884,
      "loss": 5.0169,
      "mean_token_accuracy": 0.1831623300909996,
      "num_tokens": 39738906.0,
      "step": 18145
    },
    {
      "entropy": 5.576365995407104,
      "epoch": 1.8013100436681224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004679726552987271,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18057131469249726,
      "num_tokens": 39749349.0,
      "step": 18150
    },
    {
      "entropy": 5.55555739402771,
      "epoch": 1.80180627233029,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004679544331997294,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18084515184164046,
      "num_tokens": 39760082.0,
      "step": 18155
    },
    {
      "entropy": 5.590253257751465,
      "epoch": 1.8023025009924574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046793620631587743,
      "loss": 5.098,
      "mean_token_accuracy": 0.17233668118715287,
      "num_tokens": 39771365.0,
      "step": 18160
    },
    {
      "entropy": 5.597818183898926,
      "epoch": 1.8027987296546248,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046791797464762334,
      "loss": 5.1311,
      "mean_token_accuracy": 0.16926528811454772,
      "num_tokens": 39781890.0,
      "step": 18165
    },
    {
      "entropy": 5.6373053073883055,
      "epoch": 1.8032949583167923,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046789973819541915,
      "loss": 5.1448,
      "mean_token_accuracy": 0.17292551100254058,
      "num_tokens": 39791934.0,
      "step": 18170
    },
    {
      "entropy": 5.626489114761353,
      "epoch": 1.8037911869789598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004678814969597171,
      "loss": 5.1196,
      "mean_token_accuracy": 0.16775285452604294,
      "num_tokens": 39802797.0,
      "step": 18175
    },
    {
      "entropy": 5.581368064880371,
      "epoch": 1.8042874156411273,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004678632509409696,
      "loss": 5.1035,
      "mean_token_accuracy": 0.1785384699702263,
      "num_tokens": 39813457.0,
      "step": 18180
    },
    {
      "entropy": 5.612120151519775,
      "epoch": 1.804783644303295,
      "grad_norm": 0.953125,
      "learning_rate": 0.000467845000139629,
      "loss": 5.1783,
      "mean_token_accuracy": 0.17123537808656691,
      "num_tokens": 39825771.0,
      "step": 18185
    },
    {
      "entropy": 5.609736347198487,
      "epoch": 1.8052798729654624,
      "grad_norm": 0.953125,
      "learning_rate": 0.000467826744556148,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1773964986205101,
      "num_tokens": 39836388.0,
      "step": 18190
    },
    {
      "entropy": 5.619792032241821,
      "epoch": 1.8057761016276301,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004678084841909792,
      "loss": 5.1679,
      "mean_token_accuracy": 0.17160738855600358,
      "num_tokens": 39847757.0,
      "step": 18195
    },
    {
      "entropy": 5.575406122207641,
      "epoch": 1.8062723302897976,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046779021904457555,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1678849548101425,
      "num_tokens": 39858181.0,
      "step": 18200
    },
    {
      "entropy": 5.666563224792481,
      "epoch": 1.806768558951965,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004677719491173898,
      "loss": 5.2306,
      "mean_token_accuracy": 0.16810012012720107,
      "num_tokens": 39871104.0,
      "step": 18205
    },
    {
      "entropy": 5.6151848316192625,
      "epoch": 1.8072647876141326,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046775367440987506,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17777898758649827,
      "num_tokens": 39883323.0,
      "step": 18210
    },
    {
      "entropy": 5.529569673538208,
      "epoch": 1.8077610162763,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004677353949224846,
      "loss": 5.0733,
      "mean_token_accuracy": 0.18310048580169677,
      "num_tokens": 39894246.0,
      "step": 18215
    },
    {
      "entropy": 5.555595207214355,
      "epoch": 1.8082572449384675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046771711065567167,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17335924059152602,
      "num_tokens": 39905208.0,
      "step": 18220
    },
    {
      "entropy": 5.593649911880493,
      "epoch": 1.8087534736006352,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004676988216098896,
      "loss": 5.076,
      "mean_token_accuracy": 0.17467521429061889,
      "num_tokens": 39915531.0,
      "step": 18225
    },
    {
      "entropy": 5.583991479873657,
      "epoch": 1.8092497022628027,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004676805277855921,
      "loss": 5.1832,
      "mean_token_accuracy": 0.1734203204512596,
      "num_tokens": 39926241.0,
      "step": 18230
    },
    {
      "entropy": 5.556762170791626,
      "epoch": 1.8097459309249704,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004676622291832326,
      "loss": 5.0559,
      "mean_token_accuracy": 0.18116995692253113,
      "num_tokens": 39936500.0,
      "step": 18235
    },
    {
      "entropy": 5.559406709671021,
      "epoch": 1.8102421595871379,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046764392580326506,
      "loss": 5.0835,
      "mean_token_accuracy": 0.17226486653089523,
      "num_tokens": 39948865.0,
      "step": 18240
    },
    {
      "entropy": 5.579855012893677,
      "epoch": 1.8107383882493053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046762561764614316,
      "loss": 5.154,
      "mean_token_accuracy": 0.1694946438074112,
      "num_tokens": 39960626.0,
      "step": 18245
    },
    {
      "entropy": 5.628748655319214,
      "epoch": 1.8112346169114728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000467607304712321,
      "loss": 5.1074,
      "mean_token_accuracy": 0.16774771213531495,
      "num_tokens": 39971100.0,
      "step": 18250
    },
    {
      "entropy": 5.623596858978272,
      "epoch": 1.8117308455736403,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046758898700225283,
      "loss": 5.0804,
      "mean_token_accuracy": 0.17670081108808516,
      "num_tokens": 39981332.0,
      "step": 18255
    },
    {
      "entropy": 5.593695545196534,
      "epoch": 1.8122270742358078,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004675706645163927,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17623755931854249,
      "num_tokens": 39992739.0,
      "step": 18260
    },
    {
      "entropy": 5.570254325866699,
      "epoch": 1.8127233028979752,
      "grad_norm": 1.109375,
      "learning_rate": 0.000467552337255195,
      "loss": 5.0247,
      "mean_token_accuracy": 0.1837774932384491,
      "num_tokens": 40002841.0,
      "step": 18265
    },
    {
      "entropy": 5.665992498397827,
      "epoch": 1.813219531560143,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004675340052191144,
      "loss": 5.173,
      "mean_token_accuracy": 0.1759851634502411,
      "num_tokens": 40014176.0,
      "step": 18270
    },
    {
      "entropy": 5.583106946945191,
      "epoch": 1.8137157602223104,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004675156684086052,
      "loss": 5.0576,
      "mean_token_accuracy": 0.17616566866636277,
      "num_tokens": 40024842.0,
      "step": 18275
    },
    {
      "entropy": 5.573722553253174,
      "epoch": 1.814211988884478,
      "grad_norm": 1.0,
      "learning_rate": 0.0004674973268241224,
      "loss": 5.0743,
      "mean_token_accuracy": 0.17989588379859925,
      "num_tokens": 40034838.0,
      "step": 18280
    },
    {
      "entropy": 5.52345118522644,
      "epoch": 1.8147082175466456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004674789804661205,
      "loss": 5.0869,
      "mean_token_accuracy": 0.17697727382183076,
      "num_tokens": 40045019.0,
      "step": 18285
    },
    {
      "entropy": 5.625952672958374,
      "epoch": 1.815204446208813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046746062933505477,
      "loss": 5.2349,
      "mean_token_accuracy": 0.16060252338647843,
      "num_tokens": 40057283.0,
      "step": 18290
    },
    {
      "entropy": 5.66383957862854,
      "epoch": 1.8157006748709805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004674422734313801,
      "loss": 5.1231,
      "mean_token_accuracy": 0.1726805031299591,
      "num_tokens": 40067542.0,
      "step": 18295
    },
    {
      "entropy": 5.639804697036743,
      "epoch": 1.816196903533148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046742391275555176,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1799417182803154,
      "num_tokens": 40077538.0,
      "step": 18300
    },
    {
      "entropy": 5.589988040924072,
      "epoch": 1.8166931321953155,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000467405547308025,
      "loss": 5.0481,
      "mean_token_accuracy": 0.18016397655010224,
      "num_tokens": 40088206.0,
      "step": 18305
    },
    {
      "entropy": 5.558098602294922,
      "epoch": 1.8171893608574832,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004673871770892553,
      "loss": 5.1591,
      "mean_token_accuracy": 0.16992669850587844,
      "num_tokens": 40100510.0,
      "step": 18310
    },
    {
      "entropy": 5.598103952407837,
      "epoch": 1.8176855895196506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046736880209969814,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17568527162075043,
      "num_tokens": 40110759.0,
      "step": 18315
    },
    {
      "entropy": 5.572440052032471,
      "epoch": 1.8181818181818183,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046735042233980923,
      "loss": 5.1384,
      "mean_token_accuracy": 0.1746988996863365,
      "num_tokens": 40122140.0,
      "step": 18320
    },
    {
      "entropy": 5.661036968231201,
      "epoch": 1.8186780468439858,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046733203781004426,
      "loss": 5.1705,
      "mean_token_accuracy": 0.173691326379776,
      "num_tokens": 40133479.0,
      "step": 18325
    },
    {
      "entropy": 5.606528949737549,
      "epoch": 1.8191742755061533,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046731364851085923,
      "loss": 5.0899,
      "mean_token_accuracy": 0.17181296050548553,
      "num_tokens": 40144776.0,
      "step": 18330
    },
    {
      "entropy": 5.669846391677856,
      "epoch": 1.8196705041683208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004672952544427101,
      "loss": 5.1356,
      "mean_token_accuracy": 0.17741450071334838,
      "num_tokens": 40155767.0,
      "step": 18335
    },
    {
      "entropy": 5.661758852005005,
      "epoch": 1.8201667328304882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046727685560605305,
      "loss": 5.1968,
      "mean_token_accuracy": 0.16904802918434142,
      "num_tokens": 40167040.0,
      "step": 18340
    },
    {
      "entropy": 5.589064264297486,
      "epoch": 1.8206629614926557,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004672584520013443,
      "loss": 5.07,
      "mean_token_accuracy": 0.18321648091077805,
      "num_tokens": 40177983.0,
      "step": 18345
    },
    {
      "entropy": 5.632083082199097,
      "epoch": 1.8211591901548232,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004672400436290402,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17063151746988298,
      "num_tokens": 40187781.0,
      "step": 18350
    },
    {
      "entropy": 5.579588603973389,
      "epoch": 1.821655418816991,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046722163048959725,
      "loss": 5.0924,
      "mean_token_accuracy": 0.168136066198349,
      "num_tokens": 40198186.0,
      "step": 18355
    },
    {
      "entropy": 5.533597135543824,
      "epoch": 1.8221516474791584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004672032125834721,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17321922183036803,
      "num_tokens": 40207656.0,
      "step": 18360
    },
    {
      "entropy": 5.5860552310943605,
      "epoch": 1.822647876141326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004671847899111214,
      "loss": 5.0794,
      "mean_token_accuracy": 0.1783778965473175,
      "num_tokens": 40219482.0,
      "step": 18365
    },
    {
      "entropy": 5.659935855865479,
      "epoch": 1.8231441048034935,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046716636247300206,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17707326710224153,
      "num_tokens": 40230623.0,
      "step": 18370
    },
    {
      "entropy": 5.564065980911255,
      "epoch": 1.823640333465661,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046714793026957096,
      "loss": 5.0279,
      "mean_token_accuracy": 0.1813066154718399,
      "num_tokens": 40240682.0,
      "step": 18375
    },
    {
      "entropy": 5.55633020401001,
      "epoch": 1.8241365621278285,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004671294933012853,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1732695519924164,
      "num_tokens": 40253085.0,
      "step": 18380
    },
    {
      "entropy": 5.584764623641968,
      "epoch": 1.824632790789996,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046711105156860223,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17818044126033783,
      "num_tokens": 40264056.0,
      "step": 18385
    },
    {
      "entropy": 5.651178169250488,
      "epoch": 1.8251290194521634,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046709260507197896,
      "loss": 5.1557,
      "mean_token_accuracy": 0.17710406631231307,
      "num_tokens": 40273974.0,
      "step": 18390
    },
    {
      "entropy": 5.576876211166382,
      "epoch": 1.8256252481143311,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046707415381187306,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17696934044361115,
      "num_tokens": 40284455.0,
      "step": 18395
    },
    {
      "entropy": 5.583896350860596,
      "epoch": 1.8261214767764986,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000467055697788742,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17523921728134156,
      "num_tokens": 40294598.0,
      "step": 18400
    },
    {
      "entropy": 5.627495574951172,
      "epoch": 1.826617705438666,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046703723700304345,
      "loss": 5.1223,
      "mean_token_accuracy": 0.17168800234794618,
      "num_tokens": 40306053.0,
      "step": 18405
    },
    {
      "entropy": 5.605404281616211,
      "epoch": 1.8271139341008338,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004670187714552353,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17551853358745576,
      "num_tokens": 40316714.0,
      "step": 18410
    },
    {
      "entropy": 5.6366295337677,
      "epoch": 1.8276101627630013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046700030114577533,
      "loss": 5.1282,
      "mean_token_accuracy": 0.16884971410036087,
      "num_tokens": 40327696.0,
      "step": 18415
    },
    {
      "entropy": 5.645223379135132,
      "epoch": 1.8281063914251687,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004669818260751217,
      "loss": 5.1259,
      "mean_token_accuracy": 0.17150892764329911,
      "num_tokens": 40338958.0,
      "step": 18420
    },
    {
      "entropy": 5.647758960723877,
      "epoch": 1.8286026200873362,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004669633462437324,
      "loss": 5.1999,
      "mean_token_accuracy": 0.16796416342258452,
      "num_tokens": 40350869.0,
      "step": 18425
    },
    {
      "entropy": 5.58916130065918,
      "epoch": 1.8290988487495037,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046694486165206573,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17230292707681655,
      "num_tokens": 40360932.0,
      "step": 18430
    },
    {
      "entropy": 5.597435712814331,
      "epoch": 1.8295950774116712,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004669263723005801,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17977680414915084,
      "num_tokens": 40372159.0,
      "step": 18435
    },
    {
      "entropy": 5.58283748626709,
      "epoch": 1.8300913060738389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046690787818973406,
      "loss": 5.114,
      "mean_token_accuracy": 0.18049076646566392,
      "num_tokens": 40385268.0,
      "step": 18440
    },
    {
      "entropy": 5.611348009109497,
      "epoch": 1.8305875347360063,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046688937931998613,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17418635487556458,
      "num_tokens": 40396585.0,
      "step": 18445
    },
    {
      "entropy": 5.604572105407715,
      "epoch": 1.831083763398174,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004668708756917951,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17792580872774125,
      "num_tokens": 40408348.0,
      "step": 18450
    },
    {
      "entropy": 5.663416910171509,
      "epoch": 1.8315799920603415,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046685236730561975,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17580384612083436,
      "num_tokens": 40419900.0,
      "step": 18455
    },
    {
      "entropy": 5.634486150741577,
      "epoch": 1.832076220722509,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046683385416191914,
      "loss": 5.1161,
      "mean_token_accuracy": 0.17820560485124587,
      "num_tokens": 40431110.0,
      "step": 18460
    },
    {
      "entropy": 5.529092788696289,
      "epoch": 1.8325724493846764,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046681533626115226,
      "loss": 5.027,
      "mean_token_accuracy": 0.18220625817775726,
      "num_tokens": 40442741.0,
      "step": 18465
    },
    {
      "entropy": 5.531091785430908,
      "epoch": 1.833068678046844,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004667968136037784,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18058040142059326,
      "num_tokens": 40453720.0,
      "step": 18470
    },
    {
      "entropy": 5.580479621887207,
      "epoch": 1.8335649067090114,
      "grad_norm": 0.875,
      "learning_rate": 0.00046677828619025687,
      "loss": 5.1536,
      "mean_token_accuracy": 0.16856296360492706,
      "num_tokens": 40466097.0,
      "step": 18475
    },
    {
      "entropy": 5.592116737365723,
      "epoch": 1.8340611353711789,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046675975402104706,
      "loss": 5.0903,
      "mean_token_accuracy": 0.17739699333906173,
      "num_tokens": 40475713.0,
      "step": 18480
    },
    {
      "entropy": 5.585403537750244,
      "epoch": 1.8345573640333466,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046674121709660855,
      "loss": 4.9915,
      "mean_token_accuracy": 0.18086368292570115,
      "num_tokens": 40486603.0,
      "step": 18485
    },
    {
      "entropy": 5.632459831237793,
      "epoch": 1.835053592695514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000466722675417401,
      "loss": 5.1864,
      "mean_token_accuracy": 0.17287475615739822,
      "num_tokens": 40498057.0,
      "step": 18490
    },
    {
      "entropy": 5.609178686141968,
      "epoch": 1.8355498213576817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046670412898388425,
      "loss": 5.1352,
      "mean_token_accuracy": 0.17064798027276992,
      "num_tokens": 40508798.0,
      "step": 18495
    },
    {
      "entropy": 5.551144981384278,
      "epoch": 1.8360460500198492,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046668557779651825,
      "loss": 5.0797,
      "mean_token_accuracy": 0.17619415521621704,
      "num_tokens": 40519736.0,
      "step": 18500
    },
    {
      "entropy": 5.602016925811768,
      "epoch": 1.8365422786820167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046666702185576295,
      "loss": 5.1201,
      "mean_token_accuracy": 0.17930333018302919,
      "num_tokens": 40531582.0,
      "step": 18505
    },
    {
      "entropy": 5.590337038040161,
      "epoch": 1.8370385073441842,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004666484611620785,
      "loss": 5.0375,
      "mean_token_accuracy": 0.17673102766275406,
      "num_tokens": 40542427.0,
      "step": 18510
    },
    {
      "entropy": 5.647511196136475,
      "epoch": 1.8375347360063516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004666298957159251,
      "loss": 5.2491,
      "mean_token_accuracy": 0.17346566468477248,
      "num_tokens": 40553432.0,
      "step": 18515
    },
    {
      "entropy": 5.617825889587403,
      "epoch": 1.8380309646685191,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046661132551776327,
      "loss": 5.1252,
      "mean_token_accuracy": 0.17052697390317917,
      "num_tokens": 40564995.0,
      "step": 18520
    },
    {
      "entropy": 5.609264230728149,
      "epoch": 1.8385271933306868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004665927505680535,
      "loss": 5.0903,
      "mean_token_accuracy": 0.1767307326197624,
      "num_tokens": 40574856.0,
      "step": 18525
    },
    {
      "entropy": 5.637595748901367,
      "epoch": 1.8390234219928543,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046657417086725625,
      "loss": 5.0932,
      "mean_token_accuracy": 0.18147078156471252,
      "num_tokens": 40585439.0,
      "step": 18530
    },
    {
      "entropy": 5.609993505477905,
      "epoch": 1.839519650655022,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004665555864158324,
      "loss": 5.0953,
      "mean_token_accuracy": 0.17818260192871094,
      "num_tokens": 40596275.0,
      "step": 18535
    },
    {
      "entropy": 5.553247404098511,
      "epoch": 1.8400158793171895,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004665369972142428,
      "loss": 5.0951,
      "mean_token_accuracy": 0.17518922090530395,
      "num_tokens": 40607586.0,
      "step": 18540
    },
    {
      "entropy": 5.5745409488677975,
      "epoch": 1.840512107979357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046651840326294845,
      "loss": 5.101,
      "mean_token_accuracy": 0.1798638269305229,
      "num_tokens": 40617339.0,
      "step": 18545
    },
    {
      "entropy": 5.669265985488892,
      "epoch": 1.8410083366415244,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004664998045624104,
      "loss": 5.0923,
      "mean_token_accuracy": 0.1848890095949173,
      "num_tokens": 40630426.0,
      "step": 18550
    },
    {
      "entropy": 5.633747196197509,
      "epoch": 1.8415045653036919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664812011130897,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17279489934444428,
      "num_tokens": 40640694.0,
      "step": 18555
    },
    {
      "entropy": 5.628920412063598,
      "epoch": 1.8420007939658594,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046646259291544787,
      "loss": 5.1717,
      "mean_token_accuracy": 0.16968259513378142,
      "num_tokens": 40651131.0,
      "step": 18560
    },
    {
      "entropy": 5.633462858200073,
      "epoch": 1.8424970226280268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004664439799699463,
      "loss": 5.2057,
      "mean_token_accuracy": 0.16909669637680053,
      "num_tokens": 40662228.0,
      "step": 18565
    },
    {
      "entropy": 5.615708494186402,
      "epoch": 1.8429932512901945,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046642536227704665,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17439660280942917,
      "num_tokens": 40672928.0,
      "step": 18570
    },
    {
      "entropy": 5.629876613616943,
      "epoch": 1.843489479952362,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046640673983721037,
      "loss": 5.1624,
      "mean_token_accuracy": 0.16528469771146775,
      "num_tokens": 40684578.0,
      "step": 18575
    },
    {
      "entropy": 5.591208744049072,
      "epoch": 1.8439857086145297,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046638811265089944,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17694645375013351,
      "num_tokens": 40694844.0,
      "step": 18580
    },
    {
      "entropy": 5.568539094924927,
      "epoch": 1.8444819372766972,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004663694807185758,
      "loss": 5.0743,
      "mean_token_accuracy": 0.17823754847049714,
      "num_tokens": 40706494.0,
      "step": 18585
    },
    {
      "entropy": 5.61638298034668,
      "epoch": 1.8449781659388647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046635084404070135,
      "loss": 5.1149,
      "mean_token_accuracy": 0.17234821617603302,
      "num_tokens": 40717092.0,
      "step": 18590
    },
    {
      "entropy": 5.548984336853027,
      "epoch": 1.8454743946010321,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004663322026177382,
      "loss": 5.0464,
      "mean_token_accuracy": 0.1836048260331154,
      "num_tokens": 40729131.0,
      "step": 18595
    },
    {
      "entropy": 5.536949348449707,
      "epoch": 1.8459706232631996,
      "grad_norm": 1.0,
      "learning_rate": 0.00046631355645014875,
      "loss": 5.037,
      "mean_token_accuracy": 0.17843838334083556,
      "num_tokens": 40739918.0,
      "step": 18600
    },
    {
      "entropy": 5.6413524627685545,
      "epoch": 1.846466851925367,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046629490553839537,
      "loss": 5.1495,
      "mean_token_accuracy": 0.17176228761672974,
      "num_tokens": 40752136.0,
      "step": 18605
    },
    {
      "entropy": 5.625709772109985,
      "epoch": 1.8469630805875348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004662762498829405,
      "loss": 5.054,
      "mean_token_accuracy": 0.18210064619779587,
      "num_tokens": 40761613.0,
      "step": 18610
    },
    {
      "entropy": 5.570508098602295,
      "epoch": 1.8474593092497023,
      "grad_norm": 1.0,
      "learning_rate": 0.0004662575894842469,
      "loss": 5.103,
      "mean_token_accuracy": 0.17827748656272888,
      "num_tokens": 40773405.0,
      "step": 18615
    },
    {
      "entropy": 5.619103527069091,
      "epoch": 1.84795553791187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046623892434277713,
      "loss": 5.1932,
      "mean_token_accuracy": 0.1704985409975052,
      "num_tokens": 40784795.0,
      "step": 18620
    },
    {
      "entropy": 5.63733172416687,
      "epoch": 1.8484517665740374,
      "grad_norm": 1.0,
      "learning_rate": 0.0004662202544589941,
      "loss": 5.1429,
      "mean_token_accuracy": 0.17307597100734712,
      "num_tokens": 40795946.0,
      "step": 18625
    },
    {
      "entropy": 5.613544559478759,
      "epoch": 1.848947995236205,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046620157983336087,
      "loss": 5.0705,
      "mean_token_accuracy": 0.18215764909982682,
      "num_tokens": 40807854.0,
      "step": 18630
    },
    {
      "entropy": 5.5705259323120115,
      "epoch": 1.8494442238983724,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004661829004663404,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17655593752861024,
      "num_tokens": 40818677.0,
      "step": 18635
    },
    {
      "entropy": 5.6210705757141115,
      "epoch": 1.8499404525605398,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004661642163583961,
      "loss": 5.1156,
      "mean_token_accuracy": 0.17027992010116577,
      "num_tokens": 40829927.0,
      "step": 18640
    },
    {
      "entropy": 5.555161190032959,
      "epoch": 1.8504366812227073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046614552750999093,
      "loss": 5.0477,
      "mean_token_accuracy": 0.17539135068655015,
      "num_tokens": 40840661.0,
      "step": 18645
    },
    {
      "entropy": 5.642641305923462,
      "epoch": 1.8509329098848748,
      "grad_norm": 1.0,
      "learning_rate": 0.0004661268339215887,
      "loss": 5.1443,
      "mean_token_accuracy": 0.17108574360609055,
      "num_tokens": 40851985.0,
      "step": 18650
    },
    {
      "entropy": 5.65367488861084,
      "epoch": 1.8514291385470425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661081355936528,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17627278566360474,
      "num_tokens": 40863546.0,
      "step": 18655
    },
    {
      "entropy": 5.64368839263916,
      "epoch": 1.85192536720921,
      "grad_norm": 1.015625,
      "learning_rate": 0.000466089432526647,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17552102357149124,
      "num_tokens": 40873555.0,
      "step": 18660
    },
    {
      "entropy": 5.667524290084839,
      "epoch": 1.8524215958713777,
      "grad_norm": 0.828125,
      "learning_rate": 0.00046607072472103493,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1731606900691986,
      "num_tokens": 40887483.0,
      "step": 18665
    },
    {
      "entropy": 5.657082796096802,
      "epoch": 1.8529178245335451,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046605201217728067,
      "loss": 5.2128,
      "mean_token_accuracy": 0.17065927684307097,
      "num_tokens": 40900412.0,
      "step": 18670
    },
    {
      "entropy": 5.626799154281616,
      "epoch": 1.8534140531957126,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046603329489584817,
      "loss": 5.0892,
      "mean_token_accuracy": 0.17808571755886077,
      "num_tokens": 40910069.0,
      "step": 18675
    },
    {
      "entropy": 5.6383342266082765,
      "epoch": 1.85391028185788,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004660145728772016,
      "loss": 5.0888,
      "mean_token_accuracy": 0.18198371976613997,
      "num_tokens": 40921321.0,
      "step": 18680
    },
    {
      "entropy": 5.612284851074219,
      "epoch": 1.8544065105200476,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004659958461218052,
      "loss": 5.0635,
      "mean_token_accuracy": 0.1802036315202713,
      "num_tokens": 40931564.0,
      "step": 18685
    },
    {
      "entropy": 5.587565517425537,
      "epoch": 1.854902739182215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046597711463012344,
      "loss": 5.0733,
      "mean_token_accuracy": 0.17202044129371644,
      "num_tokens": 40942537.0,
      "step": 18690
    },
    {
      "entropy": 5.635299301147461,
      "epoch": 1.8553989678443827,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004659583784026207,
      "loss": 5.1159,
      "mean_token_accuracy": 0.17749452143907546,
      "num_tokens": 40953583.0,
      "step": 18695
    },
    {
      "entropy": 5.592890119552612,
      "epoch": 1.8558951965065502,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046593963743976163,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17537719011306763,
      "num_tokens": 40964438.0,
      "step": 18700
    },
    {
      "entropy": 5.575043153762818,
      "epoch": 1.8563914251687177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000465920891742011,
      "loss": 5.0391,
      "mean_token_accuracy": 0.17703295648097991,
      "num_tokens": 40975154.0,
      "step": 18705
    },
    {
      "entropy": 5.571862077713012,
      "epoch": 1.8568876538308854,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004659021413098337,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17307765334844588,
      "num_tokens": 40986498.0,
      "step": 18710
    },
    {
      "entropy": 5.583320808410645,
      "epoch": 1.8573838824930529,
      "grad_norm": 1.0,
      "learning_rate": 0.00046588338614369465,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1719256803393364,
      "num_tokens": 40996095.0,
      "step": 18715
    },
    {
      "entropy": 5.66254153251648,
      "epoch": 1.8578801111552203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046586462624405897,
      "loss": 5.122,
      "mean_token_accuracy": 0.1704502746462822,
      "num_tokens": 41007583.0,
      "step": 18720
    },
    {
      "entropy": 5.649216938018799,
      "epoch": 1.8583763398173878,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004658458616113918,
      "loss": 5.1312,
      "mean_token_accuracy": 0.17315282225608825,
      "num_tokens": 41018251.0,
      "step": 18725
    },
    {
      "entropy": 5.650521564483642,
      "epoch": 1.8588725684795553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004658270922461585,
      "loss": 5.1437,
      "mean_token_accuracy": 0.17611207515001298,
      "num_tokens": 41029144.0,
      "step": 18730
    },
    {
      "entropy": 5.624125623703003,
      "epoch": 1.8593687971417228,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046580831814882446,
      "loss": 5.1019,
      "mean_token_accuracy": 0.16931798309087753,
      "num_tokens": 41039385.0,
      "step": 18735
    },
    {
      "entropy": 5.5743077278137205,
      "epoch": 1.8598650258038905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046578953931985536,
      "loss": 5.0497,
      "mean_token_accuracy": 0.17947263568639754,
      "num_tokens": 41049175.0,
      "step": 18740
    },
    {
      "entropy": 5.591348934173584,
      "epoch": 1.860361254466058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004657707557597167,
      "loss": 5.0847,
      "mean_token_accuracy": 0.17913589179515838,
      "num_tokens": 41060517.0,
      "step": 18745
    },
    {
      "entropy": 5.57352204322815,
      "epoch": 1.8608574831282256,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046575196746887447,
      "loss": 5.1109,
      "mean_token_accuracy": 0.1722010463476181,
      "num_tokens": 41070957.0,
      "step": 18750
    },
    {
      "entropy": 5.6214885234832765,
      "epoch": 1.861353711790393,
      "grad_norm": 0.84375,
      "learning_rate": 0.00046573317444779445,
      "loss": 5.1755,
      "mean_token_accuracy": 0.16909410655498505,
      "num_tokens": 41083079.0,
      "step": 18755
    },
    {
      "entropy": 5.694080495834351,
      "epoch": 1.8618499404525606,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004657143766969427,
      "loss": 5.206,
      "mean_token_accuracy": 0.16830307990312576,
      "num_tokens": 41095379.0,
      "step": 18760
    },
    {
      "entropy": 5.564210367202759,
      "epoch": 1.862346169114728,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004656955742167853,
      "loss": 5.0631,
      "mean_token_accuracy": 0.18218172043561937,
      "num_tokens": 41106412.0,
      "step": 18765
    },
    {
      "entropy": 5.616793203353882,
      "epoch": 1.8628423977768955,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046567676700778865,
      "loss": 5.1498,
      "mean_token_accuracy": 0.17875962108373641,
      "num_tokens": 41116820.0,
      "step": 18770
    },
    {
      "entropy": 5.606196784973145,
      "epoch": 1.863338626439063,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046565795507041904,
      "loss": 5.0983,
      "mean_token_accuracy": 0.17444873452186585,
      "num_tokens": 41128657.0,
      "step": 18775
    },
    {
      "entropy": 5.660610771179199,
      "epoch": 1.8638348551012305,
      "grad_norm": 0.984375,
      "learning_rate": 0.000465639138405143,
      "loss": 5.1138,
      "mean_token_accuracy": 0.17387838363647462,
      "num_tokens": 41139700.0,
      "step": 18780
    },
    {
      "entropy": 5.653067493438721,
      "epoch": 1.8643310837633982,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046562031701242703,
      "loss": 5.1273,
      "mean_token_accuracy": 0.17360107600688934,
      "num_tokens": 41150018.0,
      "step": 18785
    },
    {
      "entropy": 5.597535848617554,
      "epoch": 1.8648273124255657,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046560149089273803,
      "loss": 5.1611,
      "mean_token_accuracy": 0.16923482567071915,
      "num_tokens": 41160683.0,
      "step": 18790
    },
    {
      "entropy": 5.599908876419067,
      "epoch": 1.8653235410877333,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004655826600465427,
      "loss": 5.1177,
      "mean_token_accuracy": 0.17670647650957108,
      "num_tokens": 41171602.0,
      "step": 18795
    },
    {
      "entropy": 5.613927936553955,
      "epoch": 1.8658197697499008,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046556382447430816,
      "loss": 5.096,
      "mean_token_accuracy": 0.17428264319896697,
      "num_tokens": 41182087.0,
      "step": 18800
    },
    {
      "entropy": 5.571880865097046,
      "epoch": 1.8663159984120683,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046554498417650134,
      "loss": 5.0481,
      "mean_token_accuracy": 0.18275932669639589,
      "num_tokens": 41194472.0,
      "step": 18805
    },
    {
      "entropy": 5.60789647102356,
      "epoch": 1.8668122270742358,
      "grad_norm": 1.0,
      "learning_rate": 0.0004655261391535895,
      "loss": 5.0958,
      "mean_token_accuracy": 0.17838606536388396,
      "num_tokens": 41205082.0,
      "step": 18810
    },
    {
      "entropy": 5.620247220993042,
      "epoch": 1.8673084557364032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004655072894060399,
      "loss": 5.0923,
      "mean_token_accuracy": 0.18180215805768968,
      "num_tokens": 41214362.0,
      "step": 18815
    },
    {
      "entropy": 5.620284557342529,
      "epoch": 1.8678046843985707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004654884349343201,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17953689098358155,
      "num_tokens": 41225108.0,
      "step": 18820
    },
    {
      "entropy": 5.61592812538147,
      "epoch": 1.8683009130607384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046546957573889757,
      "loss": 5.0785,
      "mean_token_accuracy": 0.18160045444965361,
      "num_tokens": 41235771.0,
      "step": 18825
    },
    {
      "entropy": 5.545548915863037,
      "epoch": 1.868797141722906,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046545071182024,
      "loss": 5.0823,
      "mean_token_accuracy": 0.1780400276184082,
      "num_tokens": 41245752.0,
      "step": 18830
    },
    {
      "entropy": 5.645634174346924,
      "epoch": 1.8692933703850736,
      "grad_norm": 1.0,
      "learning_rate": 0.00046543184317881516,
      "loss": 5.2059,
      "mean_token_accuracy": 0.16750630885362625,
      "num_tokens": 41255704.0,
      "step": 18835
    },
    {
      "entropy": 5.624906778335571,
      "epoch": 1.869789599047241,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000465412969815091,
      "loss": 5.0791,
      "mean_token_accuracy": 0.17816881686449051,
      "num_tokens": 41265742.0,
      "step": 18840
    },
    {
      "entropy": 5.572498559951782,
      "epoch": 1.8702858277094085,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004653940917295354,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17346111685037613,
      "num_tokens": 41276166.0,
      "step": 18845
    },
    {
      "entropy": 5.604753494262695,
      "epoch": 1.870782056371576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653752089226166,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17507373988628389,
      "num_tokens": 41287047.0,
      "step": 18850
    },
    {
      "entropy": 5.597584772109985,
      "epoch": 1.8712782850337435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004653563213948029,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17361344993114472,
      "num_tokens": 41297317.0,
      "step": 18855
    },
    {
      "entropy": 5.635601663589478,
      "epoch": 1.871774513695911,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046533742914656254,
      "loss": 5.1493,
      "mean_token_accuracy": 0.17201642245054244,
      "num_tokens": 41308107.0,
      "step": 18860
    },
    {
      "entropy": 5.613266515731811,
      "epoch": 1.8722707423580784,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046531853217836414,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18883918970823288,
      "num_tokens": 41320026.0,
      "step": 18865
    },
    {
      "entropy": 5.672946166992188,
      "epoch": 1.8727669710202461,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046529963049067623,
      "loss": 5.1549,
      "mean_token_accuracy": 0.1735077142715454,
      "num_tokens": 41331492.0,
      "step": 18870
    },
    {
      "entropy": 5.560990238189698,
      "epoch": 1.8732631996824136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046528072408396747,
      "loss": 5.023,
      "mean_token_accuracy": 0.1755066394805908,
      "num_tokens": 41342258.0,
      "step": 18875
    },
    {
      "entropy": 5.63569393157959,
      "epoch": 1.8737594283445813,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004652618129587069,
      "loss": 5.1019,
      "mean_token_accuracy": 0.1788051873445511,
      "num_tokens": 41353041.0,
      "step": 18880
    },
    {
      "entropy": 5.59224419593811,
      "epoch": 1.8742556570067488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046524289711536327,
      "loss": 5.1324,
      "mean_token_accuracy": 0.176389442384243,
      "num_tokens": 41364517.0,
      "step": 18885
    },
    {
      "entropy": 5.587232160568237,
      "epoch": 1.8747518856689163,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004652239765544057,
      "loss": 5.1279,
      "mean_token_accuracy": 0.1794194534420967,
      "num_tokens": 41376241.0,
      "step": 18890
    },
    {
      "entropy": 5.594835186004639,
      "epoch": 1.8752481143310837,
      "grad_norm": 1.0,
      "learning_rate": 0.00046520505127630346,
      "loss": 5.1279,
      "mean_token_accuracy": 0.17645482271909713,
      "num_tokens": 41387627.0,
      "step": 18895
    },
    {
      "entropy": 5.635338926315308,
      "epoch": 1.8757443429932512,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046518612128152576,
      "loss": 5.09,
      "mean_token_accuracy": 0.17924275994300842,
      "num_tokens": 41399803.0,
      "step": 18900
    },
    {
      "entropy": 5.576120853424072,
      "epoch": 1.8762405716554187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004651671865705421,
      "loss": 5.093,
      "mean_token_accuracy": 0.1725542053580284,
      "num_tokens": 41409735.0,
      "step": 18905
    },
    {
      "entropy": 5.586027145385742,
      "epoch": 1.8767368003175864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046514824714382204,
      "loss": 5.0329,
      "mean_token_accuracy": 0.17982133477926254,
      "num_tokens": 41420928.0,
      "step": 18910
    },
    {
      "entropy": 5.59341778755188,
      "epoch": 1.8772330289797539,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004651293030018351,
      "loss": 5.118,
      "mean_token_accuracy": 0.17690795063972473,
      "num_tokens": 41432309.0,
      "step": 18915
    },
    {
      "entropy": 5.6174133777618405,
      "epoch": 1.8777292576419216,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046511035414505115,
      "loss": 5.0886,
      "mean_token_accuracy": 0.17593552321195602,
      "num_tokens": 41443064.0,
      "step": 18920
    },
    {
      "entropy": 5.618053865432739,
      "epoch": 1.878225486304089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004650914005739401,
      "loss": 5.1061,
      "mean_token_accuracy": 0.1772657960653305,
      "num_tokens": 41452698.0,
      "step": 18925
    },
    {
      "entropy": 5.575437545776367,
      "epoch": 1.8787217149662565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004650724422889719,
      "loss": 5.0707,
      "mean_token_accuracy": 0.17290614992380143,
      "num_tokens": 41463253.0,
      "step": 18930
    },
    {
      "entropy": 5.596289730072021,
      "epoch": 1.879217943628424,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004650534792906167,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17127608507871628,
      "num_tokens": 41473635.0,
      "step": 18935
    },
    {
      "entropy": 5.6114250183105465,
      "epoch": 1.8797141722905915,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004650345115793447,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1777273252606392,
      "num_tokens": 41485705.0,
      "step": 18940
    },
    {
      "entropy": 5.6716677188873295,
      "epoch": 1.880210400952759,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046501553915562644,
      "loss": 5.1823,
      "mean_token_accuracy": 0.1667814299464226,
      "num_tokens": 41497604.0,
      "step": 18945
    },
    {
      "entropy": 5.5405285358428955,
      "epoch": 1.8807066296149264,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046499656201993213,
      "loss": 5.0393,
      "mean_token_accuracy": 0.17754347622394562,
      "num_tokens": 41508960.0,
      "step": 18950
    },
    {
      "entropy": 5.578396892547607,
      "epoch": 1.881202858277094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046497758017273255,
      "loss": 5.1223,
      "mean_token_accuracy": 0.1760234221816063,
      "num_tokens": 41519494.0,
      "step": 18955
    },
    {
      "entropy": 5.6607567310333256,
      "epoch": 1.8816990869392616,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046495859361449834,
      "loss": 5.1232,
      "mean_token_accuracy": 0.1643357753753662,
      "num_tokens": 41530731.0,
      "step": 18960
    },
    {
      "entropy": 5.592073631286621,
      "epoch": 1.8821953156014293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004649396023457003,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17426057904958725,
      "num_tokens": 41541839.0,
      "step": 18965
    },
    {
      "entropy": 5.591901350021362,
      "epoch": 1.8826915442635968,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004649206063668094,
      "loss": 5.0957,
      "mean_token_accuracy": 0.18153657913208007,
      "num_tokens": 41552413.0,
      "step": 18970
    },
    {
      "entropy": 5.682925653457642,
      "epoch": 1.8831877729257642,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004649016056782968,
      "loss": 5.1987,
      "mean_token_accuracy": 0.17766997069120408,
      "num_tokens": 41563400.0,
      "step": 18975
    },
    {
      "entropy": 5.5562389373779295,
      "epoch": 1.8836840015879317,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004648826002806334,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17639435082674026,
      "num_tokens": 41575902.0,
      "step": 18980
    },
    {
      "entropy": 5.591698455810547,
      "epoch": 1.8841802302500992,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004648635901742909,
      "loss": 5.1208,
      "mean_token_accuracy": 0.174460206925869,
      "num_tokens": 41587613.0,
      "step": 18985
    },
    {
      "entropy": 5.575967788696289,
      "epoch": 1.8846764589122667,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004648445753597403,
      "loss": 5.0968,
      "mean_token_accuracy": 0.18108513802289963,
      "num_tokens": 41598674.0,
      "step": 18990
    },
    {
      "entropy": 5.630114030838013,
      "epoch": 1.8851726875744343,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046482555583745336,
      "loss": 5.142,
      "mean_token_accuracy": 0.17452540397644042,
      "num_tokens": 41609891.0,
      "step": 18995
    },
    {
      "entropy": 5.5610127449035645,
      "epoch": 1.8856689162366018,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004648065316079017,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18263779878616332,
      "num_tokens": 41622066.0,
      "step": 19000
    },
    {
      "entropy": 5.558603763580322,
      "epoch": 1.8861651448987695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046478750267155703,
      "loss": 5.0379,
      "mean_token_accuracy": 0.1814663052558899,
      "num_tokens": 41633182.0,
      "step": 19005
    },
    {
      "entropy": 5.525222969055176,
      "epoch": 1.886661373560937,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046476846902889125,
      "loss": 5.0627,
      "mean_token_accuracy": 0.18626712560653685,
      "num_tokens": 41644059.0,
      "step": 19010
    },
    {
      "entropy": 5.538286828994751,
      "epoch": 1.8871576022231045,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004647494306803764,
      "loss": 4.9998,
      "mean_token_accuracy": 0.17907464504241943,
      "num_tokens": 41654225.0,
      "step": 19015
    },
    {
      "entropy": 5.596899557113647,
      "epoch": 1.887653830885272,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046473038762648446,
      "loss": 5.1481,
      "mean_token_accuracy": 0.17314939051866532,
      "num_tokens": 41666598.0,
      "step": 19020
    },
    {
      "entropy": 5.622112512588501,
      "epoch": 1.8881500595474394,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046471133986768784,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17299881726503372,
      "num_tokens": 41677408.0,
      "step": 19025
    },
    {
      "entropy": 5.570197105407715,
      "epoch": 1.888646288209607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046469228740445874,
      "loss": 5.0217,
      "mean_token_accuracy": 0.1787801742553711,
      "num_tokens": 41687350.0,
      "step": 19030
    },
    {
      "entropy": 5.600446081161499,
      "epoch": 1.8891425168717744,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046467323023726965,
      "loss": 5.1181,
      "mean_token_accuracy": 0.1721274122595787,
      "num_tokens": 41698012.0,
      "step": 19035
    },
    {
      "entropy": 5.545606184005737,
      "epoch": 1.889638745533942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046465416836659313,
      "loss": 5.0629,
      "mean_token_accuracy": 0.17652511298656465,
      "num_tokens": 41708355.0,
      "step": 19040
    },
    {
      "entropy": 5.661450672149658,
      "epoch": 1.8901349741961095,
      "grad_norm": 1.09375,
      "learning_rate": 0.000464635101792902,
      "loss": 5.1765,
      "mean_token_accuracy": 0.16964072436094285,
      "num_tokens": 41720821.0,
      "step": 19045
    },
    {
      "entropy": 5.658682107925415,
      "epoch": 1.8906312028582772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046461603051666883,
      "loss": 5.1562,
      "mean_token_accuracy": 0.1655530408024788,
      "num_tokens": 41731167.0,
      "step": 19050
    },
    {
      "entropy": 5.611315155029297,
      "epoch": 1.8911274315204447,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046459695453836676,
      "loss": 5.1126,
      "mean_token_accuracy": 0.17031550407409668,
      "num_tokens": 41742175.0,
      "step": 19055
    },
    {
      "entropy": 5.583570337295532,
      "epoch": 1.8916236601826122,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004645778738584688,
      "loss": 5.0652,
      "mean_token_accuracy": 0.17596072405576707,
      "num_tokens": 41752435.0,
      "step": 19060
    },
    {
      "entropy": 5.6558922290802,
      "epoch": 1.8921198888447797,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004645587884774481,
      "loss": 5.1112,
      "mean_token_accuracy": 0.17589745968580245,
      "num_tokens": 41762793.0,
      "step": 19065
    },
    {
      "entropy": 5.615695476531982,
      "epoch": 1.8926161175069471,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004645396983957778,
      "loss": 5.1335,
      "mean_token_accuracy": 0.1760098785161972,
      "num_tokens": 41774139.0,
      "step": 19070
    },
    {
      "entropy": 5.678126573562622,
      "epoch": 1.8931123461691146,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046452060361393147,
      "loss": 5.1838,
      "mean_token_accuracy": 0.16264623105525972,
      "num_tokens": 41784908.0,
      "step": 19075
    },
    {
      "entropy": 5.572958660125733,
      "epoch": 1.8936085748312823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004645015041323825,
      "loss": 5.125,
      "mean_token_accuracy": 0.16994790136814117,
      "num_tokens": 41794742.0,
      "step": 19080
    },
    {
      "entropy": 5.669703722000122,
      "epoch": 1.8941048034934498,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046448239995160455,
      "loss": 5.1989,
      "mean_token_accuracy": 0.174451020359993,
      "num_tokens": 41805188.0,
      "step": 19085
    },
    {
      "entropy": 5.652312135696411,
      "epoch": 1.8946010321556173,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004644632910720715,
      "loss": 5.075,
      "mean_token_accuracy": 0.18640017062425612,
      "num_tokens": 41814689.0,
      "step": 19090
    },
    {
      "entropy": 5.581059217453003,
      "epoch": 1.895097260817785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000464444177494257,
      "loss": 5.0573,
      "mean_token_accuracy": 0.18102738708257676,
      "num_tokens": 41826623.0,
      "step": 19095
    },
    {
      "entropy": 5.648148822784424,
      "epoch": 1.8955934894799524,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004644250592186351,
      "loss": 5.1215,
      "mean_token_accuracy": 0.1805953025817871,
      "num_tokens": 41838045.0,
      "step": 19100
    },
    {
      "entropy": 5.543940734863281,
      "epoch": 1.89608971814212,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046440593624568,
      "loss": 5.0134,
      "mean_token_accuracy": 0.1823401093482971,
      "num_tokens": 41849905.0,
      "step": 19105
    },
    {
      "entropy": 5.6512871265411375,
      "epoch": 1.8965859468042874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004643868085758657,
      "loss": 5.1515,
      "mean_token_accuracy": 0.16913365572690964,
      "num_tokens": 41861679.0,
      "step": 19110
    },
    {
      "entropy": 5.636925840377808,
      "epoch": 1.8970821754664549,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046436767620966675,
      "loss": 5.1596,
      "mean_token_accuracy": 0.17035914063453675,
      "num_tokens": 41873111.0,
      "step": 19115
    },
    {
      "entropy": 5.6304350852966305,
      "epoch": 1.8975784041286223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046434853914755744,
      "loss": 5.1291,
      "mean_token_accuracy": 0.1689220905303955,
      "num_tokens": 41884181.0,
      "step": 19120
    },
    {
      "entropy": 5.65972580909729,
      "epoch": 1.89807463279079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046432939739001237,
      "loss": 5.0623,
      "mean_token_accuracy": 0.17797104269266129,
      "num_tokens": 41895086.0,
      "step": 19125
    },
    {
      "entropy": 5.5479048728942875,
      "epoch": 1.8985708614529575,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046431025093750613,
      "loss": 5.0409,
      "mean_token_accuracy": 0.18276697397232056,
      "num_tokens": 41907476.0,
      "step": 19130
    },
    {
      "entropy": 5.645580148696899,
      "epoch": 1.8990670901151252,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046429109979051367,
      "loss": 5.195,
      "mean_token_accuracy": 0.17150511890649794,
      "num_tokens": 41918992.0,
      "step": 19135
    },
    {
      "entropy": 5.726856899261475,
      "epoch": 1.8995633187772927,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046427194394950985,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16880234330892563,
      "num_tokens": 41929332.0,
      "step": 19140
    },
    {
      "entropy": 5.540611124038696,
      "epoch": 1.9000595474394602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004642527834149696,
      "loss": 4.9578,
      "mean_token_accuracy": 0.18486660718917847,
      "num_tokens": 41940363.0,
      "step": 19145
    },
    {
      "entropy": 5.592981290817261,
      "epoch": 1.9005557761016276,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046423361818736814,
      "loss": 5.104,
      "mean_token_accuracy": 0.1766674906015396,
      "num_tokens": 41951491.0,
      "step": 19150
    },
    {
      "entropy": 5.6329371452331545,
      "epoch": 1.901052004763795,
      "grad_norm": 1.0,
      "learning_rate": 0.0004642144482671807,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18512619584798812,
      "num_tokens": 41962312.0,
      "step": 19155
    },
    {
      "entropy": 5.623608541488648,
      "epoch": 1.9015482334259626,
      "grad_norm": 1.8125,
      "learning_rate": 0.00046419527365488275,
      "loss": 5.121,
      "mean_token_accuracy": 0.1747841164469719,
      "num_tokens": 41972455.0,
      "step": 19160
    },
    {
      "entropy": 5.5410747051239015,
      "epoch": 1.90204446208813,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004641760943509496,
      "loss": 5.1697,
      "mean_token_accuracy": 0.1722515419125557,
      "num_tokens": 41984608.0,
      "step": 19165
    },
    {
      "entropy": 5.6854644298553465,
      "epoch": 1.9025406907502977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000464156910355857,
      "loss": 5.181,
      "mean_token_accuracy": 0.1706845611333847,
      "num_tokens": 41996838.0,
      "step": 19170
    },
    {
      "entropy": 5.6990941047668455,
      "epoch": 1.9030369194124652,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046413772167008057,
      "loss": 5.0876,
      "mean_token_accuracy": 0.16900564432144166,
      "num_tokens": 42007720.0,
      "step": 19175
    },
    {
      "entropy": 5.5603560447692875,
      "epoch": 1.903533148074633,
      "grad_norm": 1.0,
      "learning_rate": 0.00046411852829409626,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17097101360559464,
      "num_tokens": 42018859.0,
      "step": 19180
    },
    {
      "entropy": 5.5589423179626465,
      "epoch": 1.9040293767368004,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046409933022837984,
      "loss": 5.1048,
      "mean_token_accuracy": 0.17865864783525467,
      "num_tokens": 42029616.0,
      "step": 19185
    },
    {
      "entropy": 5.6653107643127445,
      "epoch": 1.9045256053989679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004640801274734076,
      "loss": 5.1243,
      "mean_token_accuracy": 0.16712254881858826,
      "num_tokens": 42039870.0,
      "step": 19190
    },
    {
      "entropy": 5.714418172836304,
      "epoch": 1.9050218340611353,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640609200296557,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17424734830856323,
      "num_tokens": 42050418.0,
      "step": 19195
    },
    {
      "entropy": 5.5511932373046875,
      "epoch": 1.9055180627233028,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004640417078976003,
      "loss": 4.9067,
      "mean_token_accuracy": 0.1848577544093132,
      "num_tokens": 42062273.0,
      "step": 19200
    },
    {
      "entropy": 5.601956176757812,
      "epoch": 1.9060142913854703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004640224910777179,
      "loss": 5.165,
      "mean_token_accuracy": 0.17100993543863297,
      "num_tokens": 42073573.0,
      "step": 19205
    },
    {
      "entropy": 5.640049123764038,
      "epoch": 1.906510520047638,
      "grad_norm": 0.984375,
      "learning_rate": 0.000464003269570485,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17358314245939255,
      "num_tokens": 42085217.0,
      "step": 19210
    },
    {
      "entropy": 5.679013967514038,
      "epoch": 1.9070067487098055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004639840433763783,
      "loss": 5.1173,
      "mean_token_accuracy": 0.17434265315532685,
      "num_tokens": 42095993.0,
      "step": 19215
    },
    {
      "entropy": 5.572829484939575,
      "epoch": 1.9075029773719732,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046396481249587454,
      "loss": 5.031,
      "mean_token_accuracy": 0.18091069161891937,
      "num_tokens": 42107049.0,
      "step": 19220
    },
    {
      "entropy": 5.589069175720215,
      "epoch": 1.9079992060341406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004639455769294506,
      "loss": 5.1027,
      "mean_token_accuracy": 0.17392949014902115,
      "num_tokens": 42117101.0,
      "step": 19225
    },
    {
      "entropy": 5.572338438034057,
      "epoch": 1.9084954346963081,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004639263366775835,
      "loss": 5.0868,
      "mean_token_accuracy": 0.17434962689876557,
      "num_tokens": 42128432.0,
      "step": 19230
    },
    {
      "entropy": 5.583099269866944,
      "epoch": 1.9089916633584756,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046390709174075036,
      "loss": 5.0742,
      "mean_token_accuracy": 0.1758283019065857,
      "num_tokens": 42140379.0,
      "step": 19235
    },
    {
      "entropy": 5.63200192451477,
      "epoch": 1.909487892020643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004638878421194284,
      "loss": 5.0869,
      "mean_token_accuracy": 0.17497413903474807,
      "num_tokens": 42151523.0,
      "step": 19240
    },
    {
      "entropy": 5.672858619689942,
      "epoch": 1.9099841206828105,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000463868587814095,
      "loss": 5.1683,
      "mean_token_accuracy": 0.17105559557676314,
      "num_tokens": 42162002.0,
      "step": 19245
    },
    {
      "entropy": 5.627292680740356,
      "epoch": 1.910480349344978,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638493288252276,
      "loss": 5.0268,
      "mean_token_accuracy": 0.18507721722126008,
      "num_tokens": 42172363.0,
      "step": 19250
    },
    {
      "entropy": 5.6143470287323,
      "epoch": 1.9109765780071457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046383006515330374,
      "loss": 5.1144,
      "mean_token_accuracy": 0.17517024874687195,
      "num_tokens": 42184162.0,
      "step": 19255
    },
    {
      "entropy": 5.651645088195801,
      "epoch": 1.9114728066693132,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004638107967988012,
      "loss": 5.1434,
      "mean_token_accuracy": 0.17755763679742814,
      "num_tokens": 42195024.0,
      "step": 19260
    },
    {
      "entropy": 5.664618062973022,
      "epoch": 1.9119690353314809,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046379152376219775,
      "loss": 5.1319,
      "mean_token_accuracy": 0.17420412600040436,
      "num_tokens": 42205463.0,
      "step": 19265
    },
    {
      "entropy": 5.505334186553955,
      "epoch": 1.9124652639936484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046377224604397127,
      "loss": 4.9903,
      "mean_token_accuracy": 0.1812935084104538,
      "num_tokens": 42215133.0,
      "step": 19270
    },
    {
      "entropy": 5.642130517959595,
      "epoch": 1.9129614926558158,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004637529636445998,
      "loss": 5.1722,
      "mean_token_accuracy": 0.18022237420082093,
      "num_tokens": 42226822.0,
      "step": 19275
    },
    {
      "entropy": 5.589657878875732,
      "epoch": 1.9134577213179833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004637336765645617,
      "loss": 5.083,
      "mean_token_accuracy": 0.1810320496559143,
      "num_tokens": 42237298.0,
      "step": 19280
    },
    {
      "entropy": 5.585572385787964,
      "epoch": 1.9139539499801508,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046371438480433504,
      "loss": 5.027,
      "mean_token_accuracy": 0.18103785961866378,
      "num_tokens": 42247242.0,
      "step": 19285
    },
    {
      "entropy": 5.584539842605591,
      "epoch": 1.9144501786423183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004636950883643983,
      "loss": 5.0412,
      "mean_token_accuracy": 0.18435366451740265,
      "num_tokens": 42259388.0,
      "step": 19290
    },
    {
      "entropy": 5.606125926971435,
      "epoch": 1.914946407304486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046367578724523,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18575460463762283,
      "num_tokens": 42270200.0,
      "step": 19295
    },
    {
      "entropy": 5.588152027130127,
      "epoch": 1.9154426359666534,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004636564814473087,
      "loss": 5.0564,
      "mean_token_accuracy": 0.1753865271806717,
      "num_tokens": 42280723.0,
      "step": 19300
    },
    {
      "entropy": 5.575057411193848,
      "epoch": 1.9159388646288211,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004636371709711131,
      "loss": 5.0914,
      "mean_token_accuracy": 0.17365653812885284,
      "num_tokens": 42291455.0,
      "step": 19305
    },
    {
      "entropy": 5.621865224838257,
      "epoch": 1.9164350932909886,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046361785581712224,
      "loss": 5.1259,
      "mean_token_accuracy": 0.17545668184757232,
      "num_tokens": 42302807.0,
      "step": 19310
    },
    {
      "entropy": 5.590449857711792,
      "epoch": 1.916931321953156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000463598535985815,
      "loss": 5.0528,
      "mean_token_accuracy": 0.17713681608438492,
      "num_tokens": 42313521.0,
      "step": 19315
    },
    {
      "entropy": 5.630803489685059,
      "epoch": 1.9174275506153236,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004635792114776704,
      "loss": 5.0742,
      "mean_token_accuracy": 0.17179132252931595,
      "num_tokens": 42324666.0,
      "step": 19320
    },
    {
      "entropy": 5.591847467422485,
      "epoch": 1.917923779277491,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004635598822931677,
      "loss": 5.1256,
      "mean_token_accuracy": 0.17555301636457443,
      "num_tokens": 42334904.0,
      "step": 19325
    },
    {
      "entropy": 5.5474395751953125,
      "epoch": 1.9184200079396585,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046354054843278624,
      "loss": 5.0474,
      "mean_token_accuracy": 0.17984424084424971,
      "num_tokens": 42347164.0,
      "step": 19330
    },
    {
      "entropy": 5.649763011932373,
      "epoch": 1.918916236601826,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046352120989700553,
      "loss": 5.1011,
      "mean_token_accuracy": 0.17640818208456038,
      "num_tokens": 42358668.0,
      "step": 19335
    },
    {
      "entropy": 5.629749011993408,
      "epoch": 1.9194124652639937,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000463501866686305,
      "loss": 5.0997,
      "mean_token_accuracy": 0.1790978118777275,
      "num_tokens": 42371229.0,
      "step": 19340
    },
    {
      "entropy": 5.564135789871216,
      "epoch": 1.9199086939261611,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046348251880116433,
      "loss": 5.0806,
      "mean_token_accuracy": 0.17744509279727935,
      "num_tokens": 42382383.0,
      "step": 19345
    },
    {
      "entropy": 5.626365756988525,
      "epoch": 1.9204049225883288,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046346316624206325,
      "loss": 5.2127,
      "mean_token_accuracy": 0.1684655800461769,
      "num_tokens": 42393634.0,
      "step": 19350
    },
    {
      "entropy": 5.641021299362182,
      "epoch": 1.9209011512504963,
      "grad_norm": 1.0,
      "learning_rate": 0.0004634438090094819,
      "loss": 5.1188,
      "mean_token_accuracy": 0.17201555222272874,
      "num_tokens": 42404699.0,
      "step": 19355
    },
    {
      "entropy": 5.596905708312988,
      "epoch": 1.9213973799126638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046342444710390006,
      "loss": 5.0585,
      "mean_token_accuracy": 0.18506666272878647,
      "num_tokens": 42415183.0,
      "step": 19360
    },
    {
      "entropy": 5.5722380638122555,
      "epoch": 1.9218936085748313,
      "grad_norm": 1.015625,
      "learning_rate": 0.000463405080525798,
      "loss": 4.9609,
      "mean_token_accuracy": 0.18472385108470918,
      "num_tokens": 42425162.0,
      "step": 19365
    },
    {
      "entropy": 5.55789999961853,
      "epoch": 1.9223898372369987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004633857092756559,
      "loss": 5.1561,
      "mean_token_accuracy": 0.17289014160633087,
      "num_tokens": 42438044.0,
      "step": 19370
    },
    {
      "entropy": 5.60443811416626,
      "epoch": 1.9228860658991662,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004633663333539542,
      "loss": 5.0731,
      "mean_token_accuracy": 0.1808578699827194,
      "num_tokens": 42449214.0,
      "step": 19375
    },
    {
      "entropy": 5.648301649093628,
      "epoch": 1.923382294561334,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046334695276117324,
      "loss": 5.1063,
      "mean_token_accuracy": 0.1759800747036934,
      "num_tokens": 42460645.0,
      "step": 19380
    },
    {
      "entropy": 5.602192974090576,
      "epoch": 1.9238785232235014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004633275674977937,
      "loss": 5.066,
      "mean_token_accuracy": 0.1803365468978882,
      "num_tokens": 42471449.0,
      "step": 19385
    },
    {
      "entropy": 5.620052289962769,
      "epoch": 1.9243747518856689,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004633081775642964,
      "loss": 5.087,
      "mean_token_accuracy": 0.17810595780611038,
      "num_tokens": 42482002.0,
      "step": 19390
    },
    {
      "entropy": 5.563658285140991,
      "epoch": 1.9248709805478366,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046328878296116196,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1862983748316765,
      "num_tokens": 42492068.0,
      "step": 19395
    },
    {
      "entropy": 5.620643138885498,
      "epoch": 1.925367209210004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046326938368887146,
      "loss": 5.1168,
      "mean_token_accuracy": 0.1776757836341858,
      "num_tokens": 42502766.0,
      "step": 19400
    },
    {
      "entropy": 5.57090539932251,
      "epoch": 1.9258634378721715,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046324997974790596,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1833618700504303,
      "num_tokens": 42513706.0,
      "step": 19405
    },
    {
      "entropy": 5.629587888717651,
      "epoch": 1.926359666534339,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004632305711387466,
      "loss": 5.0813,
      "mean_token_accuracy": 0.17103754729032516,
      "num_tokens": 42524164.0,
      "step": 19410
    },
    {
      "entropy": 5.6833210468292235,
      "epoch": 1.9268558951965065,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046321115786187466,
      "loss": 5.1192,
      "mean_token_accuracy": 0.17411004453897477,
      "num_tokens": 42535445.0,
      "step": 19415
    },
    {
      "entropy": 5.631332921981811,
      "epoch": 1.927352123858674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046319173991777153,
      "loss": 5.0248,
      "mean_token_accuracy": 0.1730904534459114,
      "num_tokens": 42545655.0,
      "step": 19420
    },
    {
      "entropy": 5.589561176300049,
      "epoch": 1.9278483525208416,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004631723173069188,
      "loss": 5.0479,
      "mean_token_accuracy": 0.1786030113697052,
      "num_tokens": 42556424.0,
      "step": 19425
    },
    {
      "entropy": 5.5430073738098145,
      "epoch": 1.928344581183009,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004631528900297981,
      "loss": 5.0248,
      "mean_token_accuracy": 0.17676668465137482,
      "num_tokens": 42566877.0,
      "step": 19430
    },
    {
      "entropy": 5.59789342880249,
      "epoch": 1.9288408098451768,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046313345808689113,
      "loss": 5.0748,
      "mean_token_accuracy": 0.18191470503807067,
      "num_tokens": 42577202.0,
      "step": 19435
    },
    {
      "entropy": 5.640151071548462,
      "epoch": 1.9293370385073443,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004631140214786798,
      "loss": 5.091,
      "mean_token_accuracy": 0.17668157815933228,
      "num_tokens": 42588605.0,
      "step": 19440
    },
    {
      "entropy": 5.611953687667847,
      "epoch": 1.9298332671695118,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046309458020564606,
      "loss": 5.072,
      "mean_token_accuracy": 0.18047238737344742,
      "num_tokens": 42599618.0,
      "step": 19445
    },
    {
      "entropy": 5.5772178173065186,
      "epoch": 1.9303294958316792,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046307513426827214,
      "loss": 5.1001,
      "mean_token_accuracy": 0.17374377846717834,
      "num_tokens": 42610999.0,
      "step": 19450
    },
    {
      "entropy": 5.5743958950042725,
      "epoch": 1.9308257244938467,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046305568366704006,
      "loss": 5.048,
      "mean_token_accuracy": 0.17892826795578004,
      "num_tokens": 42621167.0,
      "step": 19455
    },
    {
      "entropy": 5.603986978530884,
      "epoch": 1.9313219531560142,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004630362284024322,
      "loss": 5.086,
      "mean_token_accuracy": 0.17737133353948592,
      "num_tokens": 42631787.0,
      "step": 19460
    },
    {
      "entropy": 5.663410282135009,
      "epoch": 1.9318181818181817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046301676847493116,
      "loss": 5.277,
      "mean_token_accuracy": 0.162394717335701,
      "num_tokens": 42644002.0,
      "step": 19465
    },
    {
      "entropy": 5.588808870315551,
      "epoch": 1.9323144104803494,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046299730388501935,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17255590111017227,
      "num_tokens": 42655043.0,
      "step": 19470
    },
    {
      "entropy": 5.624253654479981,
      "epoch": 1.9328106391425168,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046297783463317957,
      "loss": 5.1107,
      "mean_token_accuracy": 0.17355137318372726,
      "num_tokens": 42668041.0,
      "step": 19475
    },
    {
      "entropy": 5.602960824966431,
      "epoch": 1.9333068678046845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004629583607198945,
      "loss": 5.0965,
      "mean_token_accuracy": 0.17564914077520372,
      "num_tokens": 42678521.0,
      "step": 19480
    },
    {
      "entropy": 5.557589292526245,
      "epoch": 1.933803096466852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046293888214564704,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17674362510442734,
      "num_tokens": 42690465.0,
      "step": 19485
    },
    {
      "entropy": 5.597377729415894,
      "epoch": 1.9342993251290195,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046291939891092036,
      "loss": 5.1022,
      "mean_token_accuracy": 0.1796914279460907,
      "num_tokens": 42700712.0,
      "step": 19490
    },
    {
      "entropy": 5.647071981430054,
      "epoch": 1.934795553791187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004628999110161975,
      "loss": 5.1507,
      "mean_token_accuracy": 0.17239754945039748,
      "num_tokens": 42712189.0,
      "step": 19495
    },
    {
      "entropy": 5.552151823043824,
      "epoch": 1.9352917824533544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046288041846196174,
      "loss": 5.023,
      "mean_token_accuracy": 0.18071708381175994,
      "num_tokens": 42723366.0,
      "step": 19500
    },
    {
      "entropy": 5.590687417984009,
      "epoch": 1.935788011115522,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046286092124869647,
      "loss": 5.0871,
      "mean_token_accuracy": 0.17858830690383912,
      "num_tokens": 42733766.0,
      "step": 19505
    },
    {
      "entropy": 5.616925048828125,
      "epoch": 1.9362842397776896,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046284141937688513,
      "loss": 5.1111,
      "mean_token_accuracy": 0.18055609613656998,
      "num_tokens": 42744642.0,
      "step": 19510
    },
    {
      "entropy": 5.596711015701294,
      "epoch": 1.936780468439857,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046282191284701123,
      "loss": 5.0336,
      "mean_token_accuracy": 0.18422254472970961,
      "num_tokens": 42755832.0,
      "step": 19515
    },
    {
      "entropy": 5.638865184783936,
      "epoch": 1.9372766971020248,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004628024016595587,
      "loss": 5.1566,
      "mean_token_accuracy": 0.17410333305597306,
      "num_tokens": 42767606.0,
      "step": 19520
    },
    {
      "entropy": 5.678178882598877,
      "epoch": 1.9377729257641922,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046278288581501136,
      "loss": 5.1367,
      "mean_token_accuracy": 0.17844693958759308,
      "num_tokens": 42778845.0,
      "step": 19525
    },
    {
      "entropy": 5.600379943847656,
      "epoch": 1.9382691544263597,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046276336531385305,
      "loss": 5.1041,
      "mean_token_accuracy": 0.17906154692173004,
      "num_tokens": 42789108.0,
      "step": 19530
    },
    {
      "entropy": 5.590298795700074,
      "epoch": 1.9387653830885272,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046274384015656785,
      "loss": 5.1307,
      "mean_token_accuracy": 0.16983912587165834,
      "num_tokens": 42800483.0,
      "step": 19535
    },
    {
      "entropy": 5.602957391738892,
      "epoch": 1.9392616117506947,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046272431034363997,
      "loss": 5.0615,
      "mean_token_accuracy": 0.18090126663446426,
      "num_tokens": 42812669.0,
      "step": 19540
    },
    {
      "entropy": 5.616747903823852,
      "epoch": 1.9397578404128621,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004627047758755538,
      "loss": 5.0771,
      "mean_token_accuracy": 0.17633070200681686,
      "num_tokens": 42822981.0,
      "step": 19545
    },
    {
      "entropy": 5.615851879119873,
      "epoch": 1.9402540690750296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004626852367527935,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17741243839263915,
      "num_tokens": 42833807.0,
      "step": 19550
    },
    {
      "entropy": 5.629355335235596,
      "epoch": 1.9407502977371973,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004626656929758438,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17016363888978958,
      "num_tokens": 42844757.0,
      "step": 19555
    },
    {
      "entropy": 5.652357196807861,
      "epoch": 1.9412465263993648,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004626461445451894,
      "loss": 5.1373,
      "mean_token_accuracy": 0.17508127093315123,
      "num_tokens": 42856346.0,
      "step": 19560
    },
    {
      "entropy": 5.651303148269653,
      "epoch": 1.9417427550615325,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004626265914613149,
      "loss": 5.0638,
      "mean_token_accuracy": 0.17475503385066987,
      "num_tokens": 42866665.0,
      "step": 19565
    },
    {
      "entropy": 5.696711587905884,
      "epoch": 1.9422389837237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004626070337247052,
      "loss": 5.2257,
      "mean_token_accuracy": 0.16772837936878204,
      "num_tokens": 42879819.0,
      "step": 19570
    },
    {
      "entropy": 5.620762872695923,
      "epoch": 1.9427352123858674,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004625874713358453,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17099535912275315,
      "num_tokens": 42891924.0,
      "step": 19575
    },
    {
      "entropy": 5.613914108276367,
      "epoch": 1.943231441048035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046256790429522035,
      "loss": 5.1339,
      "mean_token_accuracy": 0.17066845446825027,
      "num_tokens": 42902369.0,
      "step": 19580
    },
    {
      "entropy": 5.696543884277344,
      "epoch": 1.9437276697102024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004625483326033155,
      "loss": 5.1639,
      "mean_token_accuracy": 0.16686271280050277,
      "num_tokens": 42913466.0,
      "step": 19585
    },
    {
      "entropy": 5.611100959777832,
      "epoch": 1.9442238983723699,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004625287562606162,
      "loss": 5.0933,
      "mean_token_accuracy": 0.1768932357430458,
      "num_tokens": 42924033.0,
      "step": 19590
    },
    {
      "entropy": 5.615634489059448,
      "epoch": 1.9447201270345376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004625091752676078,
      "loss": 5.1392,
      "mean_token_accuracy": 0.1683555766940117,
      "num_tokens": 42935746.0,
      "step": 19595
    },
    {
      "entropy": 5.57725076675415,
      "epoch": 1.945216355696705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000462489589624776,
      "loss": 4.961,
      "mean_token_accuracy": 0.1888879954814911,
      "num_tokens": 42945932.0,
      "step": 19600
    },
    {
      "entropy": 5.591265869140625,
      "epoch": 1.9457125843588727,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004624699993326063,
      "loss": 5.0514,
      "mean_token_accuracy": 0.1769207864999771,
      "num_tokens": 42957326.0,
      "step": 19605
    },
    {
      "entropy": 5.606847333908081,
      "epoch": 1.9462088130210402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046245040439158456,
      "loss": 5.164,
      "mean_token_accuracy": 0.17451813220977783,
      "num_tokens": 42968548.0,
      "step": 19610
    },
    {
      "entropy": 5.571054458618164,
      "epoch": 1.9467050416832077,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046243080480219673,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17645591050386428,
      "num_tokens": 42980627.0,
      "step": 19615
    },
    {
      "entropy": 5.601631689071655,
      "epoch": 1.9472012703453752,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004624112005649288,
      "loss": 5.0673,
      "mean_token_accuracy": 0.17796150147914885,
      "num_tokens": 42992212.0,
      "step": 19620
    },
    {
      "entropy": 5.570251560211181,
      "epoch": 1.9476974990075426,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046239159168026694,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18596976846456528,
      "num_tokens": 43002833.0,
      "step": 19625
    },
    {
      "entropy": 5.599646377563476,
      "epoch": 1.94819372766971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004623719781486974,
      "loss": 5.1286,
      "mean_token_accuracy": 0.17447592467069625,
      "num_tokens": 43013861.0,
      "step": 19630
    },
    {
      "entropy": 5.6035669326782225,
      "epoch": 1.9486899563318776,
      "grad_norm": 1.0,
      "learning_rate": 0.00046235235997070664,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17575516998767854,
      "num_tokens": 43025295.0,
      "step": 19635
    },
    {
      "entropy": 5.600982141494751,
      "epoch": 1.9491861849940453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000462332737146781,
      "loss": 5.1107,
      "mean_token_accuracy": 0.17929001450538634,
      "num_tokens": 43035956.0,
      "step": 19640
    },
    {
      "entropy": 5.596673488616943,
      "epoch": 1.9496824136562128,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004623131096774071,
      "loss": 5.1183,
      "mean_token_accuracy": 0.17920776158571244,
      "num_tokens": 43047017.0,
      "step": 19645
    },
    {
      "entropy": 5.627244758605957,
      "epoch": 1.9501786423183805,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004622934775630717,
      "loss": 5.1143,
      "mean_token_accuracy": 0.17626702189445495,
      "num_tokens": 43057600.0,
      "step": 19650
    },
    {
      "entropy": 5.617480087280273,
      "epoch": 1.950674870980548,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046227384080426163,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1745771050453186,
      "num_tokens": 43068689.0,
      "step": 19655
    },
    {
      "entropy": 5.601876640319825,
      "epoch": 1.9511710996427154,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004622541994014639,
      "loss": 5.0775,
      "mean_token_accuracy": 0.17963305711746216,
      "num_tokens": 43079576.0,
      "step": 19660
    },
    {
      "entropy": 5.592150688171387,
      "epoch": 1.9516673283048829,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004622345533551655,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1695510521531105,
      "num_tokens": 43090522.0,
      "step": 19665
    },
    {
      "entropy": 5.529030895233154,
      "epoch": 1.9521635569670504,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004622149026658537,
      "loss": 5.0265,
      "mean_token_accuracy": 0.1762513756752014,
      "num_tokens": 43100043.0,
      "step": 19670
    },
    {
      "entropy": 5.610965394973755,
      "epoch": 1.9526597856292178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004621952473340156,
      "loss": 5.121,
      "mean_token_accuracy": 0.1721089541912079,
      "num_tokens": 43111023.0,
      "step": 19675
    },
    {
      "entropy": 5.689932060241699,
      "epoch": 1.9531560142913855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004621755873601388,
      "loss": 5.1975,
      "mean_token_accuracy": 0.17060365974903108,
      "num_tokens": 43123615.0,
      "step": 19680
    },
    {
      "entropy": 5.6261239528656,
      "epoch": 1.953652242953553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004621559227447108,
      "loss": 5.1388,
      "mean_token_accuracy": 0.1692158102989197,
      "num_tokens": 43134299.0,
      "step": 19685
    },
    {
      "entropy": 5.596459484100341,
      "epoch": 1.9541484716157205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004621362534882192,
      "loss": 4.9851,
      "mean_token_accuracy": 0.18677282482385635,
      "num_tokens": 43143857.0,
      "step": 19690
    },
    {
      "entropy": 5.609898662567138,
      "epoch": 1.9546447002778882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004621165795911517,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17270647138357162,
      "num_tokens": 43154571.0,
      "step": 19695
    },
    {
      "entropy": 5.650397634506225,
      "epoch": 1.9551409289400556,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046209690105399623,
      "loss": 5.1867,
      "mean_token_accuracy": 0.17146226465702058,
      "num_tokens": 43165812.0,
      "step": 19700
    },
    {
      "entropy": 5.657338237762451,
      "epoch": 1.9556371576022231,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00046207721787724085,
      "loss": 5.1281,
      "mean_token_accuracy": 0.1761981725692749,
      "num_tokens": 43178774.0,
      "step": 19705
    },
    {
      "entropy": 5.598449945449829,
      "epoch": 1.9561333862643906,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046205753006137347,
      "loss": 4.978,
      "mean_token_accuracy": 0.1818225845694542,
      "num_tokens": 43189795.0,
      "step": 19710
    },
    {
      "entropy": 5.55854640007019,
      "epoch": 1.956629614926558,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046203783760688245,
      "loss": 5.0447,
      "mean_token_accuracy": 0.17773295938968658,
      "num_tokens": 43201093.0,
      "step": 19715
    },
    {
      "entropy": 5.68701286315918,
      "epoch": 1.9571258435887255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004620181405142561,
      "loss": 5.1154,
      "mean_token_accuracy": 0.175559601187706,
      "num_tokens": 43211370.0,
      "step": 19720
    },
    {
      "entropy": 5.534236240386963,
      "epoch": 1.9576220722508932,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004619984387839829,
      "loss": 4.996,
      "mean_token_accuracy": 0.1756545051932335,
      "num_tokens": 43222066.0,
      "step": 19725
    },
    {
      "entropy": 5.58998293876648,
      "epoch": 1.9581183009130607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046197873241655133,
      "loss": 5.0679,
      "mean_token_accuracy": 0.17605245113372803,
      "num_tokens": 43233852.0,
      "step": 19730
    },
    {
      "entropy": 5.660322666168213,
      "epoch": 1.9586145295752284,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004619590214124501,
      "loss": 5.0783,
      "mean_token_accuracy": 0.18020822405815123,
      "num_tokens": 43244090.0,
      "step": 19735
    },
    {
      "entropy": 5.724868488311768,
      "epoch": 1.9591107582373959,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046193930577216797,
      "loss": 5.1574,
      "mean_token_accuracy": 0.1742385134100914,
      "num_tokens": 43255079.0,
      "step": 19740
    },
    {
      "entropy": 5.5379070281982425,
      "epoch": 1.9596069868995634,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004619195854961939,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18465482145547868,
      "num_tokens": 43266446.0,
      "step": 19745
    },
    {
      "entropy": 5.582618284225464,
      "epoch": 1.9601032155617308,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046189986058501687,
      "loss": 4.9886,
      "mean_token_accuracy": 0.1803553432226181,
      "num_tokens": 43276854.0,
      "step": 19750
    },
    {
      "entropy": 5.672615671157837,
      "epoch": 1.9605994442238983,
      "grad_norm": 1.0,
      "learning_rate": 0.000461880131039126,
      "loss": 5.155,
      "mean_token_accuracy": 0.17689116150140763,
      "num_tokens": 43288515.0,
      "step": 19755
    },
    {
      "entropy": 5.5723401546478275,
      "epoch": 1.9610956728860658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004618603968590107,
      "loss": 5.0572,
      "mean_token_accuracy": 0.17231282740831375,
      "num_tokens": 43299433.0,
      "step": 19760
    },
    {
      "entropy": 5.556921195983887,
      "epoch": 1.9615919015482333,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046184065804516007,
      "loss": 5.0484,
      "mean_token_accuracy": 0.18379314243793488,
      "num_tokens": 43309799.0,
      "step": 19765
    },
    {
      "entropy": 5.615430116653442,
      "epoch": 1.962088130210401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004618209145980638,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17313843816518784,
      "num_tokens": 43321925.0,
      "step": 19770
    },
    {
      "entropy": 5.599119043350219,
      "epoch": 1.9625843588725684,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004618011665182114,
      "loss": 5.0462,
      "mean_token_accuracy": 0.1783667117357254,
      "num_tokens": 43333336.0,
      "step": 19775
    },
    {
      "entropy": 5.597208499908447,
      "epoch": 1.9630805875347361,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004617814138060926,
      "loss": 5.0995,
      "mean_token_accuracy": 0.1806411102414131,
      "num_tokens": 43344343.0,
      "step": 19780
    },
    {
      "entropy": 5.637654781341553,
      "epoch": 1.9635768161969036,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004617616564621972,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17922369986772538,
      "num_tokens": 43355585.0,
      "step": 19785
    },
    {
      "entropy": 5.548816061019897,
      "epoch": 1.964073044859071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046174189448701516,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17938749194145204,
      "num_tokens": 43365938.0,
      "step": 19790
    },
    {
      "entropy": 5.584402847290039,
      "epoch": 1.9645692735212386,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004617221278810365,
      "loss": 5.0476,
      "mean_token_accuracy": 0.17726254165172578,
      "num_tokens": 43377114.0,
      "step": 19795
    },
    {
      "entropy": 5.622901725769043,
      "epoch": 1.965065502183406,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046170235664475147,
      "loss": 5.0213,
      "mean_token_accuracy": 0.18012016117572785,
      "num_tokens": 43389869.0,
      "step": 19800
    },
    {
      "entropy": 5.630146932601929,
      "epoch": 1.9655617308455735,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046168258077865026,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17471566945314407,
      "num_tokens": 43400989.0,
      "step": 19805
    },
    {
      "entropy": 5.629272985458374,
      "epoch": 1.9660579595077412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004616628002832233,
      "loss": 5.1738,
      "mean_token_accuracy": 0.17662205398082734,
      "num_tokens": 43411688.0,
      "step": 19810
    },
    {
      "entropy": 5.691619634628296,
      "epoch": 1.9665541881699087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046164301515896124,
      "loss": 5.1515,
      "mean_token_accuracy": 0.17957154512405396,
      "num_tokens": 43423362.0,
      "step": 19815
    },
    {
      "entropy": 5.62332592010498,
      "epoch": 1.9670504168320764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004616232254063544,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17791622132062912,
      "num_tokens": 43433651.0,
      "step": 19820
    },
    {
      "entropy": 5.536783266067505,
      "epoch": 1.9675466454942439,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046160343102589384,
      "loss": 4.9319,
      "mean_token_accuracy": 0.19616518914699554,
      "num_tokens": 43444048.0,
      "step": 19825
    },
    {
      "entropy": 5.610284852981567,
      "epoch": 1.9680428741564113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004615836320180702,
      "loss": 5.086,
      "mean_token_accuracy": 0.1736750990152359,
      "num_tokens": 43453913.0,
      "step": 19830
    },
    {
      "entropy": 5.628982925415039,
      "epoch": 1.9685391028185788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046156382838337455,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1730814054608345,
      "num_tokens": 43466800.0,
      "step": 19835
    },
    {
      "entropy": 5.660432910919189,
      "epoch": 1.9690353314807463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046154402012229807,
      "loss": 5.115,
      "mean_token_accuracy": 0.1824721321463585,
      "num_tokens": 43477435.0,
      "step": 19840
    },
    {
      "entropy": 5.605185604095459,
      "epoch": 1.9695315601429138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046152420723533175,
      "loss": 5.0711,
      "mean_token_accuracy": 0.1799985408782959,
      "num_tokens": 43486772.0,
      "step": 19845
    },
    {
      "entropy": 5.612453126907349,
      "epoch": 1.9700277888050812,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046150438972296706,
      "loss": 5.0293,
      "mean_token_accuracy": 0.17787941098213195,
      "num_tokens": 43497240.0,
      "step": 19850
    },
    {
      "entropy": 5.655386018753052,
      "epoch": 1.970524017467249,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046148456758569536,
      "loss": 5.0995,
      "mean_token_accuracy": 0.1759825348854065,
      "num_tokens": 43507282.0,
      "step": 19855
    },
    {
      "entropy": 5.707820558547974,
      "epoch": 1.9710202461294164,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046146474082400823,
      "loss": 5.12,
      "mean_token_accuracy": 0.17353656589984895,
      "num_tokens": 43517782.0,
      "step": 19860
    },
    {
      "entropy": 5.601941776275635,
      "epoch": 1.971516474791584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046144490943839735,
      "loss": 5.1643,
      "mean_token_accuracy": 0.17355486154556274,
      "num_tokens": 43528139.0,
      "step": 19865
    },
    {
      "entropy": 5.581112957000732,
      "epoch": 1.9720127034537516,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004614250734293544,
      "loss": 5.0655,
      "mean_token_accuracy": 0.18231103271245958,
      "num_tokens": 43540395.0,
      "step": 19870
    },
    {
      "entropy": 5.567456579208374,
      "epoch": 1.972508932115919,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046140523279737146,
      "loss": 5.0236,
      "mean_token_accuracy": 0.17798611521720886,
      "num_tokens": 43551521.0,
      "step": 19875
    },
    {
      "entropy": 5.610658121109009,
      "epoch": 1.9730051607780865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004613853875429403,
      "loss": 5.0439,
      "mean_token_accuracy": 0.18555197417736052,
      "num_tokens": 43562468.0,
      "step": 19880
    },
    {
      "entropy": 5.629255533218384,
      "epoch": 1.973501389440254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004613655376665532,
      "loss": 5.0445,
      "mean_token_accuracy": 0.17663053423166275,
      "num_tokens": 43573144.0,
      "step": 19885
    },
    {
      "entropy": 5.6279380321502686,
      "epoch": 1.9739976181024215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004613456831687024,
      "loss": 5.1016,
      "mean_token_accuracy": 0.17258149534463882,
      "num_tokens": 43584482.0,
      "step": 19890
    },
    {
      "entropy": 5.652121829986572,
      "epoch": 1.9744938467645892,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046132582404988004,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1725937157869339,
      "num_tokens": 43596000.0,
      "step": 19895
    },
    {
      "entropy": 5.675826787948608,
      "epoch": 1.9749900754267566,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046130596031057885,
      "loss": 5.2419,
      "mean_token_accuracy": 0.16807832270860673,
      "num_tokens": 43606493.0,
      "step": 19900
    },
    {
      "entropy": 5.63284010887146,
      "epoch": 1.9754863040889243,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046128609195129135,
      "loss": 5.0955,
      "mean_token_accuracy": 0.17714064717292785,
      "num_tokens": 43618178.0,
      "step": 19905
    },
    {
      "entropy": 5.633785676956177,
      "epoch": 1.9759825327510918,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004612662189725101,
      "loss": 5.059,
      "mean_token_accuracy": 0.1780673235654831,
      "num_tokens": 43628978.0,
      "step": 19910
    },
    {
      "entropy": 5.630771589279175,
      "epoch": 1.9764787614132593,
      "grad_norm": 0.953125,
      "learning_rate": 0.000461246341374728,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17840541452169417,
      "num_tokens": 43639283.0,
      "step": 19915
    },
    {
      "entropy": 5.641792917251587,
      "epoch": 1.9769749900754268,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046122645915843796,
      "loss": 5.1383,
      "mean_token_accuracy": 0.17022849768400192,
      "num_tokens": 43650226.0,
      "step": 19920
    },
    {
      "entropy": 5.611616420745849,
      "epoch": 1.9774712187375942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046120657232413305,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1748488336801529,
      "num_tokens": 43660289.0,
      "step": 19925
    },
    {
      "entropy": 5.595848417282104,
      "epoch": 1.9779674473997617,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004611866808723064,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1711698591709137,
      "num_tokens": 43671716.0,
      "step": 19930
    },
    {
      "entropy": 5.544299364089966,
      "epoch": 1.9784636760619292,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046116678480345117,
      "loss": 5.0247,
      "mean_token_accuracy": 0.1808907687664032,
      "num_tokens": 43681906.0,
      "step": 19935
    },
    {
      "entropy": 5.562044620513916,
      "epoch": 1.9789599047240969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046114688411806083,
      "loss": 5.0497,
      "mean_token_accuracy": 0.17248550057411194,
      "num_tokens": 43693765.0,
      "step": 19940
    },
    {
      "entropy": 5.6163774013519285,
      "epoch": 1.9794561333862644,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046112697881662895,
      "loss": 5.0405,
      "mean_token_accuracy": 0.18224265575408935,
      "num_tokens": 43703986.0,
      "step": 19945
    },
    {
      "entropy": 5.562637519836426,
      "epoch": 1.979952362048432,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046110706889964905,
      "loss": 5.0566,
      "mean_token_accuracy": 0.17694421112537384,
      "num_tokens": 43715361.0,
      "step": 19950
    },
    {
      "entropy": 5.661002111434937,
      "epoch": 1.9804485907105995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004610871543676149,
      "loss": 5.0976,
      "mean_token_accuracy": 0.18404200673103333,
      "num_tokens": 43725674.0,
      "step": 19955
    },
    {
      "entropy": 5.620563316345215,
      "epoch": 1.980944819372767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046106723522102024,
      "loss": 5.0813,
      "mean_token_accuracy": 0.17459651529788972,
      "num_tokens": 43736004.0,
      "step": 19960
    },
    {
      "entropy": 5.629887437820434,
      "epoch": 1.9814410480349345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046104731146035904,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17238666564226152,
      "num_tokens": 43747216.0,
      "step": 19965
    },
    {
      "entropy": 5.577605867385865,
      "epoch": 1.981937276697102,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004610273830861255,
      "loss": 5.0709,
      "mean_token_accuracy": 0.17797560840845109,
      "num_tokens": 43758865.0,
      "step": 19970
    },
    {
      "entropy": 5.571738195419312,
      "epoch": 1.9824335053592694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004610074500988137,
      "loss": 5.001,
      "mean_token_accuracy": 0.1801302269101143,
      "num_tokens": 43769428.0,
      "step": 19975
    },
    {
      "entropy": 5.636602401733398,
      "epoch": 1.9829297340214371,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004609875124989179,
      "loss": 5.1016,
      "mean_token_accuracy": 0.17357080280780793,
      "num_tokens": 43779334.0,
      "step": 19980
    },
    {
      "entropy": 5.6906229019165036,
      "epoch": 1.9834259626836046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004609675702869326,
      "loss": 5.1528,
      "mean_token_accuracy": 0.17552360147237778,
      "num_tokens": 43791308.0,
      "step": 19985
    },
    {
      "entropy": 5.640877819061279,
      "epoch": 1.983922191345772,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004609476234633523,
      "loss": 5.0937,
      "mean_token_accuracy": 0.1740354359149933,
      "num_tokens": 43801871.0,
      "step": 19990
    },
    {
      "entropy": 5.54983229637146,
      "epoch": 1.9844184200079398,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046092767202867156,
      "loss": 4.9407,
      "mean_token_accuracy": 0.183018459379673,
      "num_tokens": 43812897.0,
      "step": 19995
    },
    {
      "entropy": 5.573059797286987,
      "epoch": 1.9849146486701073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046090771598338525,
      "loss": 5.1507,
      "mean_token_accuracy": 0.1769825905561447,
      "num_tokens": 43823104.0,
      "step": 20000
    },
    {
      "entropy": 5.599181985855102,
      "epoch": 1.9854108773322747,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004608877553279881,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18237437456846237,
      "num_tokens": 43833889.0,
      "step": 20005
    },
    {
      "entropy": 5.718679428100586,
      "epoch": 1.9859071059944422,
      "grad_norm": 1.0,
      "learning_rate": 0.0004608677900629753,
      "loss": 5.1752,
      "mean_token_accuracy": 0.17263693511486053,
      "num_tokens": 43845407.0,
      "step": 20010
    },
    {
      "entropy": 5.655726766586303,
      "epoch": 1.9864033346566097,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046084782018884177,
      "loss": 5.1035,
      "mean_token_accuracy": 0.18041329681873322,
      "num_tokens": 43857304.0,
      "step": 20015
    },
    {
      "entropy": 5.6530725955963135,
      "epoch": 1.9868995633187772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046082784570608266,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1787797197699547,
      "num_tokens": 43867993.0,
      "step": 20020
    },
    {
      "entropy": 5.602781105041504,
      "epoch": 1.9873957919809448,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046080786661519354,
      "loss": 5.0428,
      "mean_token_accuracy": 0.18363250494003297,
      "num_tokens": 43877724.0,
      "step": 20025
    },
    {
      "entropy": 5.5647139072418215,
      "epoch": 1.9878920206431123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004607878829166696,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18626073747873306,
      "num_tokens": 43887595.0,
      "step": 20030
    },
    {
      "entropy": 5.678546714782715,
      "epoch": 1.98838824930528,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046076789461100653,
      "loss": 5.1172,
      "mean_token_accuracy": 0.17824324816465378,
      "num_tokens": 43898218.0,
      "step": 20035
    },
    {
      "entropy": 5.6460648536682125,
      "epoch": 1.9888844779674475,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004607479016987001,
      "loss": 5.0404,
      "mean_token_accuracy": 0.18273699134588242,
      "num_tokens": 43908154.0,
      "step": 20040
    },
    {
      "entropy": 5.536100149154663,
      "epoch": 1.989380706629615,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004607279041802458,
      "loss": 5.0471,
      "mean_token_accuracy": 0.17320722937583924,
      "num_tokens": 43918840.0,
      "step": 20045
    },
    {
      "entropy": 5.609504985809326,
      "epoch": 1.9898769352917824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046070790205613967,
      "loss": 5.0665,
      "mean_token_accuracy": 0.1730301186442375,
      "num_tokens": 43928836.0,
      "step": 20050
    },
    {
      "entropy": 5.686213541030884,
      "epoch": 1.99037316395395,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046068789532687784,
      "loss": 5.1466,
      "mean_token_accuracy": 0.17253938615322112,
      "num_tokens": 43941213.0,
      "step": 20055
    },
    {
      "entropy": 5.641171979904175,
      "epoch": 1.9908693926161174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004606678839929563,
      "loss": 5.1436,
      "mean_token_accuracy": 0.17477365285158158,
      "num_tokens": 43953254.0,
      "step": 20060
    },
    {
      "entropy": 5.602974700927734,
      "epoch": 1.9913656212782849,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004606478680548713,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17936568856239318,
      "num_tokens": 43963512.0,
      "step": 20065
    },
    {
      "entropy": 5.640328407287598,
      "epoch": 1.9918618499404526,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046062784751311917,
      "loss": 5.1721,
      "mean_token_accuracy": 0.16895440965890884,
      "num_tokens": 43973735.0,
      "step": 20070
    },
    {
      "entropy": 5.671923828125,
      "epoch": 1.99235807860262,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046060782236819647,
      "loss": 5.1495,
      "mean_token_accuracy": 0.16893809139728547,
      "num_tokens": 43985619.0,
      "step": 20075
    },
    {
      "entropy": 5.637044668197632,
      "epoch": 1.9928543072647877,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004605877926205996,
      "loss": 5.0736,
      "mean_token_accuracy": 0.17144529670476913,
      "num_tokens": 43996524.0,
      "step": 20080
    },
    {
      "entropy": 5.553088092803955,
      "epoch": 1.9933505359269552,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004605677582708256,
      "loss": 5.094,
      "mean_token_accuracy": 0.17524840533733368,
      "num_tokens": 44007649.0,
      "step": 20085
    },
    {
      "entropy": 5.6164374351501465,
      "epoch": 1.9938467645891227,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004605477193193709,
      "loss": 5.0792,
      "mean_token_accuracy": 0.1798313722014427,
      "num_tokens": 44018049.0,
      "step": 20090
    },
    {
      "entropy": 5.671327304840088,
      "epoch": 1.9943429932512902,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004605276757667326,
      "loss": 5.0994,
      "mean_token_accuracy": 0.1663486912846565,
      "num_tokens": 44027600.0,
      "step": 20095
    },
    {
      "entropy": 5.590271759033203,
      "epoch": 1.9948392219134576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004605076276134078,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17954279929399491,
      "num_tokens": 44038244.0,
      "step": 20100
    },
    {
      "entropy": 5.5420889377594,
      "epoch": 1.9953354505756251,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046048757485989345,
      "loss": 4.9981,
      "mean_token_accuracy": 0.18235064297914505,
      "num_tokens": 44049565.0,
      "step": 20105
    },
    {
      "entropy": 5.633370208740234,
      "epoch": 1.9958316792377928,
      "grad_norm": 1.109375,
      "learning_rate": 0.000460467517506687,
      "loss": 5.1292,
      "mean_token_accuracy": 0.17617220878601075,
      "num_tokens": 44060955.0,
      "step": 20110
    },
    {
      "entropy": 5.540948343276978,
      "epoch": 1.9963279078999603,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004604474555542858,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18059320598840714,
      "num_tokens": 44071115.0,
      "step": 20115
    },
    {
      "entropy": 5.665598678588867,
      "epoch": 1.996824136562128,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004604273890031872,
      "loss": 5.1704,
      "mean_token_accuracy": 0.17605490535497664,
      "num_tokens": 44083114.0,
      "step": 20120
    },
    {
      "entropy": 5.6139599800109865,
      "epoch": 1.9973203652242955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000460407317853889,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18079412132501602,
      "num_tokens": 44093790.0,
      "step": 20125
    },
    {
      "entropy": 5.607929754257202,
      "epoch": 1.997816593886463,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004603872421068888,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17734451740980148,
      "num_tokens": 44104688.0,
      "step": 20130
    },
    {
      "entropy": 5.640383625030518,
      "epoch": 1.9983128225486304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004603671617626845,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17531801015138626,
      "num_tokens": 44114487.0,
      "step": 20135
    },
    {
      "entropy": 5.586035919189453,
      "epoch": 1.9988090512107979,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046034707682177393,
      "loss": 5.0597,
      "mean_token_accuracy": 0.17528702318668365,
      "num_tokens": 44126070.0,
      "step": 20140
    },
    {
      "entropy": 5.5745361804962155,
      "epoch": 1.9993052798729654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004603269872846554,
      "loss": 5.0675,
      "mean_token_accuracy": 0.17879535853862763,
      "num_tokens": 44137538.0,
      "step": 20145
    },
    {
      "entropy": 5.644154977798462,
      "epoch": 1.9998015085351328,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046030689315182676,
      "loss": 5.0652,
      "mean_token_accuracy": 0.18087309896945952,
      "num_tokens": 44148322.0,
      "step": 20150
    },
    {
      "entropy": 5.716168308258057,
      "epoch": 2.0002977371973003,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004602867944237866,
      "loss": 5.1919,
      "mean_token_accuracy": 0.17407033294439317,
      "num_tokens": 44159085.0,
      "step": 20155
    },
    {
      "entropy": 5.629267787933349,
      "epoch": 2.0007939658594682,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046026669110103315,
      "loss": 5.0512,
      "mean_token_accuracy": 0.17556682080030442,
      "num_tokens": 44170269.0,
      "step": 20160
    },
    {
      "entropy": 5.629829216003418,
      "epoch": 2.0012901945216357,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046024658318406493,
      "loss": 5.0105,
      "mean_token_accuracy": 0.17339108884334564,
      "num_tokens": 44181395.0,
      "step": 20165
    },
    {
      "entropy": 5.618002843856812,
      "epoch": 2.001786423183803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004602264706733807,
      "loss": 5.0466,
      "mean_token_accuracy": 0.18079222738742828,
      "num_tokens": 44192655.0,
      "step": 20170
    },
    {
      "entropy": 5.59625473022461,
      "epoch": 2.0022826518459707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004602063535694791,
      "loss": 4.9523,
      "mean_token_accuracy": 0.1862561970949173,
      "num_tokens": 44202628.0,
      "step": 20175
    },
    {
      "entropy": 5.6821709156036375,
      "epoch": 2.002778880508138,
      "grad_norm": 1.140625,
      "learning_rate": 0.000460186231872859,
      "loss": 5.1335,
      "mean_token_accuracy": 0.17253224700689315,
      "num_tokens": 44213281.0,
      "step": 20180
    },
    {
      "entropy": 5.576064348220825,
      "epoch": 2.0032751091703056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046016610558401936,
      "loss": 4.9818,
      "mean_token_accuracy": 0.17936881631612778,
      "num_tokens": 44223919.0,
      "step": 20185
    },
    {
      "entropy": 5.564826345443725,
      "epoch": 2.003771337832473,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046014597470345935,
      "loss": 5.0628,
      "mean_token_accuracy": 0.18734797686338425,
      "num_tokens": 44233637.0,
      "step": 20190
    },
    {
      "entropy": 5.638918781280518,
      "epoch": 2.0042675664946406,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004601258392316781,
      "loss": 5.0368,
      "mean_token_accuracy": 0.18087815791368483,
      "num_tokens": 44243119.0,
      "step": 20195
    },
    {
      "entropy": 5.694678163528442,
      "epoch": 2.0047637951568085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046010569916917496,
      "loss": 5.1052,
      "mean_token_accuracy": 0.1798362448811531,
      "num_tokens": 44253827.0,
      "step": 20200
    },
    {
      "entropy": 5.574078226089478,
      "epoch": 2.005260023818976,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004600855545164493,
      "loss": 4.883,
      "mean_token_accuracy": 0.20115431249141694,
      "num_tokens": 44264721.0,
      "step": 20205
    },
    {
      "entropy": 5.545940971374511,
      "epoch": 2.0057562524811434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004600654052740008,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18107434809207917,
      "num_tokens": 44275510.0,
      "step": 20210
    },
    {
      "entropy": 5.701707983016968,
      "epoch": 2.006252481143311,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004600452514423289,
      "loss": 5.1466,
      "mean_token_accuracy": 0.16957348883152007,
      "num_tokens": 44288391.0,
      "step": 20215
    },
    {
      "entropy": 5.665515899658203,
      "epoch": 2.0067487098054784,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004600250930219336,
      "loss": 5.096,
      "mean_token_accuracy": 0.1768262892961502,
      "num_tokens": 44300062.0,
      "step": 20220
    },
    {
      "entropy": 5.617221260070801,
      "epoch": 2.007244938467646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046000493001331465,
      "loss": 5.005,
      "mean_token_accuracy": 0.18162814527750015,
      "num_tokens": 44310341.0,
      "step": 20225
    },
    {
      "entropy": 5.599055051803589,
      "epoch": 2.0077411671298133,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045998476241697214,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18597615957260133,
      "num_tokens": 44320618.0,
      "step": 20230
    },
    {
      "entropy": 5.576207494735717,
      "epoch": 2.008237395791981,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004599645902334061,
      "loss": 5.0198,
      "mean_token_accuracy": 0.19035600274801254,
      "num_tokens": 44331627.0,
      "step": 20235
    },
    {
      "entropy": 5.604830551147461,
      "epoch": 2.0087336244541483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045994441346311677,
      "loss": 5.0131,
      "mean_token_accuracy": 0.1838614448904991,
      "num_tokens": 44342880.0,
      "step": 20240
    },
    {
      "entropy": 5.642623996734619,
      "epoch": 2.009229853116316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004599242321066045,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17637291550636292,
      "num_tokens": 44354150.0,
      "step": 20245
    },
    {
      "entropy": 5.604832029342651,
      "epoch": 2.0097260817784837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045990404616436976,
      "loss": 5.0236,
      "mean_token_accuracy": 0.18253791630268096,
      "num_tokens": 44364849.0,
      "step": 20250
    },
    {
      "entropy": 5.596246910095215,
      "epoch": 2.010222310440651,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004598838556369131,
      "loss": 5.0671,
      "mean_token_accuracy": 0.17052896916866303,
      "num_tokens": 44374954.0,
      "step": 20255
    },
    {
      "entropy": 5.669431066513061,
      "epoch": 2.0107185391028186,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045986366052473525,
      "loss": 5.0909,
      "mean_token_accuracy": 0.17615959346294402,
      "num_tokens": 44387653.0,
      "step": 20260
    },
    {
      "entropy": 5.683819198608399,
      "epoch": 2.011214767764986,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004598434608283369,
      "loss": 5.0711,
      "mean_token_accuracy": 0.1834455117583275,
      "num_tokens": 44398673.0,
      "step": 20265
    },
    {
      "entropy": 5.598685646057129,
      "epoch": 2.0117109964271536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004598232565482191,
      "loss": 5.0554,
      "mean_token_accuracy": 0.175421079993248,
      "num_tokens": 44409764.0,
      "step": 20270
    },
    {
      "entropy": 5.636963891983032,
      "epoch": 2.012207225089321,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045980304768488274,
      "loss": 5.1158,
      "mean_token_accuracy": 0.1739790305495262,
      "num_tokens": 44421998.0,
      "step": 20275
    },
    {
      "entropy": 5.6159971237182615,
      "epoch": 2.0127034537514885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045978283423882906,
      "loss": 5.0194,
      "mean_token_accuracy": 0.18298039585351944,
      "num_tokens": 44434236.0,
      "step": 20280
    },
    {
      "entropy": 5.654560470581055,
      "epoch": 2.0131996824136564,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045976261621055915,
      "loss": 4.9711,
      "mean_token_accuracy": 0.18500525057315825,
      "num_tokens": 44444171.0,
      "step": 20285
    },
    {
      "entropy": 5.604700565338135,
      "epoch": 2.013695911075824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045974239360057457,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18429137617349625,
      "num_tokens": 44454924.0,
      "step": 20290
    },
    {
      "entropy": 5.574910068511963,
      "epoch": 2.0141921397379914,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045972216640937676,
      "loss": 5.0594,
      "mean_token_accuracy": 0.18228431195020675,
      "num_tokens": 44467072.0,
      "step": 20295
    },
    {
      "entropy": 5.6258447647094725,
      "epoch": 2.014688368400159,
      "grad_norm": 1.125,
      "learning_rate": 0.0004597019346374672,
      "loss": 5.07,
      "mean_token_accuracy": 0.1702168822288513,
      "num_tokens": 44477735.0,
      "step": 20300
    },
    {
      "entropy": 5.600796031951904,
      "epoch": 2.0151845970623263,
      "grad_norm": 1.125,
      "learning_rate": 0.0004596816982853477,
      "loss": 5.069,
      "mean_token_accuracy": 0.17305066287517548,
      "num_tokens": 44489028.0,
      "step": 20305
    },
    {
      "entropy": 5.693414878845215,
      "epoch": 2.015680825724494,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045966145735352,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17785266041755676,
      "num_tokens": 44499313.0,
      "step": 20310
    },
    {
      "entropy": 5.629087638854981,
      "epoch": 2.0161770543866613,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045964121184248604,
      "loss": 4.9659,
      "mean_token_accuracy": 0.1837633430957794,
      "num_tokens": 44510331.0,
      "step": 20315
    },
    {
      "entropy": 5.553265857696533,
      "epoch": 2.0166732830488288,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004596209617527479,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17705406099557877,
      "num_tokens": 44521249.0,
      "step": 20320
    },
    {
      "entropy": 5.576003837585449,
      "epoch": 2.0171695117109962,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004596007070848078,
      "loss": 5.1035,
      "mean_token_accuracy": 0.17309673130512238,
      "num_tokens": 44531117.0,
      "step": 20325
    },
    {
      "entropy": 5.631967163085937,
      "epoch": 2.017665740373164,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045958044783916795,
      "loss": 4.9826,
      "mean_token_accuracy": 0.18253767341375352,
      "num_tokens": 44543807.0,
      "step": 20330
    },
    {
      "entropy": 5.631833839416504,
      "epoch": 2.0181619690353316,
      "grad_norm": 1.125,
      "learning_rate": 0.0004595601840163306,
      "loss": 5.0442,
      "mean_token_accuracy": 0.17350894063711167,
      "num_tokens": 44553247.0,
      "step": 20335
    },
    {
      "entropy": 5.604042625427246,
      "epoch": 2.018658197697499,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045953991561679855,
      "loss": 5.0187,
      "mean_token_accuracy": 0.17845599949359894,
      "num_tokens": 44564566.0,
      "step": 20340
    },
    {
      "entropy": 5.602689170837403,
      "epoch": 2.0191544263596666,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004595196426410742,
      "loss": 5.0605,
      "mean_token_accuracy": 0.1808190479874611,
      "num_tokens": 44575968.0,
      "step": 20345
    },
    {
      "entropy": 5.707440567016602,
      "epoch": 2.019650655021834,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045949936508966015,
      "loss": 5.1408,
      "mean_token_accuracy": 0.17520863860845565,
      "num_tokens": 44587797.0,
      "step": 20350
    },
    {
      "entropy": 5.607734155654907,
      "epoch": 2.0201468836840015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004594790829630596,
      "loss": 4.972,
      "mean_token_accuracy": 0.18462805449962616,
      "num_tokens": 44598741.0,
      "step": 20355
    },
    {
      "entropy": 5.665106868743896,
      "epoch": 2.020643112346169,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004594587962617752,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17568286806344985,
      "num_tokens": 44609813.0,
      "step": 20360
    },
    {
      "entropy": 5.599295949935913,
      "epoch": 2.0211393410083365,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004594385049863102,
      "loss": 4.9729,
      "mean_token_accuracy": 0.18127224445343018,
      "num_tokens": 44619996.0,
      "step": 20365
    },
    {
      "entropy": 5.553371715545654,
      "epoch": 2.021635569670504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045941820913716766,
      "loss": 4.9808,
      "mean_token_accuracy": 0.1822769582271576,
      "num_tokens": 44629596.0,
      "step": 20370
    },
    {
      "entropy": 5.5720545768737795,
      "epoch": 2.022131798332672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045939790871485093,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18138013184070587,
      "num_tokens": 44639204.0,
      "step": 20375
    },
    {
      "entropy": 5.609366416931152,
      "epoch": 2.0226280269948393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045937760371986334,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17522464394569398,
      "num_tokens": 44650140.0,
      "step": 20380
    },
    {
      "entropy": 5.555718278884887,
      "epoch": 2.023124255657007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045935729415270857,
      "loss": 4.9481,
      "mean_token_accuracy": 0.18697573542594909,
      "num_tokens": 44659962.0,
      "step": 20385
    },
    {
      "entropy": 5.577268981933594,
      "epoch": 2.0236204843191743,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004593369800138901,
      "loss": 4.9171,
      "mean_token_accuracy": 0.18299557119607926,
      "num_tokens": 44669608.0,
      "step": 20390
    },
    {
      "entropy": 5.5346098899841305,
      "epoch": 2.0241167129813418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004593166613039118,
      "loss": 4.9669,
      "mean_token_accuracy": 0.1861206367611885,
      "num_tokens": 44679044.0,
      "step": 20395
    },
    {
      "entropy": 5.658390522003174,
      "epoch": 2.0246129416435092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045929633802327743,
      "loss": 5.1534,
      "mean_token_accuracy": 0.18184249997138976,
      "num_tokens": 44691186.0,
      "step": 20400
    },
    {
      "entropy": 5.63915114402771,
      "epoch": 2.0251091703056767,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000459276010172491,
      "loss": 5.0305,
      "mean_token_accuracy": 0.17548326253890992,
      "num_tokens": 44702827.0,
      "step": 20405
    },
    {
      "entropy": 5.6714911460876465,
      "epoch": 2.025605398967844,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045925567775205656,
      "loss": 5.0298,
      "mean_token_accuracy": 0.17729886174201964,
      "num_tokens": 44713903.0,
      "step": 20410
    },
    {
      "entropy": 5.627742338180542,
      "epoch": 2.026101627630012,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045923534076247846,
      "loss": 5.0621,
      "mean_token_accuracy": 0.18023286014795303,
      "num_tokens": 44724527.0,
      "step": 20415
    },
    {
      "entropy": 5.577289342880249,
      "epoch": 2.0265978562921796,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045921499920426083,
      "loss": 5.069,
      "mean_token_accuracy": 0.1770881950855255,
      "num_tokens": 44734092.0,
      "step": 20420
    },
    {
      "entropy": 5.584189748764038,
      "epoch": 2.027094084954347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045919465307790817,
      "loss": 4.9625,
      "mean_token_accuracy": 0.19051628857851027,
      "num_tokens": 44745046.0,
      "step": 20425
    },
    {
      "entropy": 5.646173667907715,
      "epoch": 2.0275903136165145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045917430238392505,
      "loss": 4.9418,
      "mean_token_accuracy": 0.19334718436002732,
      "num_tokens": 44754603.0,
      "step": 20430
    },
    {
      "entropy": 5.59629864692688,
      "epoch": 2.028086542278682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000459153947122816,
      "loss": 5.0323,
      "mean_token_accuracy": 0.182754223048687,
      "num_tokens": 44765263.0,
      "step": 20435
    },
    {
      "entropy": 5.65412163734436,
      "epoch": 2.0285827709408495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004591335872950859,
      "loss": 5.0999,
      "mean_token_accuracy": 0.16839880496263504,
      "num_tokens": 44775861.0,
      "step": 20440
    },
    {
      "entropy": 5.704663801193237,
      "epoch": 2.029078999603017,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004591132229012397,
      "loss": 5.1089,
      "mean_token_accuracy": 0.17653557956218718,
      "num_tokens": 44785132.0,
      "step": 20445
    },
    {
      "entropy": 5.647472667694092,
      "epoch": 2.0295752282651844,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045909285394178217,
      "loss": 5.0339,
      "mean_token_accuracy": 0.18023869395256042,
      "num_tokens": 44795634.0,
      "step": 20450
    },
    {
      "entropy": 5.604823112487793,
      "epoch": 2.030071456927352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004590724804172187,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1797090321779251,
      "num_tokens": 44806315.0,
      "step": 20455
    },
    {
      "entropy": 5.557442712783813,
      "epoch": 2.03056768558952,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004590521023280543,
      "loss": 4.9969,
      "mean_token_accuracy": 0.18357927948236466,
      "num_tokens": 44816670.0,
      "step": 20460
    },
    {
      "entropy": 5.642659997940063,
      "epoch": 2.0310639142516873,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045903171967479425,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17648125141859056,
      "num_tokens": 44827292.0,
      "step": 20465
    },
    {
      "entropy": 5.619122266769409,
      "epoch": 2.031560142913855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045901133245794426,
      "loss": 5.048,
      "mean_token_accuracy": 0.1775427892804146,
      "num_tokens": 44838501.0,
      "step": 20470
    },
    {
      "entropy": 5.63469386100769,
      "epoch": 2.0320563715760223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004589909406780096,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1767989844083786,
      "num_tokens": 44849416.0,
      "step": 20475
    },
    {
      "entropy": 5.632760810852051,
      "epoch": 2.0325526002381897,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045897054433549617,
      "loss": 5.1353,
      "mean_token_accuracy": 0.17158064246177673,
      "num_tokens": 44860365.0,
      "step": 20480
    },
    {
      "entropy": 5.59315938949585,
      "epoch": 2.033048828900357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045895014343090963,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1813046246767044,
      "num_tokens": 44870415.0,
      "step": 20485
    },
    {
      "entropy": 5.591394090652466,
      "epoch": 2.0335450575625247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045892973796475594,
      "loss": 4.9633,
      "mean_token_accuracy": 0.1848526880145073,
      "num_tokens": 44881409.0,
      "step": 20490
    },
    {
      "entropy": 5.675364589691162,
      "epoch": 2.034041286224692,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000458909327937541,
      "loss": 5.1182,
      "mean_token_accuracy": 0.17476049959659576,
      "num_tokens": 44893580.0,
      "step": 20495
    },
    {
      "entropy": 5.654539632797241,
      "epoch": 2.03453751488686,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045888891334977105,
      "loss": 5.0053,
      "mean_token_accuracy": 0.18154372423887252,
      "num_tokens": 44905154.0,
      "step": 20500
    },
    {
      "entropy": 5.63834319114685,
      "epoch": 2.0350337435490276,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004588684942019524,
      "loss": 4.9964,
      "mean_token_accuracy": 0.1821243941783905,
      "num_tokens": 44916796.0,
      "step": 20505
    },
    {
      "entropy": 5.62962532043457,
      "epoch": 2.035529972211195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045884807049459116,
      "loss": 5.1185,
      "mean_token_accuracy": 0.17625821381807327,
      "num_tokens": 44929065.0,
      "step": 20510
    },
    {
      "entropy": 5.547454500198365,
      "epoch": 2.0360262008733625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000458827642228194,
      "loss": 4.9413,
      "mean_token_accuracy": 0.18885983526706696,
      "num_tokens": 44938948.0,
      "step": 20515
    },
    {
      "entropy": 5.652587127685547,
      "epoch": 2.03652242953553,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045880720940326744,
      "loss": 5.0695,
      "mean_token_accuracy": 0.1773996338248253,
      "num_tokens": 44949536.0,
      "step": 20520
    },
    {
      "entropy": 5.684528923034668,
      "epoch": 2.0370186581976975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045878677202031804,
      "loss": 5.0314,
      "mean_token_accuracy": 0.18271245211362838,
      "num_tokens": 44960667.0,
      "step": 20525
    },
    {
      "entropy": 5.617602157592773,
      "epoch": 2.037514886859865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004587663300798529,
      "loss": 5.0168,
      "mean_token_accuracy": 0.1845632791519165,
      "num_tokens": 44972289.0,
      "step": 20530
    },
    {
      "entropy": 5.673601341247559,
      "epoch": 2.0380111155220324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004587458835823786,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1795769065618515,
      "num_tokens": 44983028.0,
      "step": 20535
    },
    {
      "entropy": 5.554367208480835,
      "epoch": 2.0385073441842,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045872543252840236,
      "loss": 4.9942,
      "mean_token_accuracy": 0.17637073248624802,
      "num_tokens": 44994935.0,
      "step": 20540
    },
    {
      "entropy": 5.599671792984009,
      "epoch": 2.039003572846368,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045870497691843134,
      "loss": 5.0182,
      "mean_token_accuracy": 0.17384425550699234,
      "num_tokens": 45005403.0,
      "step": 20545
    },
    {
      "entropy": 5.566119718551636,
      "epoch": 2.0394998015085353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004586845167529727,
      "loss": 4.978,
      "mean_token_accuracy": 0.18006177395582199,
      "num_tokens": 45016960.0,
      "step": 20550
    },
    {
      "entropy": 5.651681041717529,
      "epoch": 2.0399960301707027,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045866405203253384,
      "loss": 5.0568,
      "mean_token_accuracy": 0.1824905604124069,
      "num_tokens": 45027820.0,
      "step": 20555
    },
    {
      "entropy": 5.623155784606934,
      "epoch": 2.04049225883287,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004586435827576223,
      "loss": 5.0607,
      "mean_token_accuracy": 0.17433976382017136,
      "num_tokens": 45039442.0,
      "step": 20560
    },
    {
      "entropy": 5.5958226203918455,
      "epoch": 2.0409884874950377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004586231089287456,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17911491245031358,
      "num_tokens": 45050565.0,
      "step": 20565
    },
    {
      "entropy": 5.573035955429077,
      "epoch": 2.041484716157205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004586026305464115,
      "loss": 4.9705,
      "mean_token_accuracy": 0.18767289221286773,
      "num_tokens": 45060992.0,
      "step": 20570
    },
    {
      "entropy": 5.609989023208618,
      "epoch": 2.0419809448193726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045858214761112777,
      "loss": 5.0351,
      "mean_token_accuracy": 0.18002334088087082,
      "num_tokens": 45071595.0,
      "step": 20575
    },
    {
      "entropy": 5.580452108383179,
      "epoch": 2.04247717348154,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045856166012340235,
      "loss": 5.0063,
      "mean_token_accuracy": 0.17683750689029692,
      "num_tokens": 45082021.0,
      "step": 20580
    },
    {
      "entropy": 5.5752197265625,
      "epoch": 2.0429734021437076,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004585411680837433,
      "loss": 4.9706,
      "mean_token_accuracy": 0.17959051579236984,
      "num_tokens": 45092064.0,
      "step": 20585
    },
    {
      "entropy": 5.592089462280273,
      "epoch": 2.0434696308058755,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045852067149265875,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1802539512515068,
      "num_tokens": 45104389.0,
      "step": 20590
    },
    {
      "entropy": 5.63570237159729,
      "epoch": 2.043965859468043,
      "grad_norm": 1.09375,
      "learning_rate": 0.000458500170350657,
      "loss": 5.0775,
      "mean_token_accuracy": 0.17549042105674745,
      "num_tokens": 45113627.0,
      "step": 20595
    },
    {
      "entropy": 5.628873348236084,
      "epoch": 2.0444620881302105,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045847966465824654,
      "loss": 5.0488,
      "mean_token_accuracy": 0.1891790509223938,
      "num_tokens": 45124835.0,
      "step": 20600
    },
    {
      "entropy": 5.482684755325318,
      "epoch": 2.044958316792378,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004584591544159356,
      "loss": 4.8681,
      "mean_token_accuracy": 0.1929523915052414,
      "num_tokens": 45134966.0,
      "step": 20605
    },
    {
      "entropy": 5.560111141204834,
      "epoch": 2.0454545454545454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004584386396242331,
      "loss": 5.0229,
      "mean_token_accuracy": 0.1788581430912018,
      "num_tokens": 45145626.0,
      "step": 20610
    },
    {
      "entropy": 5.608994007110596,
      "epoch": 2.045950774116713,
      "grad_norm": 1.0,
      "learning_rate": 0.00045841812028364756,
      "loss": 5.0173,
      "mean_token_accuracy": 0.18345005065202713,
      "num_tokens": 45157054.0,
      "step": 20615
    },
    {
      "entropy": 5.593127298355102,
      "epoch": 2.0464470027788804,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045839759639468774,
      "loss": 5.0474,
      "mean_token_accuracy": 0.17866908013820648,
      "num_tokens": 45168275.0,
      "step": 20620
    },
    {
      "entropy": 5.6532848358154295,
      "epoch": 2.046943231441048,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004583770679578628,
      "loss": 5.0128,
      "mean_token_accuracy": 0.19033531844615936,
      "num_tokens": 45178346.0,
      "step": 20625
    },
    {
      "entropy": 5.565115833282471,
      "epoch": 2.0474394601032158,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004583565349736818,
      "loss": 4.9755,
      "mean_token_accuracy": 0.1890598252415657,
      "num_tokens": 45189033.0,
      "step": 20630
    },
    {
      "entropy": 5.513562297821045,
      "epoch": 2.0479356887653832,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004583359974426538,
      "loss": 4.9017,
      "mean_token_accuracy": 0.1974255070090294,
      "num_tokens": 45200470.0,
      "step": 20635
    },
    {
      "entropy": 5.578318929672241,
      "epoch": 2.0484319174275507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045831545536528806,
      "loss": 4.9829,
      "mean_token_accuracy": 0.1798306241631508,
      "num_tokens": 45211550.0,
      "step": 20640
    },
    {
      "entropy": 5.5868409156799315,
      "epoch": 2.048928146089718,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045829490874209404,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1874051421880722,
      "num_tokens": 45222230.0,
      "step": 20645
    },
    {
      "entropy": 5.566236972808838,
      "epoch": 2.0494243747518857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045827435757358136,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18879947811365128,
      "num_tokens": 45233397.0,
      "step": 20650
    },
    {
      "entropy": 5.589874219894409,
      "epoch": 2.049920603414053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004582538018602594,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1746046543121338,
      "num_tokens": 45244617.0,
      "step": 20655
    },
    {
      "entropy": 5.5974701881408695,
      "epoch": 2.0504168320762206,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004582332416026381,
      "loss": 4.8914,
      "mean_token_accuracy": 0.1978221669793129,
      "num_tokens": 45254377.0,
      "step": 20660
    },
    {
      "entropy": 5.640861225128174,
      "epoch": 2.050913060738388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045821267680122725,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18290061354637147,
      "num_tokens": 45264945.0,
      "step": 20665
    },
    {
      "entropy": 5.567238187789917,
      "epoch": 2.0514092894005556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004581921074565368,
      "loss": 5.0396,
      "mean_token_accuracy": 0.1811390519142151,
      "num_tokens": 45275596.0,
      "step": 20670
    },
    {
      "entropy": 5.598221731185913,
      "epoch": 2.0519055180627235,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004581715335690768,
      "loss": 5.0226,
      "mean_token_accuracy": 0.18420050293207169,
      "num_tokens": 45285775.0,
      "step": 20675
    },
    {
      "entropy": 5.662706089019776,
      "epoch": 2.052401746724891,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045815095513935747,
      "loss": 5.0889,
      "mean_token_accuracy": 0.1795325383543968,
      "num_tokens": 45297760.0,
      "step": 20680
    },
    {
      "entropy": 5.705981683731079,
      "epoch": 2.0528979753870584,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004581303721678891,
      "loss": 5.0712,
      "mean_token_accuracy": 0.1760007619857788,
      "num_tokens": 45308903.0,
      "step": 20685
    },
    {
      "entropy": 5.583258628845215,
      "epoch": 2.053394204049226,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004581097846551821,
      "loss": 4.9929,
      "mean_token_accuracy": 0.18427660316228867,
      "num_tokens": 45321004.0,
      "step": 20690
    },
    {
      "entropy": 5.590235042572021,
      "epoch": 2.0538904327113934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000458089192601747,
      "loss": 4.9896,
      "mean_token_accuracy": 0.1794118255376816,
      "num_tokens": 45331353.0,
      "step": 20695
    },
    {
      "entropy": 5.632206296920776,
      "epoch": 2.054386661373561,
      "grad_norm": 1.125,
      "learning_rate": 0.00045806859600809455,
      "loss": 5.067,
      "mean_token_accuracy": 0.17694442868232726,
      "num_tokens": 45342086.0,
      "step": 20700
    },
    {
      "entropy": 5.651246881484985,
      "epoch": 2.0548828900357283,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004580479948747353,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1781681925058365,
      "num_tokens": 45353018.0,
      "step": 20705
    },
    {
      "entropy": 5.587949514389038,
      "epoch": 2.055379118697896,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045802738920218015,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18832326531410218,
      "num_tokens": 45364261.0,
      "step": 20710
    },
    {
      "entropy": 5.584009218215942,
      "epoch": 2.0558753473600637,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004580067789909403,
      "loss": 5.0101,
      "mean_token_accuracy": 0.18190052211284638,
      "num_tokens": 45374888.0,
      "step": 20715
    },
    {
      "entropy": 5.625341987609863,
      "epoch": 2.056371576022231,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045798616424152654,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17823589593172073,
      "num_tokens": 45385780.0,
      "step": 20720
    },
    {
      "entropy": 5.6558130264282225,
      "epoch": 2.0568678046843987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045796554495445023,
      "loss": 5.035,
      "mean_token_accuracy": 0.1812782734632492,
      "num_tokens": 45396554.0,
      "step": 20725
    },
    {
      "entropy": 5.61595664024353,
      "epoch": 2.057364033346566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004579449211302228,
      "loss": 5.0102,
      "mean_token_accuracy": 0.1859840989112854,
      "num_tokens": 45407258.0,
      "step": 20730
    },
    {
      "entropy": 5.654914617538452,
      "epoch": 2.0578602620087336,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004579242927693554,
      "loss": 5.0891,
      "mean_token_accuracy": 0.177579765021801,
      "num_tokens": 45417836.0,
      "step": 20735
    },
    {
      "entropy": 5.559084177017212,
      "epoch": 2.058356490670901,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004579036598723598,
      "loss": 4.932,
      "mean_token_accuracy": 0.18479044139385223,
      "num_tokens": 45428826.0,
      "step": 20740
    },
    {
      "entropy": 5.571610927581787,
      "epoch": 2.0588527193330686,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045788302243974745,
      "loss": 5.0116,
      "mean_token_accuracy": 0.18209773898124695,
      "num_tokens": 45441023.0,
      "step": 20745
    },
    {
      "entropy": 5.621649646759034,
      "epoch": 2.059348947995236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004578623804720302,
      "loss": 5.0496,
      "mean_token_accuracy": 0.17821156829595566,
      "num_tokens": 45451529.0,
      "step": 20750
    },
    {
      "entropy": 5.705277681350708,
      "epoch": 2.0598451766574035,
      "grad_norm": 1.125,
      "learning_rate": 0.0004578417339697201,
      "loss": 5.052,
      "mean_token_accuracy": 0.17604229599237442,
      "num_tokens": 45461127.0,
      "step": 20755
    },
    {
      "entropy": 5.673625040054321,
      "epoch": 2.0603414053195714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045782108293332884,
      "loss": 5.0768,
      "mean_token_accuracy": 0.18350817859172822,
      "num_tokens": 45471649.0,
      "step": 20760
    },
    {
      "entropy": 5.565325927734375,
      "epoch": 2.060837633981739,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004578004273633687,
      "loss": 4.9405,
      "mean_token_accuracy": 0.1790801018476486,
      "num_tokens": 45482267.0,
      "step": 20765
    },
    {
      "entropy": 5.641433572769165,
      "epoch": 2.0613338626439064,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045777976726035194,
      "loss": 5.1256,
      "mean_token_accuracy": 0.1725946396589279,
      "num_tokens": 45493935.0,
      "step": 20770
    },
    {
      "entropy": 5.595594596862793,
      "epoch": 2.061830091306074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004577591026247907,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17816049754619598,
      "num_tokens": 45505347.0,
      "step": 20775
    },
    {
      "entropy": 5.566076612472534,
      "epoch": 2.0623263199682413,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004577384334571976,
      "loss": 4.9707,
      "mean_token_accuracy": 0.18364004641771317,
      "num_tokens": 45515168.0,
      "step": 20780
    },
    {
      "entropy": 5.659224987030029,
      "epoch": 2.062822548630409,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004577177597580851,
      "loss": 5.1247,
      "mean_token_accuracy": 0.17064681202173232,
      "num_tokens": 45525792.0,
      "step": 20785
    },
    {
      "entropy": 5.557657861709595,
      "epoch": 2.0633187772925763,
      "grad_norm": 1.125,
      "learning_rate": 0.00045769708152796593,
      "loss": 4.9428,
      "mean_token_accuracy": 0.1843093603849411,
      "num_tokens": 45536704.0,
      "step": 20790
    },
    {
      "entropy": 5.651213550567627,
      "epoch": 2.0638150059547438,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045767639876735276,
      "loss": 5.0155,
      "mean_token_accuracy": 0.18837999403476716,
      "num_tokens": 45548400.0,
      "step": 20795
    },
    {
      "entropy": 5.619901132583618,
      "epoch": 2.0643112346169117,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004576557114767585,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17858463078737258,
      "num_tokens": 45559417.0,
      "step": 20800
    },
    {
      "entropy": 5.679437828063965,
      "epoch": 2.064807463279079,
      "grad_norm": 1.0,
      "learning_rate": 0.00045763501965669627,
      "loss": 5.1375,
      "mean_token_accuracy": 0.16579360365867615,
      "num_tokens": 45571040.0,
      "step": 20805
    },
    {
      "entropy": 5.60668158531189,
      "epoch": 2.0653036919412466,
      "grad_norm": 1.046875,
      "learning_rate": 0.000457614323307679,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1817839503288269,
      "num_tokens": 45581180.0,
      "step": 20810
    },
    {
      "entropy": 5.59911961555481,
      "epoch": 2.065799920603414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045759362243022,
      "loss": 5.0298,
      "mean_token_accuracy": 0.18241226822137832,
      "num_tokens": 45592671.0,
      "step": 20815
    },
    {
      "entropy": 5.543575954437256,
      "epoch": 2.0662961492655816,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004575729170248327,
      "loss": 4.9732,
      "mean_token_accuracy": 0.18788623660802842,
      "num_tokens": 45605002.0,
      "step": 20820
    },
    {
      "entropy": 5.6147651195526125,
      "epoch": 2.066792377927749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004575522070920305,
      "loss": 5.0439,
      "mean_token_accuracy": 0.18116793781518936,
      "num_tokens": 45615238.0,
      "step": 20825
    },
    {
      "entropy": 5.622368621826172,
      "epoch": 2.0672886065899165,
      "grad_norm": 1.125,
      "learning_rate": 0.00045753149263232687,
      "loss": 5.1013,
      "mean_token_accuracy": 0.17714103758335115,
      "num_tokens": 45625272.0,
      "step": 20830
    },
    {
      "entropy": 5.673278665542602,
      "epoch": 2.067784835252084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004575107736462355,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18955665528774263,
      "num_tokens": 45638591.0,
      "step": 20835
    },
    {
      "entropy": 5.629866361618042,
      "epoch": 2.0682810639142515,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004574900501342703,
      "loss": 5.0509,
      "mean_token_accuracy": 0.18077087998390198,
      "num_tokens": 45648802.0,
      "step": 20840
    },
    {
      "entropy": 5.622836589813232,
      "epoch": 2.0687772925764194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004574693220969451,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1765446186065674,
      "num_tokens": 45660297.0,
      "step": 20845
    },
    {
      "entropy": 5.555187463760376,
      "epoch": 2.069273521238587,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004574485895347738,
      "loss": 4.9792,
      "mean_token_accuracy": 0.1776080310344696,
      "num_tokens": 45671782.0,
      "step": 20850
    },
    {
      "entropy": 5.615747880935669,
      "epoch": 2.0697697499007544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045742785244827066,
      "loss": 5.0747,
      "mean_token_accuracy": 0.18001775443553925,
      "num_tokens": 45682018.0,
      "step": 20855
    },
    {
      "entropy": 5.613861322402954,
      "epoch": 2.070265978562922,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004574071108379499,
      "loss": 5.0592,
      "mean_token_accuracy": 0.1787185087800026,
      "num_tokens": 45692746.0,
      "step": 20860
    },
    {
      "entropy": 5.626727056503296,
      "epoch": 2.0707622072250893,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004573863647043259,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17908562570810319,
      "num_tokens": 45702452.0,
      "step": 20865
    },
    {
      "entropy": 5.636052656173706,
      "epoch": 2.0712584358872568,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004573656140479129,
      "loss": 4.9968,
      "mean_token_accuracy": 0.17588534206151962,
      "num_tokens": 45712931.0,
      "step": 20870
    },
    {
      "entropy": 5.584581661224365,
      "epoch": 2.0717546645494243,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004573448588692258,
      "loss": 4.9801,
      "mean_token_accuracy": 0.19226926416158677,
      "num_tokens": 45723677.0,
      "step": 20875
    },
    {
      "entropy": 5.565772438049317,
      "epoch": 2.0722508932115917,
      "grad_norm": 1.015625,
      "learning_rate": 0.000457324099168779,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18418121635913848,
      "num_tokens": 45734835.0,
      "step": 20880
    },
    {
      "entropy": 5.609440803527832,
      "epoch": 2.0727471218737596,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004573033349470874,
      "loss": 5.0078,
      "mean_token_accuracy": 0.1826714336872101,
      "num_tokens": 45746900.0,
      "step": 20885
    },
    {
      "entropy": 5.688608264923095,
      "epoch": 2.073243350535927,
      "grad_norm": 1.125,
      "learning_rate": 0.0004572825662046661,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17885234355926513,
      "num_tokens": 45757565.0,
      "step": 20890
    },
    {
      "entropy": 5.681149053573608,
      "epoch": 2.0737395791980946,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004572617929420298,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1663748636841774,
      "num_tokens": 45770468.0,
      "step": 20895
    },
    {
      "entropy": 5.616214466094971,
      "epoch": 2.074235807860262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045724101515969375,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18375278562307357,
      "num_tokens": 45781245.0,
      "step": 20900
    },
    {
      "entropy": 5.619049882888794,
      "epoch": 2.0747320365224295,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045722023285817327,
      "loss": 5.052,
      "mean_token_accuracy": 0.1858016535639763,
      "num_tokens": 45793190.0,
      "step": 20905
    },
    {
      "entropy": 5.561623001098633,
      "epoch": 2.075228265184597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004571994460379837,
      "loss": 4.9444,
      "mean_token_accuracy": 0.1837600737810135,
      "num_tokens": 45804268.0,
      "step": 20910
    },
    {
      "entropy": 5.562750625610351,
      "epoch": 2.0757244938467645,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045717865469964045,
      "loss": 5.0684,
      "mean_token_accuracy": 0.18217221200466155,
      "num_tokens": 45816107.0,
      "step": 20915
    },
    {
      "entropy": 5.6864824295043945,
      "epoch": 2.076220722508932,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045715785884365914,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1745594173669815,
      "num_tokens": 45827919.0,
      "step": 20920
    },
    {
      "entropy": 5.673999881744384,
      "epoch": 2.0767169511710994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045713705847055537,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1731465280056,
      "num_tokens": 45839343.0,
      "step": 20925
    },
    {
      "entropy": 5.652263212203979,
      "epoch": 2.0772131798332674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004571162535808451,
      "loss": 5.0759,
      "mean_token_accuracy": 0.17674693018198012,
      "num_tokens": 45849903.0,
      "step": 20930
    },
    {
      "entropy": 5.588974237442017,
      "epoch": 2.077709408495435,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004570954441750441,
      "loss": 5.0284,
      "mean_token_accuracy": 0.18744733035564423,
      "num_tokens": 45860115.0,
      "step": 20935
    },
    {
      "entropy": 5.619974565505982,
      "epoch": 2.0782056371576023,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004570746302536685,
      "loss": 5.101,
      "mean_token_accuracy": 0.17608762681484222,
      "num_tokens": 45870772.0,
      "step": 20940
    },
    {
      "entropy": 5.656980371475219,
      "epoch": 2.07870186581977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045705381181723444,
      "loss": 5.0432,
      "mean_token_accuracy": 0.1759821057319641,
      "num_tokens": 45882498.0,
      "step": 20945
    },
    {
      "entropy": 5.643893814086914,
      "epoch": 2.0791980944819373,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045703298886625804,
      "loss": 5.0455,
      "mean_token_accuracy": 0.17923768758773803,
      "num_tokens": 45893829.0,
      "step": 20950
    },
    {
      "entropy": 5.584176826477051,
      "epoch": 2.0796943231441047,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004570121614012558,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18707724064588546,
      "num_tokens": 45902941.0,
      "step": 20955
    },
    {
      "entropy": 5.580630731582642,
      "epoch": 2.080190551806272,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045699132942274425,
      "loss": 5.0003,
      "mean_token_accuracy": 0.1806248351931572,
      "num_tokens": 45915024.0,
      "step": 20960
    },
    {
      "entropy": 5.664514350891113,
      "epoch": 2.0806867804684397,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045697049293123984,
      "loss": 5.065,
      "mean_token_accuracy": 0.1753416657447815,
      "num_tokens": 45925019.0,
      "step": 20965
    },
    {
      "entropy": 5.591357183456421,
      "epoch": 2.0811830091306076,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004569496519272593,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18411210030317307,
      "num_tokens": 45936824.0,
      "step": 20970
    },
    {
      "entropy": 5.545995235443115,
      "epoch": 2.081679237792775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045692880641131945,
      "loss": 4.9602,
      "mean_token_accuracy": 0.1802825465798378,
      "num_tokens": 45947513.0,
      "step": 20975
    },
    {
      "entropy": 5.535391712188721,
      "epoch": 2.0821754664549426,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004569079563839372,
      "loss": 4.936,
      "mean_token_accuracy": 0.19287471175193788,
      "num_tokens": 45957190.0,
      "step": 20980
    },
    {
      "entropy": 5.611828470230103,
      "epoch": 2.08267169511711,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045688710184562965,
      "loss": 5.0673,
      "mean_token_accuracy": 0.1885467380285263,
      "num_tokens": 45968461.0,
      "step": 20985
    },
    {
      "entropy": 5.5437909126281735,
      "epoch": 2.0831679237792775,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045686624279691387,
      "loss": 4.9818,
      "mean_token_accuracy": 0.1844064250588417,
      "num_tokens": 45978782.0,
      "step": 20990
    },
    {
      "entropy": 5.587511730194092,
      "epoch": 2.083664152441445,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004568453792383072,
      "loss": 4.9908,
      "mean_token_accuracy": 0.18779300302267074,
      "num_tokens": 45988042.0,
      "step": 20995
    },
    {
      "entropy": 5.6331392288208,
      "epoch": 2.0841603811036125,
      "grad_norm": 1.015625,
      "learning_rate": 0.000456824511170327,
      "loss": 5.119,
      "mean_token_accuracy": 0.17321881353855134,
      "num_tokens": 46000250.0,
      "step": 21000
    },
    {
      "epoch": 2.0841603811036125,
      "eval_entropy": 5.407602022764206,
      "eval_loss": 5.177760124206543,
      "eval_mean_token_accuracy": 0.1783353925577944,
      "eval_num_tokens": 46000250.0,
      "eval_runtime": 20.3767,
      "eval_samples_per_second": 1652.328,
      "eval_steps_per_second": 206.559,
      "step": 21000
    },
    {
      "entropy": 5.660132026672363,
      "epoch": 2.08465660976578,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004568036385934906,
      "loss": 5.0792,
      "mean_token_accuracy": 0.17312037348747253,
      "num_tokens": 46011707.0,
      "step": 21005
    },
    {
      "entropy": 5.652016592025757,
      "epoch": 2.0851528384279474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045678276150831586,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18170465528964996,
      "num_tokens": 46022673.0,
      "step": 21010
    },
    {
      "entropy": 5.679728078842163,
      "epoch": 2.0856490670901153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004567618799153202,
      "loss": 5.1559,
      "mean_token_accuracy": 0.16912010908126832,
      "num_tokens": 46033073.0,
      "step": 21015
    },
    {
      "entropy": 5.625650024414062,
      "epoch": 2.086145295752283,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045674099381502167,
      "loss": 5.0437,
      "mean_token_accuracy": 0.18636615872383117,
      "num_tokens": 46043395.0,
      "step": 21020
    },
    {
      "entropy": 5.608670854568482,
      "epoch": 2.0866415244144503,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004567201032079381,
      "loss": 5.1188,
      "mean_token_accuracy": 0.1729451984167099,
      "num_tokens": 46054808.0,
      "step": 21025
    },
    {
      "entropy": 5.6416669368743895,
      "epoch": 2.0871377530766178,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045669920809458754,
      "loss": 5.0893,
      "mean_token_accuracy": 0.1772936388850212,
      "num_tokens": 46065726.0,
      "step": 21030
    },
    {
      "entropy": 5.6168341636657715,
      "epoch": 2.0876339817387852,
      "grad_norm": 1.0,
      "learning_rate": 0.0004566783084754881,
      "loss": 5.0553,
      "mean_token_accuracy": 0.18020621836185455,
      "num_tokens": 46076470.0,
      "step": 21035
    },
    {
      "entropy": 5.578331565856933,
      "epoch": 2.0881302104009527,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045665740435115815,
      "loss": 5.0108,
      "mean_token_accuracy": 0.1826112151145935,
      "num_tokens": 46087348.0,
      "step": 21040
    },
    {
      "entropy": 5.627524375915527,
      "epoch": 2.08862643906312,
      "grad_norm": 0.984375,
      "learning_rate": 0.000456636495722116,
      "loss": 5.0562,
      "mean_token_accuracy": 0.1768150046467781,
      "num_tokens": 46099225.0,
      "step": 21045
    },
    {
      "entropy": 5.678006744384765,
      "epoch": 2.0891226677252877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004566155825888801,
      "loss": 5.1285,
      "mean_token_accuracy": 0.17698326110839843,
      "num_tokens": 46110212.0,
      "step": 21050
    },
    {
      "entropy": 5.635766077041626,
      "epoch": 2.089618896387455,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004565946649519691,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18256428837776184,
      "num_tokens": 46120197.0,
      "step": 21055
    },
    {
      "entropy": 5.6216089725494385,
      "epoch": 2.090115125049623,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045657374281190177,
      "loss": 5.0317,
      "mean_token_accuracy": 0.18054956942796707,
      "num_tokens": 46133043.0,
      "step": 21060
    },
    {
      "entropy": 5.647916603088379,
      "epoch": 2.0906113537117905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004565528161691969,
      "loss": 5.027,
      "mean_token_accuracy": 0.178712897002697,
      "num_tokens": 46144594.0,
      "step": 21065
    },
    {
      "entropy": 5.629165744781494,
      "epoch": 2.091107582373958,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004565318850243733,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17905355244874954,
      "num_tokens": 46155400.0,
      "step": 21070
    },
    {
      "entropy": 5.565516567230224,
      "epoch": 2.0916038110361255,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004565109493779502,
      "loss": 5.0245,
      "mean_token_accuracy": 0.17760595679283142,
      "num_tokens": 46166354.0,
      "step": 21075
    },
    {
      "entropy": 5.664071846008301,
      "epoch": 2.092100039698293,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045649000923044664,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1838023379445076,
      "num_tokens": 46177472.0,
      "step": 21080
    },
    {
      "entropy": 5.633112478256225,
      "epoch": 2.0925962683604604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045646906458238196,
      "loss": 5.109,
      "mean_token_accuracy": 0.17648178339004517,
      "num_tokens": 46189625.0,
      "step": 21085
    },
    {
      "entropy": 5.660123538970947,
      "epoch": 2.093092497022628,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004564481154342754,
      "loss": 5.0465,
      "mean_token_accuracy": 0.17643104046583175,
      "num_tokens": 46201639.0,
      "step": 21090
    },
    {
      "entropy": 5.6108380317687985,
      "epoch": 2.0935887256847954,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004564271617866467,
      "loss": 4.997,
      "mean_token_accuracy": 0.1813057392835617,
      "num_tokens": 46212248.0,
      "step": 21095
    },
    {
      "entropy": 5.6433710098266605,
      "epoch": 2.0940849543469633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045640620364001524,
      "loss": 5.0601,
      "mean_token_accuracy": 0.17673568725585936,
      "num_tokens": 46223171.0,
      "step": 21100
    },
    {
      "entropy": 5.6608987808227536,
      "epoch": 2.0945811830091308,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045638524099490086,
      "loss": 5.045,
      "mean_token_accuracy": 0.1788820281624794,
      "num_tokens": 46233782.0,
      "step": 21105
    },
    {
      "entropy": 5.626981639862061,
      "epoch": 2.0950774116712982,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045636427385182334,
      "loss": 5.0778,
      "mean_token_accuracy": 0.17701856195926666,
      "num_tokens": 46245983.0,
      "step": 21110
    },
    {
      "entropy": 5.578864717483521,
      "epoch": 2.0955736403334657,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045634330221130265,
      "loss": 5.0145,
      "mean_token_accuracy": 0.1835564762353897,
      "num_tokens": 46256855.0,
      "step": 21115
    },
    {
      "entropy": 5.627338981628418,
      "epoch": 2.096069868995633,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004563223260738589,
      "loss": 4.9955,
      "mean_token_accuracy": 0.1852916032075882,
      "num_tokens": 46268103.0,
      "step": 21120
    },
    {
      "entropy": 5.681714153289795,
      "epoch": 2.0965660976578007,
      "grad_norm": 1.0625,
      "learning_rate": 0.000456301345440012,
      "loss": 5.1219,
      "mean_token_accuracy": 0.17974307239055634,
      "num_tokens": 46278647.0,
      "step": 21125
    },
    {
      "entropy": 5.607840347290039,
      "epoch": 2.097062326319968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004562803603102825,
      "loss": 5.0121,
      "mean_token_accuracy": 0.17975080162286758,
      "num_tokens": 46289018.0,
      "step": 21130
    },
    {
      "entropy": 5.593312740325928,
      "epoch": 2.0975585549821356,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045625937068519075,
      "loss": 5.0016,
      "mean_token_accuracy": 0.17977760285139083,
      "num_tokens": 46300491.0,
      "step": 21135
    },
    {
      "entropy": 5.610566902160644,
      "epoch": 2.098054783644303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004562383765652571,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17392949461936952,
      "num_tokens": 46313322.0,
      "step": 21140
    },
    {
      "entropy": 5.688094806671143,
      "epoch": 2.098551012306471,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004562173779510023,
      "loss": 5.1683,
      "mean_token_accuracy": 0.17288658618927003,
      "num_tokens": 46326581.0,
      "step": 21145
    },
    {
      "entropy": 5.634061527252197,
      "epoch": 2.0990472409686385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000456196374842947,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18436722755432128,
      "num_tokens": 46337723.0,
      "step": 21150
    },
    {
      "entropy": 5.604186391830444,
      "epoch": 2.099543469630806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045617536724161203,
      "loss": 4.986,
      "mean_token_accuracy": 0.18553417325019836,
      "num_tokens": 46347290.0,
      "step": 21155
    },
    {
      "entropy": 5.599689674377442,
      "epoch": 2.1000396982929734,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045615435514751834,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18593654483556749,
      "num_tokens": 46358040.0,
      "step": 21160
    },
    {
      "entropy": 5.631352472305298,
      "epoch": 2.100535926955141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000456133338561187,
      "loss": 5.0663,
      "mean_token_accuracy": 0.1794649824500084,
      "num_tokens": 46369692.0,
      "step": 21165
    },
    {
      "entropy": 5.600844764709473,
      "epoch": 2.1010321556173084,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004561123174831391,
      "loss": 5.0717,
      "mean_token_accuracy": 0.18038979172706604,
      "num_tokens": 46381030.0,
      "step": 21170
    },
    {
      "entropy": 5.579244375228882,
      "epoch": 2.101528384279476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004560912919138962,
      "loss": 4.967,
      "mean_token_accuracy": 0.17979490906000137,
      "num_tokens": 46391321.0,
      "step": 21175
    },
    {
      "entropy": 5.582192659378052,
      "epoch": 2.1020246129416433,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045607026185397924,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1770518258213997,
      "num_tokens": 46403750.0,
      "step": 21180
    },
    {
      "entropy": 5.601483535766602,
      "epoch": 2.102520841603811,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004560492273039101,
      "loss": 4.9752,
      "mean_token_accuracy": 0.18002072125673294,
      "num_tokens": 46413619.0,
      "step": 21185
    },
    {
      "entropy": 5.600888919830322,
      "epoch": 2.1030170702659787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045602818826421024,
      "loss": 5.0065,
      "mean_token_accuracy": 0.1847044885158539,
      "num_tokens": 46423757.0,
      "step": 21190
    },
    {
      "entropy": 5.5917733669281,
      "epoch": 2.103513298928146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004560071447354013,
      "loss": 4.9556,
      "mean_token_accuracy": 0.1904408663511276,
      "num_tokens": 46433885.0,
      "step": 21195
    },
    {
      "entropy": 5.572417497634888,
      "epoch": 2.1040095275903137,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004559860967180053,
      "loss": 4.9669,
      "mean_token_accuracy": 0.18722977340221406,
      "num_tokens": 46445149.0,
      "step": 21200
    },
    {
      "entropy": 5.558375978469849,
      "epoch": 2.104505756252481,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004559650442125441,
      "loss": 4.9984,
      "mean_token_accuracy": 0.18225550502538682,
      "num_tokens": 46456417.0,
      "step": 21205
    },
    {
      "entropy": 5.539513635635376,
      "epoch": 2.1050019849146486,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004559439872195397,
      "loss": 5.0354,
      "mean_token_accuracy": 0.18154238909482956,
      "num_tokens": 46469093.0,
      "step": 21210
    },
    {
      "entropy": 5.677435636520386,
      "epoch": 2.105498213576816,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004559229257395144,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17117315530776978,
      "num_tokens": 46479940.0,
      "step": 21215
    },
    {
      "entropy": 5.62581696510315,
      "epoch": 2.1059944422389836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045590185977299025,
      "loss": 4.992,
      "mean_token_accuracy": 0.18301042467355727,
      "num_tokens": 46490231.0,
      "step": 21220
    },
    {
      "entropy": 5.606288099288941,
      "epoch": 2.106490670901151,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004558807893204899,
      "loss": 5.0235,
      "mean_token_accuracy": 0.17482503950595857,
      "num_tokens": 46501229.0,
      "step": 21225
    },
    {
      "entropy": 5.656154823303223,
      "epoch": 2.106986899563319,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004558597143825357,
      "loss": 5.1469,
      "mean_token_accuracy": 0.17541915327310562,
      "num_tokens": 46513657.0,
      "step": 21230
    },
    {
      "entropy": 5.625735950469971,
      "epoch": 2.1074831282254864,
      "grad_norm": 1.125,
      "learning_rate": 0.0004558386349596503,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1876465007662773,
      "num_tokens": 46525655.0,
      "step": 21235
    },
    {
      "entropy": 5.664568328857422,
      "epoch": 2.107979356887654,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004558175510523564,
      "loss": 5.0677,
      "mean_token_accuracy": 0.17532590478658677,
      "num_tokens": 46538313.0,
      "step": 21240
    },
    {
      "entropy": 5.710178327560425,
      "epoch": 2.1084755855498214,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045579646266117696,
      "loss": 5.1151,
      "mean_token_accuracy": 0.17573546916246413,
      "num_tokens": 46551311.0,
      "step": 21245
    },
    {
      "entropy": 5.596586513519287,
      "epoch": 2.108971814211989,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045577536978663475,
      "loss": 4.9762,
      "mean_token_accuracy": 0.18724369406700134,
      "num_tokens": 46561468.0,
      "step": 21250
    },
    {
      "entropy": 5.578286933898926,
      "epoch": 2.1094680428741563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004557542724292529,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18574392795562744,
      "num_tokens": 46572447.0,
      "step": 21255
    },
    {
      "entropy": 5.587227630615234,
      "epoch": 2.109964271536324,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004557331705895545,
      "loss": 4.9806,
      "mean_token_accuracy": 0.1840259090065956,
      "num_tokens": 46583736.0,
      "step": 21260
    },
    {
      "entropy": 5.534249496459961,
      "epoch": 2.1104605001984913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004557120642680631,
      "loss": 4.9483,
      "mean_token_accuracy": 0.18266794830560684,
      "num_tokens": 46595182.0,
      "step": 21265
    },
    {
      "entropy": 5.643159055709839,
      "epoch": 2.1109567288606588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004556909534653018,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18170541375875474,
      "num_tokens": 46606058.0,
      "step": 21270
    },
    {
      "entropy": 5.603690767288208,
      "epoch": 2.1114529575228267,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004556698381817942,
      "loss": 4.9708,
      "mean_token_accuracy": 0.18036742955446244,
      "num_tokens": 46617862.0,
      "step": 21275
    },
    {
      "entropy": 5.564745569229126,
      "epoch": 2.111949186184994,
      "grad_norm": 1.09375,
      "learning_rate": 0.000455648718418064,
      "loss": 5.0182,
      "mean_token_accuracy": 0.17796906530857087,
      "num_tokens": 46627708.0,
      "step": 21280
    },
    {
      "entropy": 5.561023902893067,
      "epoch": 2.1124454148471616,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004556275941746348,
      "loss": 4.9681,
      "mean_token_accuracy": 0.18391259461641313,
      "num_tokens": 46638263.0,
      "step": 21285
    },
    {
      "entropy": 5.684447813034057,
      "epoch": 2.112941643509329,
      "grad_norm": 1.125,
      "learning_rate": 0.00045560646545203044,
      "loss": 5.0318,
      "mean_token_accuracy": 0.17842359244823455,
      "num_tokens": 46648554.0,
      "step": 21290
    },
    {
      "entropy": 5.621147727966308,
      "epoch": 2.1134378721714966,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045558533225077496,
      "loss": 5.0906,
      "mean_token_accuracy": 0.1761091783642769,
      "num_tokens": 46659081.0,
      "step": 21295
    },
    {
      "entropy": 5.655944061279297,
      "epoch": 2.113934100833664,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045556419457139235,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17872451096773148,
      "num_tokens": 46670452.0,
      "step": 21300
    },
    {
      "entropy": 5.678097724914551,
      "epoch": 2.1144303294958315,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045554305241440684,
      "loss": 5.1142,
      "mean_token_accuracy": 0.17388912290334702,
      "num_tokens": 46680911.0,
      "step": 21305
    },
    {
      "entropy": 5.654494953155518,
      "epoch": 2.114926558157999,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004555219057803426,
      "loss": 5.0054,
      "mean_token_accuracy": 0.1883489280939102,
      "num_tokens": 46692367.0,
      "step": 21310
    },
    {
      "entropy": 5.579535245895386,
      "epoch": 2.115422786820167,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004555007546697242,
      "loss": 4.9795,
      "mean_token_accuracy": 0.18641866147518157,
      "num_tokens": 46702198.0,
      "step": 21315
    },
    {
      "entropy": 5.600734567642212,
      "epoch": 2.1159190154823344,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045547959908307585,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18715034127235414,
      "num_tokens": 46714887.0,
      "step": 21320
    },
    {
      "entropy": 5.636063575744629,
      "epoch": 2.116415244144502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004554584390209225,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1752977192401886,
      "num_tokens": 46725725.0,
      "step": 21325
    },
    {
      "entropy": 5.633595371246338,
      "epoch": 2.1169114728066694,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004554372744837886,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1768911749124527,
      "num_tokens": 46737105.0,
      "step": 21330
    },
    {
      "entropy": 5.593023014068604,
      "epoch": 2.117407701468837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004554161054721992,
      "loss": 5.0614,
      "mean_token_accuracy": 0.17969293594360353,
      "num_tokens": 46747287.0,
      "step": 21335
    },
    {
      "entropy": 5.586045360565185,
      "epoch": 2.1179039301310043,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004553949319866791,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18497083187103272,
      "num_tokens": 46757818.0,
      "step": 21340
    },
    {
      "entropy": 5.589415168762207,
      "epoch": 2.118400158793172,
      "grad_norm": 1.0,
      "learning_rate": 0.00045537375402775334,
      "loss": 4.9725,
      "mean_token_accuracy": 0.1829960823059082,
      "num_tokens": 46768086.0,
      "step": 21345
    },
    {
      "entropy": 5.532119369506836,
      "epoch": 2.1188963874553393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045535257159594724,
      "loss": 4.9759,
      "mean_token_accuracy": 0.18849024623632432,
      "num_tokens": 46779591.0,
      "step": 21350
    },
    {
      "entropy": 5.628573846817017,
      "epoch": 2.1193926161175067,
      "grad_norm": 1.015625,
      "learning_rate": 0.000455331384691786,
      "loss": 5.0473,
      "mean_token_accuracy": 0.17739235311746598,
      "num_tokens": 46790843.0,
      "step": 21355
    },
    {
      "entropy": 5.659381103515625,
      "epoch": 2.1198888447796747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004553101933157949,
      "loss": 5.0139,
      "mean_token_accuracy": 0.18132198303937913,
      "num_tokens": 46802046.0,
      "step": 21360
    },
    {
      "entropy": 5.601346778869629,
      "epoch": 2.120385073441842,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045528899746849965,
      "loss": 4.9578,
      "mean_token_accuracy": 0.18534009009599686,
      "num_tokens": 46812955.0,
      "step": 21365
    },
    {
      "entropy": 5.631674098968506,
      "epoch": 2.1208813021040096,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004552677971504257,
      "loss": 5.0606,
      "mean_token_accuracy": 0.18400646597146988,
      "num_tokens": 46824206.0,
      "step": 21370
    },
    {
      "entropy": 5.543349075317383,
      "epoch": 2.121377530766177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004552465923620988,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1858616963028908,
      "num_tokens": 46835352.0,
      "step": 21375
    },
    {
      "entropy": 5.603107595443726,
      "epoch": 2.1218737594283446,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004552253831040448,
      "loss": 5.0286,
      "mean_token_accuracy": 0.18181214928627015,
      "num_tokens": 46847071.0,
      "step": 21380
    },
    {
      "entropy": 5.594523000717163,
      "epoch": 2.122369988090512,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045520416937678973,
      "loss": 4.9436,
      "mean_token_accuracy": 0.1822444975376129,
      "num_tokens": 46856802.0,
      "step": 21385
    },
    {
      "entropy": 5.579753637313843,
      "epoch": 2.1228662167526795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045518295118085955,
      "loss": 4.9562,
      "mean_token_accuracy": 0.1850918084383011,
      "num_tokens": 46867549.0,
      "step": 21390
    },
    {
      "entropy": 5.600651502609253,
      "epoch": 2.123362445414847,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045516172851678036,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17267704904079437,
      "num_tokens": 46877162.0,
      "step": 21395
    },
    {
      "entropy": 5.568783140182495,
      "epoch": 2.123858674077015,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045514050138507857,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18882086127996445,
      "num_tokens": 46889418.0,
      "step": 21400
    },
    {
      "entropy": 5.624409484863281,
      "epoch": 2.1243549027391824,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551192697862805,
      "loss": 5.0315,
      "mean_token_accuracy": 0.17802696526050568,
      "num_tokens": 46900778.0,
      "step": 21405
    },
    {
      "entropy": 5.560769128799438,
      "epoch": 2.12485113140135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004550980337209127,
      "loss": 4.997,
      "mean_token_accuracy": 0.1829338863492012,
      "num_tokens": 46911383.0,
      "step": 21410
    },
    {
      "entropy": 5.6172548770904545,
      "epoch": 2.1253473600635173,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045507679318950166,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1805106669664383,
      "num_tokens": 46922066.0,
      "step": 21415
    },
    {
      "entropy": 5.628243064880371,
      "epoch": 2.125843588725685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004550555481925743,
      "loss": 5.1359,
      "mean_token_accuracy": 0.17289932668209076,
      "num_tokens": 46933317.0,
      "step": 21420
    },
    {
      "entropy": 5.630956506729126,
      "epoch": 2.1263398173878523,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045503429873065723,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18688475489616393,
      "num_tokens": 46945437.0,
      "step": 21425
    },
    {
      "entropy": 5.585095405578613,
      "epoch": 2.1268360460500197,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004550130448042775,
      "loss": 4.991,
      "mean_token_accuracy": 0.18130427449941636,
      "num_tokens": 46957455.0,
      "step": 21430
    },
    {
      "entropy": 5.632899761199951,
      "epoch": 2.127332274712187,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004549917864139622,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17498193979263305,
      "num_tokens": 46967626.0,
      "step": 21435
    },
    {
      "entropy": 5.555804300308227,
      "epoch": 2.1278285033743547,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004549705235602384,
      "loss": 5.0012,
      "mean_token_accuracy": 0.18070120960474015,
      "num_tokens": 46979492.0,
      "step": 21440
    },
    {
      "entropy": 5.5990368843078615,
      "epoch": 2.1283247320365226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004549492562436334,
      "loss": 5.0135,
      "mean_token_accuracy": 0.1863516479730606,
      "num_tokens": 46990803.0,
      "step": 21445
    },
    {
      "entropy": 5.597627115249634,
      "epoch": 2.12882096069869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045492798446467465,
      "loss": 5.0154,
      "mean_token_accuracy": 0.18997643887996674,
      "num_tokens": 47000771.0,
      "step": 21450
    },
    {
      "entropy": 5.60591082572937,
      "epoch": 2.1293171893608576,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045490670822388957,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18513640016317368,
      "num_tokens": 47011461.0,
      "step": 21455
    },
    {
      "entropy": 5.507507944107056,
      "epoch": 2.129813418023025,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004548854275218059,
      "loss": 4.8663,
      "mean_token_accuracy": 0.19810278564691544,
      "num_tokens": 47021164.0,
      "step": 21460
    },
    {
      "entropy": 5.58200888633728,
      "epoch": 2.1303096466851925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045486414235895115,
      "loss": 5.064,
      "mean_token_accuracy": 0.1755203440785408,
      "num_tokens": 47032150.0,
      "step": 21465
    },
    {
      "entropy": 5.648106861114502,
      "epoch": 2.13080587534736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004548428527358532,
      "loss": 5.016,
      "mean_token_accuracy": 0.1850614756345749,
      "num_tokens": 47042853.0,
      "step": 21470
    },
    {
      "entropy": 5.6467469215393065,
      "epoch": 2.1313021040095275,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004548215586530401,
      "loss": 4.942,
      "mean_token_accuracy": 0.18079358339309692,
      "num_tokens": 47053197.0,
      "step": 21475
    },
    {
      "entropy": 5.6171653270721436,
      "epoch": 2.131798332671695,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045480026011103975,
      "loss": 5.0408,
      "mean_token_accuracy": 0.1817584976553917,
      "num_tokens": 47063234.0,
      "step": 21480
    },
    {
      "entropy": 5.6184220790863035,
      "epoch": 2.132294561333863,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045477895711038045,
      "loss": 4.997,
      "mean_token_accuracy": 0.18391293585300444,
      "num_tokens": 47074291.0,
      "step": 21485
    },
    {
      "entropy": 5.590197038650513,
      "epoch": 2.1327907899960303,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004547576496515904,
      "loss": 5.0111,
      "mean_token_accuracy": 0.178131602704525,
      "num_tokens": 47084647.0,
      "step": 21490
    },
    {
      "entropy": 5.5880921363830565,
      "epoch": 2.133287018658198,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045473633773519795,
      "loss": 5.0507,
      "mean_token_accuracy": 0.18466037958860398,
      "num_tokens": 47096590.0,
      "step": 21495
    },
    {
      "entropy": 5.669073534011841,
      "epoch": 2.1337832473203653,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045471502136173173,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17812456637620927,
      "num_tokens": 47107740.0,
      "step": 21500
    },
    {
      "entropy": 5.663896465301514,
      "epoch": 2.1342794759825328,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004546937005317201,
      "loss": 4.9888,
      "mean_token_accuracy": 0.17892660647630693,
      "num_tokens": 47117665.0,
      "step": 21505
    },
    {
      "entropy": 5.704553461074829,
      "epoch": 2.1347757046447002,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045467237524569193,
      "loss": 5.116,
      "mean_token_accuracy": 0.17152834683656693,
      "num_tokens": 47129018.0,
      "step": 21510
    },
    {
      "entropy": 5.634072017669678,
      "epoch": 2.1352719333068677,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045465104550417604,
      "loss": 4.9904,
      "mean_token_accuracy": 0.18819151520729066,
      "num_tokens": 47139882.0,
      "step": 21515
    },
    {
      "entropy": 5.593241500854492,
      "epoch": 2.135768161969035,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045462971130770123,
      "loss": 4.9855,
      "mean_token_accuracy": 0.1822924792766571,
      "num_tokens": 47149294.0,
      "step": 21520
    },
    {
      "entropy": 5.72894253730774,
      "epoch": 2.1362643906312027,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004546083726567968,
      "loss": 5.2451,
      "mean_token_accuracy": 0.16815827190876007,
      "num_tokens": 47161604.0,
      "step": 21525
    },
    {
      "entropy": 5.704651641845703,
      "epoch": 2.1367606192933706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004545870295519917,
      "loss": 5.08,
      "mean_token_accuracy": 0.18445676416158677,
      "num_tokens": 47171730.0,
      "step": 21530
    },
    {
      "entropy": 5.55524468421936,
      "epoch": 2.137256847955538,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045456568199381524,
      "loss": 4.9573,
      "mean_token_accuracy": 0.1840907096862793,
      "num_tokens": 47181167.0,
      "step": 21535
    },
    {
      "entropy": 5.619512319564819,
      "epoch": 2.1377530766177055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045454432998279676,
      "loss": 5.0783,
      "mean_token_accuracy": 0.18187799155712128,
      "num_tokens": 47192904.0,
      "step": 21540
    },
    {
      "entropy": 5.748489713668823,
      "epoch": 2.138249305279873,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004545229735194658,
      "loss": 5.1386,
      "mean_token_accuracy": 0.17197660356760025,
      "num_tokens": 47203653.0,
      "step": 21545
    },
    {
      "entropy": 5.652518463134766,
      "epoch": 2.1387455339420405,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045450161260435196,
      "loss": 5.0304,
      "mean_token_accuracy": 0.1883913442492485,
      "num_tokens": 47215835.0,
      "step": 21550
    },
    {
      "entropy": 5.615349912643433,
      "epoch": 2.139241762604208,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004544802472379849,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17797956764698028,
      "num_tokens": 47227588.0,
      "step": 21555
    },
    {
      "entropy": 5.603651523590088,
      "epoch": 2.1397379912663754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004544588774208945,
      "loss": 5.0618,
      "mean_token_accuracy": 0.1775931015610695,
      "num_tokens": 47239119.0,
      "step": 21560
    },
    {
      "entropy": 5.6764568328857425,
      "epoch": 2.140234219928543,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045443750315361055,
      "loss": 5.0472,
      "mean_token_accuracy": 0.18722014576196672,
      "num_tokens": 47251047.0,
      "step": 21565
    },
    {
      "entropy": 5.695380735397339,
      "epoch": 2.140730448590711,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004544161244366632,
      "loss": 5.0833,
      "mean_token_accuracy": 0.18063765317201613,
      "num_tokens": 47260811.0,
      "step": 21570
    },
    {
      "entropy": 5.6383385181427,
      "epoch": 2.1412266772528783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004543947412705826,
      "loss": 4.9808,
      "mean_token_accuracy": 0.1814132109284401,
      "num_tokens": 47271175.0,
      "step": 21575
    },
    {
      "entropy": 5.5625471591949465,
      "epoch": 2.1417229059150458,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045437335365589905,
      "loss": 4.9152,
      "mean_token_accuracy": 0.1898696467280388,
      "num_tokens": 47282191.0,
      "step": 21580
    },
    {
      "entropy": 5.529517126083374,
      "epoch": 2.1422191345772132,
      "grad_norm": 2.53125,
      "learning_rate": 0.00045435196159314276,
      "loss": 4.9466,
      "mean_token_accuracy": 0.19128209948539734,
      "num_tokens": 47292222.0,
      "step": 21585
    },
    {
      "entropy": 5.554851102828979,
      "epoch": 2.1427153632393807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045433056508284425,
      "loss": 5.0579,
      "mean_token_accuracy": 0.18099926710128783,
      "num_tokens": 47302936.0,
      "step": 21590
    },
    {
      "entropy": 5.621550703048706,
      "epoch": 2.143211591901548,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004543091641255342,
      "loss": 4.9578,
      "mean_token_accuracy": 0.18945959955453873,
      "num_tokens": 47314126.0,
      "step": 21595
    },
    {
      "entropy": 5.626703214645386,
      "epoch": 2.1437078205637157,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004542877587217432,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18560295253992082,
      "num_tokens": 47324038.0,
      "step": 21600
    },
    {
      "entropy": 5.554405546188354,
      "epoch": 2.144204049225883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045426634887200224,
      "loss": 4.9855,
      "mean_token_accuracy": 0.18637784868478774,
      "num_tokens": 47334820.0,
      "step": 21605
    },
    {
      "entropy": 5.579324102401733,
      "epoch": 2.1447002778880506,
      "grad_norm": 1.078125,
      "learning_rate": 0.000454244934576842,
      "loss": 5.0188,
      "mean_token_accuracy": 0.18049055486917495,
      "num_tokens": 47345039.0,
      "step": 21610
    },
    {
      "entropy": 5.6812659740448,
      "epoch": 2.1451965065502185,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045422351583679365,
      "loss": 5.0508,
      "mean_token_accuracy": 0.18169586956501008,
      "num_tokens": 47354668.0,
      "step": 21615
    },
    {
      "entropy": 5.6552345752716064,
      "epoch": 2.145692735212386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004542020926523883,
      "loss": 5.0572,
      "mean_token_accuracy": 0.18152814358472824,
      "num_tokens": 47366002.0,
      "step": 21620
    },
    {
      "entropy": 5.6395583152771,
      "epoch": 2.1461889638745535,
      "grad_norm": 1.0,
      "learning_rate": 0.0004541806650241572,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18581708818674086,
      "num_tokens": 47377052.0,
      "step": 21625
    },
    {
      "entropy": 5.614376544952393,
      "epoch": 2.146685192536721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541592329526317,
      "loss": 5.0323,
      "mean_token_accuracy": 0.17310748100280762,
      "num_tokens": 47387598.0,
      "step": 21630
    },
    {
      "entropy": 5.596726799011231,
      "epoch": 2.1471814211988884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004541377964383434,
      "loss": 5.0314,
      "mean_token_accuracy": 0.18029216378927232,
      "num_tokens": 47398565.0,
      "step": 21635
    },
    {
      "entropy": 5.619342422485351,
      "epoch": 2.147677649861056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541163554818237,
      "loss": 5.005,
      "mean_token_accuracy": 0.1802336886525154,
      "num_tokens": 47408715.0,
      "step": 21640
    },
    {
      "entropy": 5.719729423522949,
      "epoch": 2.1481738785232234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045409491008360423,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17501294016838073,
      "num_tokens": 47418690.0,
      "step": 21645
    },
    {
      "entropy": 5.637933206558228,
      "epoch": 2.148670107185391,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004540734602442169,
      "loss": 5.0412,
      "mean_token_accuracy": 0.17870535850524902,
      "num_tokens": 47429415.0,
      "step": 21650
    },
    {
      "entropy": 5.602420711517334,
      "epoch": 2.149166335847559,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004540520059641937,
      "loss": 5.0413,
      "mean_token_accuracy": 0.1837388589978218,
      "num_tokens": 47440852.0,
      "step": 21655
    },
    {
      "entropy": 5.606856107711792,
      "epoch": 2.1496625645097263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004540305472440666,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18571242541074753,
      "num_tokens": 47450800.0,
      "step": 21660
    },
    {
      "entropy": 5.674269962310791,
      "epoch": 2.1501587931718937,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004540090840843677,
      "loss": 5.0948,
      "mean_token_accuracy": 0.18023694753646852,
      "num_tokens": 47460900.0,
      "step": 21665
    },
    {
      "entropy": 5.634238433837891,
      "epoch": 2.150655021834061,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004539876164856292,
      "loss": 5.0356,
      "mean_token_accuracy": 0.18543240427970886,
      "num_tokens": 47471182.0,
      "step": 21670
    },
    {
      "entropy": 5.5706182479858395,
      "epoch": 2.1511512504962287,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004539661444483835,
      "loss": 4.9729,
      "mean_token_accuracy": 0.18496343195438386,
      "num_tokens": 47482348.0,
      "step": 21675
    },
    {
      "entropy": 5.696905469894409,
      "epoch": 2.151647479158396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004539446679731631,
      "loss": 5.0635,
      "mean_token_accuracy": 0.18529872596263885,
      "num_tokens": 47494642.0,
      "step": 21680
    },
    {
      "entropy": 5.66732964515686,
      "epoch": 2.1521437078205636,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004539231870605005,
      "loss": 4.956,
      "mean_token_accuracy": 0.18650461584329606,
      "num_tokens": 47504965.0,
      "step": 21685
    },
    {
      "entropy": 5.630079078674316,
      "epoch": 2.152639936482731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045390170171092845,
      "loss": 5.0322,
      "mean_token_accuracy": 0.18401335775852204,
      "num_tokens": 47516071.0,
      "step": 21690
    },
    {
      "entropy": 5.699949932098389,
      "epoch": 2.1531361651448986,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045388021192497956,
      "loss": 5.1128,
      "mean_token_accuracy": 0.17248342484235762,
      "num_tokens": 47527765.0,
      "step": 21695
    },
    {
      "entropy": 5.620795106887817,
      "epoch": 2.1536323938070665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000453858717703187,
      "loss": 4.9833,
      "mean_token_accuracy": 0.17555208653211593,
      "num_tokens": 47538526.0,
      "step": 21700
    },
    {
      "entropy": 5.706483364105225,
      "epoch": 2.154128622469234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004538372190460836,
      "loss": 5.1302,
      "mean_token_accuracy": 0.17395668625831603,
      "num_tokens": 47550061.0,
      "step": 21705
    },
    {
      "entropy": 5.600920963287353,
      "epoch": 2.1546248511314015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004538157159542025,
      "loss": 4.9686,
      "mean_token_accuracy": 0.1863330602645874,
      "num_tokens": 47559955.0,
      "step": 21710
    },
    {
      "entropy": 5.623594570159912,
      "epoch": 2.155121079793569,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004537942084280771,
      "loss": 5.046,
      "mean_token_accuracy": 0.17800690978765488,
      "num_tokens": 47570768.0,
      "step": 21715
    },
    {
      "entropy": 5.578223085403442,
      "epoch": 2.1556173084557364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045377269646824036,
      "loss": 4.9621,
      "mean_token_accuracy": 0.18024650514125823,
      "num_tokens": 47581296.0,
      "step": 21720
    },
    {
      "entropy": 5.651197528839111,
      "epoch": 2.156113537117904,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045375118007522607,
      "loss": 5.0909,
      "mean_token_accuracy": 0.1741702973842621,
      "num_tokens": 47591481.0,
      "step": 21725
    },
    {
      "entropy": 5.590255212783814,
      "epoch": 2.1566097657800714,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004537296592495678,
      "loss": 5.0531,
      "mean_token_accuracy": 0.1833728551864624,
      "num_tokens": 47601920.0,
      "step": 21730
    },
    {
      "entropy": 5.573930263519287,
      "epoch": 2.157105994442239,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000453708133991799,
      "loss": 4.9482,
      "mean_token_accuracy": 0.18905789405107498,
      "num_tokens": 47612454.0,
      "step": 21735
    },
    {
      "entropy": 5.612695837020874,
      "epoch": 2.1576022231044063,
      "grad_norm": 1.0,
      "learning_rate": 0.00045368660430245364,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17382052838802337,
      "num_tokens": 47624134.0,
      "step": 21740
    },
    {
      "entropy": 5.62518539428711,
      "epoch": 2.158098451766574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004536650701820654,
      "loss": 5.0494,
      "mean_token_accuracy": 0.18036366254091263,
      "num_tokens": 47635789.0,
      "step": 21745
    },
    {
      "entropy": 5.607472515106201,
      "epoch": 2.1585946804287417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004536435316311685,
      "loss": 5.0034,
      "mean_token_accuracy": 0.18499118834733963,
      "num_tokens": 47645738.0,
      "step": 21750
    },
    {
      "entropy": 5.627914810180664,
      "epoch": 2.159090909090909,
      "grad_norm": 1.0,
      "learning_rate": 0.00045362198865029696,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17780595123767853,
      "num_tokens": 47657027.0,
      "step": 21755
    },
    {
      "entropy": 5.638756942749024,
      "epoch": 2.1595871377530766,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045360044123998495,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18356901109218599,
      "num_tokens": 47667257.0,
      "step": 21760
    },
    {
      "entropy": 5.605485725402832,
      "epoch": 2.160083366415244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045357888940076685,
      "loss": 5.0313,
      "mean_token_accuracy": 0.17905918955802919,
      "num_tokens": 47677247.0,
      "step": 21765
    },
    {
      "entropy": 5.651819324493408,
      "epoch": 2.1605795950774116,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004535573331331771,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17568423151969909,
      "num_tokens": 47688751.0,
      "step": 21770
    },
    {
      "entropy": 5.637729549407959,
      "epoch": 2.161075823739579,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045353577243775035,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17730163484811784,
      "num_tokens": 47700347.0,
      "step": 21775
    },
    {
      "entropy": 5.631749296188355,
      "epoch": 2.1615720524017465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004535142073150211,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18833346664905548,
      "num_tokens": 47710187.0,
      "step": 21780
    },
    {
      "entropy": 5.668229055404663,
      "epoch": 2.162068281063914,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004534926377655241,
      "loss": 5.0743,
      "mean_token_accuracy": 0.18138637095689775,
      "num_tokens": 47720461.0,
      "step": 21785
    },
    {
      "entropy": 5.616532325744629,
      "epoch": 2.162564509726082,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045347106378979435,
      "loss": 4.984,
      "mean_token_accuracy": 0.18854275345802307,
      "num_tokens": 47730457.0,
      "step": 21790
    },
    {
      "entropy": 5.595559740066529,
      "epoch": 2.1630607383882494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045344948538836674,
      "loss": 5.0197,
      "mean_token_accuracy": 0.1798556312918663,
      "num_tokens": 47742390.0,
      "step": 21795
    },
    {
      "entropy": 5.673142337799073,
      "epoch": 2.163556967050417,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045342790256177646,
      "loss": 5.1472,
      "mean_token_accuracy": 0.17768627107143403,
      "num_tokens": 47752783.0,
      "step": 21800
    },
    {
      "entropy": 5.612023067474365,
      "epoch": 2.1640531957125844,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004534063153105587,
      "loss": 5.0214,
      "mean_token_accuracy": 0.18754320293664933,
      "num_tokens": 47762997.0,
      "step": 21805
    },
    {
      "entropy": 5.614993715286255,
      "epoch": 2.164549424374752,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004533847236352487,
      "loss": 5.096,
      "mean_token_accuracy": 0.17683973163366318,
      "num_tokens": 47773152.0,
      "step": 21810
    },
    {
      "entropy": 5.588564205169678,
      "epoch": 2.1650456530369193,
      "grad_norm": 1.0,
      "learning_rate": 0.000453363127536382,
      "loss": 4.9772,
      "mean_token_accuracy": 0.18774286806583404,
      "num_tokens": 47784392.0,
      "step": 21815
    },
    {
      "entropy": 5.631738805770874,
      "epoch": 2.165541881699087,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045334152701449407,
      "loss": 4.962,
      "mean_token_accuracy": 0.18159456104040145,
      "num_tokens": 47795775.0,
      "step": 21820
    },
    {
      "entropy": 5.701515102386475,
      "epoch": 2.1660381103612543,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004533199220701205,
      "loss": 5.1631,
      "mean_token_accuracy": 0.17102012038230896,
      "num_tokens": 47808340.0,
      "step": 21825
    },
    {
      "entropy": 5.623595333099365,
      "epoch": 2.166534339023422,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045329831270379714,
      "loss": 5.0778,
      "mean_token_accuracy": 0.17671456485986708,
      "num_tokens": 47819591.0,
      "step": 21830
    },
    {
      "entropy": 5.6318803310394285,
      "epoch": 2.1670305676855897,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045327669891605975,
      "loss": 5.072,
      "mean_token_accuracy": 0.17305649518966676,
      "num_tokens": 47830551.0,
      "step": 21835
    },
    {
      "entropy": 5.645876121520996,
      "epoch": 2.167526796347757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004532550807074445,
      "loss": 5.0255,
      "mean_token_accuracy": 0.1783945381641388,
      "num_tokens": 47839553.0,
      "step": 21840
    },
    {
      "entropy": 5.582621288299561,
      "epoch": 2.1680230250099246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004532334580784873,
      "loss": 5.0312,
      "mean_token_accuracy": 0.182449272274971,
      "num_tokens": 47850449.0,
      "step": 21845
    },
    {
      "entropy": 5.648117399215698,
      "epoch": 2.168519253672092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532118310297244,
      "loss": 5.0425,
      "mean_token_accuracy": 0.1752578854560852,
      "num_tokens": 47861751.0,
      "step": 21850
    },
    {
      "entropy": 5.605733489990234,
      "epoch": 2.1690154823342596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004531901995616921,
      "loss": 4.9545,
      "mean_token_accuracy": 0.1841782733798027,
      "num_tokens": 47872019.0,
      "step": 21855
    },
    {
      "entropy": 5.633587169647217,
      "epoch": 2.169511710996427,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004531685636749268,
      "loss": 5.0488,
      "mean_token_accuracy": 0.17591358423233033,
      "num_tokens": 47882271.0,
      "step": 21860
    },
    {
      "entropy": 5.57750973701477,
      "epoch": 2.1700079396585945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004531469233699651,
      "loss": 5.014,
      "mean_token_accuracy": 0.1824018359184265,
      "num_tokens": 47892036.0,
      "step": 21865
    },
    {
      "entropy": 5.610058403015136,
      "epoch": 2.170504168320762,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004531252786473436,
      "loss": 5.069,
      "mean_token_accuracy": 0.18263627588748932,
      "num_tokens": 47902760.0,
      "step": 21870
    },
    {
      "entropy": 5.628575563430786,
      "epoch": 2.17100039698293,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004531036295075989,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18048920035362243,
      "num_tokens": 47913954.0,
      "step": 21875
    },
    {
      "entropy": 5.659623098373413,
      "epoch": 2.1714966256450974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045308197595126797,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17891908437013626,
      "num_tokens": 47925167.0,
      "step": 21880
    },
    {
      "entropy": 5.621177911758423,
      "epoch": 2.171992854307265,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004530603179788879,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18778740912675856,
      "num_tokens": 47934828.0,
      "step": 21885
    },
    {
      "entropy": 5.59321141242981,
      "epoch": 2.1724890829694323,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004530386555909955,
      "loss": 5.0369,
      "mean_token_accuracy": 0.1802393153309822,
      "num_tokens": 47945716.0,
      "step": 21890
    },
    {
      "entropy": 5.689624977111817,
      "epoch": 2.1729853116316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045301698878812813,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1745626375079155,
      "num_tokens": 47956653.0,
      "step": 21895
    },
    {
      "entropy": 5.574101114273072,
      "epoch": 2.1734815402937673,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004529953175708231,
      "loss": 4.8605,
      "mean_token_accuracy": 0.1930001348257065,
      "num_tokens": 47967404.0,
      "step": 21900
    },
    {
      "entropy": 5.600957441329956,
      "epoch": 2.1739777689559348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045297364193961763,
      "loss": 4.9305,
      "mean_token_accuracy": 0.18839868009090424,
      "num_tokens": 47977281.0,
      "step": 21905
    },
    {
      "entropy": 5.62956132888794,
      "epoch": 2.1744739976181022,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045295196189504934,
      "loss": 5.033,
      "mean_token_accuracy": 0.18009577989578246,
      "num_tokens": 47988009.0,
      "step": 21910
    },
    {
      "entropy": 5.64723048210144,
      "epoch": 2.17497022628027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045293027743765595,
      "loss": 5.1304,
      "mean_token_accuracy": 0.17281622737646102,
      "num_tokens": 47999403.0,
      "step": 21915
    },
    {
      "entropy": 5.665315961837768,
      "epoch": 2.1754664549424376,
      "grad_norm": 1.0,
      "learning_rate": 0.00045290858856797503,
      "loss": 5.0555,
      "mean_token_accuracy": 0.18501553684473038,
      "num_tokens": 48011377.0,
      "step": 21920
    },
    {
      "entropy": 5.624405717849731,
      "epoch": 2.175962683604605,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045288689528654444,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18168228566646577,
      "num_tokens": 48023425.0,
      "step": 21925
    },
    {
      "entropy": 5.702217149734497,
      "epoch": 2.1764589122667726,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045286519759390217,
      "loss": 5.1371,
      "mean_token_accuracy": 0.17502327859401703,
      "num_tokens": 48035743.0,
      "step": 21930
    },
    {
      "entropy": 5.620946645736694,
      "epoch": 2.17695514092894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004528434954905863,
      "loss": 4.9109,
      "mean_token_accuracy": 0.1877771019935608,
      "num_tokens": 48045799.0,
      "step": 21935
    },
    {
      "entropy": 5.623396873474121,
      "epoch": 2.1774513695911075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004528217889771349,
      "loss": 5.083,
      "mean_token_accuracy": 0.1773052215576172,
      "num_tokens": 48056514.0,
      "step": 21940
    },
    {
      "entropy": 5.588036680221558,
      "epoch": 2.177947598253275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045280007805408636,
      "loss": 5.0133,
      "mean_token_accuracy": 0.17875383347272872,
      "num_tokens": 48067493.0,
      "step": 21945
    },
    {
      "entropy": 5.723169898986816,
      "epoch": 2.1784438269154425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045277836272197895,
      "loss": 5.17,
      "mean_token_accuracy": 0.17643133997917176,
      "num_tokens": 48078593.0,
      "step": 21950
    },
    {
      "entropy": 5.59760594367981,
      "epoch": 2.17894005557761,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045275664298135134,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18585691452026368,
      "num_tokens": 48089840.0,
      "step": 21955
    },
    {
      "entropy": 5.574075937271118,
      "epoch": 2.179436284239778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004527349188327419,
      "loss": 5.0137,
      "mean_token_accuracy": 0.17835981249809266,
      "num_tokens": 48100965.0,
      "step": 21960
    },
    {
      "entropy": 5.603652763366699,
      "epoch": 2.1799325129019453,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004527131902766895,
      "loss": 5.0317,
      "mean_token_accuracy": 0.1799773871898651,
      "num_tokens": 48111391.0,
      "step": 21965
    },
    {
      "entropy": 5.594666767120361,
      "epoch": 2.180428741564113,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004526914573137329,
      "loss": 5.0556,
      "mean_token_accuracy": 0.17783844470977783,
      "num_tokens": 48122213.0,
      "step": 21970
    },
    {
      "entropy": 5.652326059341431,
      "epoch": 2.1809249702262803,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045266971994441116,
      "loss": 5.0876,
      "mean_token_accuracy": 0.17998792380094528,
      "num_tokens": 48132966.0,
      "step": 21975
    },
    {
      "entropy": 5.642933702468872,
      "epoch": 2.1814211988884478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000452647978169263,
      "loss": 5.0072,
      "mean_token_accuracy": 0.17921024113893508,
      "num_tokens": 48144253.0,
      "step": 21980
    },
    {
      "entropy": 5.650934362411499,
      "epoch": 2.1819174275506152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045262623198882797,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18038337975740432,
      "num_tokens": 48154512.0,
      "step": 21985
    },
    {
      "entropy": 5.584538745880127,
      "epoch": 2.1824136562127827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004526044814036451,
      "loss": 5.0377,
      "mean_token_accuracy": 0.18000043481588363,
      "num_tokens": 48165296.0,
      "step": 21990
    },
    {
      "entropy": 5.596587467193603,
      "epoch": 2.18290988487495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045258272641425377,
      "loss": 5.0639,
      "mean_token_accuracy": 0.17803102731704712,
      "num_tokens": 48175878.0,
      "step": 21995
    },
    {
      "entropy": 5.633815288543701,
      "epoch": 2.183406113537118,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004525609670211934,
      "loss": 5.0638,
      "mean_token_accuracy": 0.1808975711464882,
      "num_tokens": 48187400.0,
      "step": 22000
    },
    {
      "entropy": 5.728505659103393,
      "epoch": 2.1839023421992856,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045253920322500373,
      "loss": 5.1787,
      "mean_token_accuracy": 0.17853147983551027,
      "num_tokens": 48196956.0,
      "step": 22005
    },
    {
      "entropy": 5.625637865066528,
      "epoch": 2.184398570861453,
      "grad_norm": 1.0,
      "learning_rate": 0.00045251743502622445,
      "loss": 5.0298,
      "mean_token_accuracy": 0.18231543600559236,
      "num_tokens": 48208947.0,
      "step": 22010
    },
    {
      "entropy": 5.664026308059692,
      "epoch": 2.1848947995236205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004524956624253952,
      "loss": 5.1299,
      "mean_token_accuracy": 0.17732467800378798,
      "num_tokens": 48220370.0,
      "step": 22015
    },
    {
      "entropy": 5.606225395202637,
      "epoch": 2.185391028185788,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045247388542305595,
      "loss": 5.0327,
      "mean_token_accuracy": 0.17459086775779725,
      "num_tokens": 48231977.0,
      "step": 22020
    },
    {
      "entropy": 5.631925010681153,
      "epoch": 2.1858872568479555,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004524521040197469,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17527895122766496,
      "num_tokens": 48242791.0,
      "step": 22025
    },
    {
      "entropy": 5.607951736450195,
      "epoch": 2.186383485510123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045243031821600794,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18511460721492767,
      "num_tokens": 48253683.0,
      "step": 22030
    },
    {
      "entropy": 5.6538414478302,
      "epoch": 2.1868797141722904,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004524085280123795,
      "loss": 5.0637,
      "mean_token_accuracy": 0.17579394429922104,
      "num_tokens": 48264681.0,
      "step": 22035
    },
    {
      "entropy": 5.6739747524261475,
      "epoch": 2.187375942834458,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004523867334094018,
      "loss": 5.064,
      "mean_token_accuracy": 0.18457618206739426,
      "num_tokens": 48275766.0,
      "step": 22040
    },
    {
      "entropy": 5.6848657608032225,
      "epoch": 2.187872171496626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004523649344076154,
      "loss": 5.1888,
      "mean_token_accuracy": 0.17721029222011567,
      "num_tokens": 48288273.0,
      "step": 22045
    },
    {
      "entropy": 5.566167640686035,
      "epoch": 2.1883684001587933,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045234313100756077,
      "loss": 4.9424,
      "mean_token_accuracy": 0.19660272896289827,
      "num_tokens": 48299668.0,
      "step": 22050
    },
    {
      "entropy": 5.63266167640686,
      "epoch": 2.188864628820961,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045232132320977854,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18445695489645003,
      "num_tokens": 48311296.0,
      "step": 22055
    },
    {
      "entropy": 5.592179250717163,
      "epoch": 2.1893608574831283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004522995110148097,
      "loss": 4.9937,
      "mean_token_accuracy": 0.1819663643836975,
      "num_tokens": 48321549.0,
      "step": 22060
    },
    {
      "entropy": 5.62314395904541,
      "epoch": 2.1898570861452957,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000452277694423195,
      "loss": 5.0749,
      "mean_token_accuracy": 0.17589035332202912,
      "num_tokens": 48334391.0,
      "step": 22065
    },
    {
      "entropy": 5.645931816101074,
      "epoch": 2.190353314807463,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004522558734354755,
      "loss": 5.0532,
      "mean_token_accuracy": 0.18434374779462814,
      "num_tokens": 48345102.0,
      "step": 22070
    },
    {
      "entropy": 5.681306791305542,
      "epoch": 2.1908495434696307,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045223404805219225,
      "loss": 5.0364,
      "mean_token_accuracy": 0.1735142722725868,
      "num_tokens": 48357175.0,
      "step": 22075
    },
    {
      "entropy": 5.649609613418579,
      "epoch": 2.191345772131798,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045221221827388655,
      "loss": 5.1144,
      "mean_token_accuracy": 0.18218374252319336,
      "num_tokens": 48369597.0,
      "step": 22080
    },
    {
      "entropy": 5.6597161293029785,
      "epoch": 2.191842000793966,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004521903841010997,
      "loss": 5.0503,
      "mean_token_accuracy": 0.18235548138618468,
      "num_tokens": 48380339.0,
      "step": 22085
    },
    {
      "entropy": 5.662203693389893,
      "epoch": 2.1923382294561335,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004521685455343731,
      "loss": 5.0514,
      "mean_token_accuracy": 0.18061643987894058,
      "num_tokens": 48393001.0,
      "step": 22090
    },
    {
      "entropy": 5.574957656860351,
      "epoch": 2.192834458118301,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045214670257424835,
      "loss": 5.0088,
      "mean_token_accuracy": 0.18802878856658936,
      "num_tokens": 48403089.0,
      "step": 22095
    },
    {
      "entropy": 5.679512929916382,
      "epoch": 2.1933306867804685,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004521248552212671,
      "loss": 5.0611,
      "mean_token_accuracy": 0.1804522082209587,
      "num_tokens": 48414017.0,
      "step": 22100
    },
    {
      "entropy": 5.661268472671509,
      "epoch": 2.193826915442636,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004521030034759711,
      "loss": 5.084,
      "mean_token_accuracy": 0.19138626754283905,
      "num_tokens": 48426812.0,
      "step": 22105
    },
    {
      "entropy": 5.535390949249267,
      "epoch": 2.1943231441048034,
      "grad_norm": 1.0,
      "learning_rate": 0.0004520811473389023,
      "loss": 4.9437,
      "mean_token_accuracy": 0.18634756207466124,
      "num_tokens": 48437373.0,
      "step": 22110
    },
    {
      "entropy": 5.600811767578125,
      "epoch": 2.194819372766971,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045205928681060255,
      "loss": 5.0088,
      "mean_token_accuracy": 0.1817712217569351,
      "num_tokens": 48448250.0,
      "step": 22115
    },
    {
      "entropy": 5.640335464477539,
      "epoch": 2.1953156014291384,
      "grad_norm": 1.109375,
      "learning_rate": 0.000452037421891614,
      "loss": 5.0774,
      "mean_token_accuracy": 0.17754901349544525,
      "num_tokens": 48459026.0,
      "step": 22120
    },
    {
      "entropy": 5.550931024551391,
      "epoch": 2.195811830091306,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045201555258247896,
      "loss": 4.9868,
      "mean_token_accuracy": 0.18458283245563506,
      "num_tokens": 48469319.0,
      "step": 22125
    },
    {
      "entropy": 5.612356758117675,
      "epoch": 2.196308058753474,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004519936788837396,
      "loss": 5.0801,
      "mean_token_accuracy": 0.18300363570451736,
      "num_tokens": 48479777.0,
      "step": 22130
    },
    {
      "entropy": 5.670968151092529,
      "epoch": 2.1968042874156413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045197180079593836,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17946938425302505,
      "num_tokens": 48491434.0,
      "step": 22135
    },
    {
      "entropy": 5.691042423248291,
      "epoch": 2.1973005160778087,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045194991831961777,
      "loss": 5.1341,
      "mean_token_accuracy": 0.17612102627754211,
      "num_tokens": 48502526.0,
      "step": 22140
    },
    {
      "entropy": 5.6510144710540775,
      "epoch": 2.197796744739976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004519280314553206,
      "loss": 5.0429,
      "mean_token_accuracy": 0.18047257214784623,
      "num_tokens": 48512897.0,
      "step": 22145
    },
    {
      "entropy": 5.644176959991455,
      "epoch": 2.1982929734021437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004519061402035894,
      "loss": 5.0148,
      "mean_token_accuracy": 0.17796958684921266,
      "num_tokens": 48523659.0,
      "step": 22150
    },
    {
      "entropy": 5.606931591033936,
      "epoch": 2.198789202064311,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045188424456496723,
      "loss": 5.0534,
      "mean_token_accuracy": 0.17892654240131378,
      "num_tokens": 48533899.0,
      "step": 22155
    },
    {
      "entropy": 5.57880482673645,
      "epoch": 2.1992854307264786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045186234453999693,
      "loss": 5.0421,
      "mean_token_accuracy": 0.1860852375626564,
      "num_tokens": 48545693.0,
      "step": 22160
    },
    {
      "entropy": 5.7078752517700195,
      "epoch": 2.199781659388646,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045184044012922154,
      "loss": 5.0399,
      "mean_token_accuracy": 0.17923077791929246,
      "num_tokens": 48556411.0,
      "step": 22165
    },
    {
      "entropy": 5.665908002853394,
      "epoch": 2.200277888050814,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045181853133318425,
      "loss": 5.0548,
      "mean_token_accuracy": 0.17667514234781265,
      "num_tokens": 48569385.0,
      "step": 22170
    },
    {
      "entropy": 5.703004789352417,
      "epoch": 2.2007741167129815,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004517966181524285,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1732465460896492,
      "num_tokens": 48580586.0,
      "step": 22175
    },
    {
      "entropy": 5.6407904624938965,
      "epoch": 2.201270345375149,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004517747005874975,
      "loss": 5.0272,
      "mean_token_accuracy": 0.17773015052080154,
      "num_tokens": 48589557.0,
      "step": 22180
    },
    {
      "entropy": 5.6478746891021725,
      "epoch": 2.2017665740373165,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004517527786389349,
      "loss": 5.0247,
      "mean_token_accuracy": 0.18182750046253204,
      "num_tokens": 48601880.0,
      "step": 22185
    },
    {
      "entropy": 5.612695932388306,
      "epoch": 2.202262802699484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045173085230728417,
      "loss": 5.0139,
      "mean_token_accuracy": 0.19008533656597137,
      "num_tokens": 48612469.0,
      "step": 22190
    },
    {
      "entropy": 5.659466791152954,
      "epoch": 2.2027590313616514,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004517089215930892,
      "loss": 5.1143,
      "mean_token_accuracy": 0.18250493258237838,
      "num_tokens": 48622852.0,
      "step": 22195
    },
    {
      "entropy": 5.676016807556152,
      "epoch": 2.203255260023819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045168698649689375,
      "loss": 5.0125,
      "mean_token_accuracy": 0.18265240788459777,
      "num_tokens": 48634062.0,
      "step": 22200
    },
    {
      "entropy": 5.640259504318237,
      "epoch": 2.2037514886859864,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045166504701924176,
      "loss": 4.9917,
      "mean_token_accuracy": 0.1777860477566719,
      "num_tokens": 48644538.0,
      "step": 22205
    },
    {
      "entropy": 5.644992399215698,
      "epoch": 2.204247717348154,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045164310316067727,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1815647691488266,
      "num_tokens": 48656053.0,
      "step": 22210
    },
    {
      "entropy": 5.628798151016236,
      "epoch": 2.2047439460103218,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004516211549217444,
      "loss": 5.0267,
      "mean_token_accuracy": 0.17714459151029588,
      "num_tokens": 48667846.0,
      "step": 22215
    },
    {
      "entropy": 5.638632583618164,
      "epoch": 2.2052401746724892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004515992023029875,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17752687633037567,
      "num_tokens": 48678087.0,
      "step": 22220
    },
    {
      "entropy": 5.526142120361328,
      "epoch": 2.2057364033346567,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000451577245304951,
      "loss": 4.9752,
      "mean_token_accuracy": 0.19560491144657136,
      "num_tokens": 48688727.0,
      "step": 22225
    },
    {
      "entropy": 5.542825365066529,
      "epoch": 2.206232631996824,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004515552839281792,
      "loss": 4.9085,
      "mean_token_accuracy": 0.18698004633188248,
      "num_tokens": 48700370.0,
      "step": 22230
    },
    {
      "entropy": 5.604220581054688,
      "epoch": 2.2067288606589917,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004515333181732169,
      "loss": 4.975,
      "mean_token_accuracy": 0.18607507199048995,
      "num_tokens": 48710988.0,
      "step": 22235
    },
    {
      "entropy": 5.6201331615448,
      "epoch": 2.207225089321159,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045151134804060863,
      "loss": 4.994,
      "mean_token_accuracy": 0.18344981521368026,
      "num_tokens": 48721133.0,
      "step": 22240
    },
    {
      "entropy": 5.625381851196289,
      "epoch": 2.2077213179833266,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004514893735308993,
      "loss": 4.969,
      "mean_token_accuracy": 0.18999962210655214,
      "num_tokens": 48732051.0,
      "step": 22245
    },
    {
      "entropy": 5.659356737136841,
      "epoch": 2.208217546645494,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004514673946446338,
      "loss": 5.0638,
      "mean_token_accuracy": 0.17934478372335433,
      "num_tokens": 48742847.0,
      "step": 22250
    },
    {
      "entropy": 5.650118732452393,
      "epoch": 2.208713775307662,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004514454113823572,
      "loss": 5.0569,
      "mean_token_accuracy": 0.17946675568819045,
      "num_tokens": 48754773.0,
      "step": 22255
    },
    {
      "entropy": 5.607219505310058,
      "epoch": 2.2092100039698295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004514234237446146,
      "loss": 5.0113,
      "mean_token_accuracy": 0.17866269797086715,
      "num_tokens": 48765387.0,
      "step": 22260
    },
    {
      "entropy": 5.57786717414856,
      "epoch": 2.209706232631997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045140143173195117,
      "loss": 4.9828,
      "mean_token_accuracy": 0.1901062473654747,
      "num_tokens": 48776532.0,
      "step": 22265
    },
    {
      "entropy": 5.655235815048218,
      "epoch": 2.2102024612941644,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004513794353449124,
      "loss": 5.1349,
      "mean_token_accuracy": 0.1788174867630005,
      "num_tokens": 48788690.0,
      "step": 22270
    },
    {
      "entropy": 5.636922216415405,
      "epoch": 2.210698689956332,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004513574345840437,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17941029518842697,
      "num_tokens": 48799375.0,
      "step": 22275
    },
    {
      "entropy": 5.685873889923096,
      "epoch": 2.2111949186184994,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004513354294498907,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17758168876171113,
      "num_tokens": 48810918.0,
      "step": 22280
    },
    {
      "entropy": 5.596605682373047,
      "epoch": 2.211691147280667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045131341994299895,
      "loss": 5.039,
      "mean_token_accuracy": 0.17952029258012772,
      "num_tokens": 48821225.0,
      "step": 22285
    },
    {
      "entropy": 5.6495081901550295,
      "epoch": 2.2121873759428343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004512914060639143,
      "loss": 5.0336,
      "mean_token_accuracy": 0.18804351538419722,
      "num_tokens": 48831457.0,
      "step": 22290
    },
    {
      "entropy": 5.600747060775757,
      "epoch": 2.212683604605002,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004512693878131827,
      "loss": 4.9653,
      "mean_token_accuracy": 0.19119712561368943,
      "num_tokens": 48841795.0,
      "step": 22295
    },
    {
      "entropy": 5.671405935287476,
      "epoch": 2.2131798332671697,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045124736519135003,
      "loss": 5.1078,
      "mean_token_accuracy": 0.1725558742880821,
      "num_tokens": 48852921.0,
      "step": 22300
    },
    {
      "entropy": 5.632822132110595,
      "epoch": 2.213676061929337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045122533819896257,
      "loss": 5.0203,
      "mean_token_accuracy": 0.1749219313263893,
      "num_tokens": 48863456.0,
      "step": 22305
    },
    {
      "entropy": 5.768363904953003,
      "epoch": 2.2141722905915047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004512033068365664,
      "loss": 5.1433,
      "mean_token_accuracy": 0.17843584418296815,
      "num_tokens": 48873295.0,
      "step": 22310
    },
    {
      "entropy": 5.634160137176513,
      "epoch": 2.214668519253672,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004511812711047079,
      "loss": 5.0061,
      "mean_token_accuracy": 0.17754214853048325,
      "num_tokens": 48884701.0,
      "step": 22315
    },
    {
      "entropy": 5.557489395141602,
      "epoch": 2.2151647479158396,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045115923100393356,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17914108335971832,
      "num_tokens": 48896393.0,
      "step": 22320
    },
    {
      "entropy": 5.589876651763916,
      "epoch": 2.215660976578007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045113718653478984,
      "loss": 5.0195,
      "mean_token_accuracy": 0.17684440761804582,
      "num_tokens": 48907710.0,
      "step": 22325
    },
    {
      "entropy": 5.6676544666290285,
      "epoch": 2.2161572052401746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004511151376978234,
      "loss": 5.0743,
      "mean_token_accuracy": 0.1745292291045189,
      "num_tokens": 48918665.0,
      "step": 22330
    },
    {
      "entropy": 5.61656346321106,
      "epoch": 2.216653433902342,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045109308449358114,
      "loss": 4.9824,
      "mean_token_accuracy": 0.18588371723890304,
      "num_tokens": 48931057.0,
      "step": 22335
    },
    {
      "entropy": 5.581650352478027,
      "epoch": 2.21714966256451,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004510710269226098,
      "loss": 5.0321,
      "mean_token_accuracy": 0.17384046018123628,
      "num_tokens": 48941470.0,
      "step": 22340
    },
    {
      "entropy": 5.556636571884155,
      "epoch": 2.2176458912266774,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004510489649854564,
      "loss": 4.9775,
      "mean_token_accuracy": 0.18445798009634018,
      "num_tokens": 48953278.0,
      "step": 22345
    },
    {
      "entropy": 5.586061096191406,
      "epoch": 2.218142119888845,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000451026898682668,
      "loss": 5.0212,
      "mean_token_accuracy": 0.1817585974931717,
      "num_tokens": 48964265.0,
      "step": 22350
    },
    {
      "entropy": 5.647521257400513,
      "epoch": 2.2186383485510124,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004510048280147918,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17588482648134232,
      "num_tokens": 48975307.0,
      "step": 22355
    },
    {
      "entropy": 5.655486965179444,
      "epoch": 2.21913457721318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004509827529823751,
      "loss": 5.0707,
      "mean_token_accuracy": 0.17881902307271957,
      "num_tokens": 48985770.0,
      "step": 22360
    },
    {
      "entropy": 5.589617919921875,
      "epoch": 2.2196308058753473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045096067358596536,
      "loss": 4.9637,
      "mean_token_accuracy": 0.18041966259479522,
      "num_tokens": 48996638.0,
      "step": 22365
    },
    {
      "entropy": 5.650857639312744,
      "epoch": 2.220127034537515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045093858982611006,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17653706967830657,
      "num_tokens": 49006974.0,
      "step": 22370
    },
    {
      "entropy": 5.57751054763794,
      "epoch": 2.2206232631996823,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004509165017033569,
      "loss": 4.9943,
      "mean_token_accuracy": 0.18197499066591263,
      "num_tokens": 49018344.0,
      "step": 22375
    },
    {
      "entropy": 5.659957695007324,
      "epoch": 2.2211194918618498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045089440921825354,
      "loss": 5.0725,
      "mean_token_accuracy": 0.18067340105772017,
      "num_tokens": 49028720.0,
      "step": 22380
    },
    {
      "entropy": 5.64381628036499,
      "epoch": 2.2216157205240172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045087231237134787,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1832860916852951,
      "num_tokens": 49038884.0,
      "step": 22385
    },
    {
      "entropy": 5.655926656723023,
      "epoch": 2.222111949186185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045085021116318783,
      "loss": 5.0272,
      "mean_token_accuracy": 0.18669366985559463,
      "num_tokens": 49050031.0,
      "step": 22390
    },
    {
      "entropy": 5.7237590789794925,
      "epoch": 2.2226081778483526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004508281055943214,
      "loss": 5.1393,
      "mean_token_accuracy": 0.1728743717074394,
      "num_tokens": 49060770.0,
      "step": 22395
    },
    {
      "entropy": 5.701142454147339,
      "epoch": 2.22310440651052,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004508059956652969,
      "loss": 5.0324,
      "mean_token_accuracy": 0.18361424952745437,
      "num_tokens": 49070621.0,
      "step": 22400
    },
    {
      "entropy": 5.643007278442383,
      "epoch": 2.2236006351726876,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004507838813766625,
      "loss": 5.0711,
      "mean_token_accuracy": 0.179632368683815,
      "num_tokens": 49080452.0,
      "step": 22405
    },
    {
      "entropy": 5.6532622337341305,
      "epoch": 2.224096863834855,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004507617627289666,
      "loss": 5.1038,
      "mean_token_accuracy": 0.18231165707111358,
      "num_tokens": 49091344.0,
      "step": 22410
    },
    {
      "entropy": 5.655296468734742,
      "epoch": 2.2245930924970225,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004507396397227577,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1749944880604744,
      "num_tokens": 49103590.0,
      "step": 22415
    },
    {
      "entropy": 5.590015125274658,
      "epoch": 2.22508932115919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045071751235858454,
      "loss": 4.9526,
      "mean_token_accuracy": 0.18348598033189772,
      "num_tokens": 49113923.0,
      "step": 22420
    },
    {
      "entropy": 5.7138457775115965,
      "epoch": 2.2255855498213575,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004506953806369956,
      "loss": 5.0893,
      "mean_token_accuracy": 0.17694988697767258,
      "num_tokens": 49124690.0,
      "step": 22425
    },
    {
      "entropy": 5.718528175354004,
      "epoch": 2.2260817784835254,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004506732445585399,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1740540623664856,
      "num_tokens": 49135546.0,
      "step": 22430
    },
    {
      "entropy": 5.66335620880127,
      "epoch": 2.226578007145693,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004506511041237662,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18825536221265793,
      "num_tokens": 49147446.0,
      "step": 22435
    },
    {
      "entropy": 5.547195482254028,
      "epoch": 2.2270742358078603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045062895933322366,
      "loss": 5.0002,
      "mean_token_accuracy": 0.17991036027669907,
      "num_tokens": 49157224.0,
      "step": 22440
    },
    {
      "entropy": 5.587775564193725,
      "epoch": 2.227570464470028,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004506068101874614,
      "loss": 5.0893,
      "mean_token_accuracy": 0.1757865935564041,
      "num_tokens": 49169961.0,
      "step": 22445
    },
    {
      "entropy": 5.69843373298645,
      "epoch": 2.2280666931321953,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045058465668702856,
      "loss": 5.0506,
      "mean_token_accuracy": 0.18165793269872665,
      "num_tokens": 49180486.0,
      "step": 22450
    },
    {
      "entropy": 5.63394513130188,
      "epoch": 2.2285629217943628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045056249883247457,
      "loss": 4.9833,
      "mean_token_accuracy": 0.17738423496484756,
      "num_tokens": 49191792.0,
      "step": 22455
    },
    {
      "entropy": 5.599758291244507,
      "epoch": 2.2290591504565302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045054033662434903,
      "loss": 5.0309,
      "mean_token_accuracy": 0.1877695396542549,
      "num_tokens": 49202268.0,
      "step": 22460
    },
    {
      "entropy": 5.630701017379761,
      "epoch": 2.2295553791186977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045051817006320127,
      "loss": 5.1143,
      "mean_token_accuracy": 0.17618466764688492,
      "num_tokens": 49212677.0,
      "step": 22465
    },
    {
      "entropy": 5.56439175605774,
      "epoch": 2.230051607780865,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045049599914958124,
      "loss": 5.0258,
      "mean_token_accuracy": 0.18051226884126664,
      "num_tokens": 49224543.0,
      "step": 22470
    },
    {
      "entropy": 5.596431541442871,
      "epoch": 2.230547836443033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004504738238840385,
      "loss": 5.026,
      "mean_token_accuracy": 0.17417484372854233,
      "num_tokens": 49235495.0,
      "step": 22475
    },
    {
      "entropy": 5.682494020462036,
      "epoch": 2.2310440651052006,
      "grad_norm": 1.125,
      "learning_rate": 0.0004504516442671231,
      "loss": 4.985,
      "mean_token_accuracy": 0.1864611864089966,
      "num_tokens": 49248378.0,
      "step": 22480
    },
    {
      "entropy": 5.5747222900390625,
      "epoch": 2.231540293767368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045042946029938495,
      "loss": 4.9443,
      "mean_token_accuracy": 0.18801382780075074,
      "num_tokens": 49258519.0,
      "step": 22485
    },
    {
      "entropy": 5.637068223953247,
      "epoch": 2.2320365224295355,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045040727198137416,
      "loss": 5.0692,
      "mean_token_accuracy": 0.17496583759784698,
      "num_tokens": 49268418.0,
      "step": 22490
    },
    {
      "entropy": 5.643234348297119,
      "epoch": 2.232532751091703,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045038507931364103,
      "loss": 5.065,
      "mean_token_accuracy": 0.178111532330513,
      "num_tokens": 49278290.0,
      "step": 22495
    },
    {
      "entropy": 5.6476686000823975,
      "epoch": 2.2330289797538705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045036288229673583,
      "loss": 5.0019,
      "mean_token_accuracy": 0.1775826930999756,
      "num_tokens": 49288523.0,
      "step": 22500
    },
    {
      "entropy": 5.6053393363952635,
      "epoch": 2.233525208416038,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045034068093120907,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1808995246887207,
      "num_tokens": 49299123.0,
      "step": 22505
    },
    {
      "entropy": 5.638334941864014,
      "epoch": 2.2340214370782054,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004503184752176112,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17958689332008362,
      "num_tokens": 49310674.0,
      "step": 22510
    },
    {
      "entropy": 5.674486875534058,
      "epoch": 2.2345176657403734,
      "grad_norm": 1.03125,
      "learning_rate": 0.000450296265156493,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1795735090970993,
      "num_tokens": 49322606.0,
      "step": 22515
    },
    {
      "entropy": 5.612466049194336,
      "epoch": 2.235013894402541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045027405074840506,
      "loss": 4.9861,
      "mean_token_accuracy": 0.18990458250045777,
      "num_tokens": 49332817.0,
      "step": 22520
    },
    {
      "entropy": 5.595915460586548,
      "epoch": 2.2355101230647083,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004502518319938984,
      "loss": 5.0382,
      "mean_token_accuracy": 0.17822938412427902,
      "num_tokens": 49344276.0,
      "step": 22525
    },
    {
      "entropy": 5.647004652023315,
      "epoch": 2.236006351726876,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045022960889352394,
      "loss": 5.0617,
      "mean_token_accuracy": 0.18050379157066346,
      "num_tokens": 49354179.0,
      "step": 22530
    },
    {
      "entropy": 5.725394630432129,
      "epoch": 2.2365025803890433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045020738144783273,
      "loss": 5.0969,
      "mean_token_accuracy": 0.1813880294561386,
      "num_tokens": 49364940.0,
      "step": 22535
    },
    {
      "entropy": 5.633791303634643,
      "epoch": 2.2369988090512107,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004501851496573761,
      "loss": 5.0587,
      "mean_token_accuracy": 0.18153682798147203,
      "num_tokens": 49374992.0,
      "step": 22540
    },
    {
      "entropy": 5.618936824798584,
      "epoch": 2.237495037713378,
      "grad_norm": 1.0,
      "learning_rate": 0.0004501629135227051,
      "loss": 5.0105,
      "mean_token_accuracy": 0.17463302463293076,
      "num_tokens": 49386521.0,
      "step": 22545
    },
    {
      "entropy": 5.5586323738098145,
      "epoch": 2.2379912663755457,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004501406730443714,
      "loss": 4.9326,
      "mean_token_accuracy": 0.1845379054546356,
      "num_tokens": 49396246.0,
      "step": 22550
    },
    {
      "entropy": 5.570860242843628,
      "epoch": 2.238487495037713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045011842822292636,
      "loss": 4.953,
      "mean_token_accuracy": 0.1822250172495842,
      "num_tokens": 49406158.0,
      "step": 22555
    },
    {
      "entropy": 5.685667896270752,
      "epoch": 2.238983723699881,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004500961790589216,
      "loss": 5.085,
      "mean_token_accuracy": 0.18150988221168518,
      "num_tokens": 49418189.0,
      "step": 22560
    },
    {
      "entropy": 5.651168012619019,
      "epoch": 2.2394799523620486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000450073925552909,
      "loss": 5.0533,
      "mean_token_accuracy": 0.18459042459726333,
      "num_tokens": 49430385.0,
      "step": 22565
    },
    {
      "entropy": 5.652307558059692,
      "epoch": 2.239976181024216,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004500516677054403,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18383918553590775,
      "num_tokens": 49440279.0,
      "step": 22570
    },
    {
      "entropy": 5.653861045837402,
      "epoch": 2.2404724096863835,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045002940551706726,
      "loss": 5.0692,
      "mean_token_accuracy": 0.17891368567943572,
      "num_tokens": 49450895.0,
      "step": 22575
    },
    {
      "entropy": 5.660978698730469,
      "epoch": 2.240968638348551,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045000713898834225,
      "loss": 5.0873,
      "mean_token_accuracy": 0.17524576038122178,
      "num_tokens": 49462113.0,
      "step": 22580
    },
    {
      "entropy": 5.653906393051147,
      "epoch": 2.2414648670107185,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004499848681198173,
      "loss": 5.0669,
      "mean_token_accuracy": 0.1767621397972107,
      "num_tokens": 49473972.0,
      "step": 22585
    },
    {
      "entropy": 5.671089553833008,
      "epoch": 2.241961095672886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004499625929120447,
      "loss": 5.0505,
      "mean_token_accuracy": 0.18103896975517272,
      "num_tokens": 49484706.0,
      "step": 22590
    },
    {
      "entropy": 5.682682657241822,
      "epoch": 2.2424573243350534,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004499403133655768,
      "loss": 5.0995,
      "mean_token_accuracy": 0.1757372036576271,
      "num_tokens": 49496732.0,
      "step": 22595
    },
    {
      "entropy": 5.565688705444336,
      "epoch": 2.2429535529972213,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004499180294809661,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18198151737451554,
      "num_tokens": 49508615.0,
      "step": 22600
    },
    {
      "entropy": 5.619434452056884,
      "epoch": 2.243449781659389,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044989574125876516,
      "loss": 5.0139,
      "mean_token_accuracy": 0.17980944812297822,
      "num_tokens": 49520081.0,
      "step": 22605
    },
    {
      "entropy": 5.606912612915039,
      "epoch": 2.2439460103215563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004498734486995267,
      "loss": 4.9399,
      "mean_token_accuracy": 0.18589258342981338,
      "num_tokens": 49531537.0,
      "step": 22610
    },
    {
      "entropy": 5.570858478546143,
      "epoch": 2.2444422389837237,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044985115180380355,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18593387305736542,
      "num_tokens": 49541829.0,
      "step": 22615
    },
    {
      "entropy": 5.5832202434539795,
      "epoch": 2.2449384676458912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044982885057214857,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18275251239538193,
      "num_tokens": 49554056.0,
      "step": 22620
    },
    {
      "entropy": 5.676235055923462,
      "epoch": 2.2454346963080587,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044980654500511485,
      "loss": 5.0633,
      "mean_token_accuracy": 0.17623039335012436,
      "num_tokens": 49565427.0,
      "step": 22625
    },
    {
      "entropy": 5.661865854263306,
      "epoch": 2.245930924970226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044978423510325544,
      "loss": 5.0337,
      "mean_token_accuracy": 0.18324103057384492,
      "num_tokens": 49574903.0,
      "step": 22630
    },
    {
      "entropy": 5.664002323150635,
      "epoch": 2.2464271536323936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004497619208671237,
      "loss": 5.1104,
      "mean_token_accuracy": 0.18109564930200578,
      "num_tokens": 49586721.0,
      "step": 22635
    },
    {
      "entropy": 5.660956954956054,
      "epoch": 2.246923382294561,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004497396022972728,
      "loss": 5.0899,
      "mean_token_accuracy": 0.18515109568834304,
      "num_tokens": 49597212.0,
      "step": 22640
    },
    {
      "entropy": 5.655280923843383,
      "epoch": 2.247419610956729,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004497172793942564,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18177321553230286,
      "num_tokens": 49608235.0,
      "step": 22645
    },
    {
      "entropy": 5.569576263427734,
      "epoch": 2.2479158396188965,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004496949521586279,
      "loss": 4.9871,
      "mean_token_accuracy": 0.1865206241607666,
      "num_tokens": 49618460.0,
      "step": 22650
    },
    {
      "entropy": 5.608824348449707,
      "epoch": 2.248412068281064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044967262059094097,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1836908757686615,
      "num_tokens": 49629704.0,
      "step": 22655
    },
    {
      "entropy": 5.600371313095093,
      "epoch": 2.2489082969432315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044965028469174956,
      "loss": 5.0476,
      "mean_token_accuracy": 0.1775353267788887,
      "num_tokens": 49640307.0,
      "step": 22660
    },
    {
      "entropy": 5.628655290603637,
      "epoch": 2.249404525605399,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044962794446160736,
      "loss": 5.0721,
      "mean_token_accuracy": 0.1736655652523041,
      "num_tokens": 49651119.0,
      "step": 22665
    },
    {
      "entropy": 5.687271976470948,
      "epoch": 2.2499007542675664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044960559990106844,
      "loss": 5.0553,
      "mean_token_accuracy": 0.17793308794498444,
      "num_tokens": 49661822.0,
      "step": 22670
    },
    {
      "entropy": 5.737827444076538,
      "epoch": 2.250396982929734,
      "grad_norm": 1.0,
      "learning_rate": 0.0004495832510106869,
      "loss": 5.0812,
      "mean_token_accuracy": 0.18286984264850617,
      "num_tokens": 49672864.0,
      "step": 22675
    },
    {
      "entropy": 5.603489351272583,
      "epoch": 2.2508932115919014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004495608977910169,
      "loss": 5.0398,
      "mean_token_accuracy": 0.1863543063402176,
      "num_tokens": 49683570.0,
      "step": 22680
    },
    {
      "entropy": 5.614112472534179,
      "epoch": 2.2513894402540693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004495385402426128,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18808760344982148,
      "num_tokens": 49693297.0,
      "step": 22685
    },
    {
      "entropy": 5.597784805297851,
      "epoch": 2.2518856689162368,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000449516178366029,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18556829690933227,
      "num_tokens": 49705506.0,
      "step": 22690
    },
    {
      "entropy": 5.656980276107788,
      "epoch": 2.2523818975784042,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044949381216182,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1785978838801384,
      "num_tokens": 49715594.0,
      "step": 22695
    },
    {
      "entropy": 5.663366889953613,
      "epoch": 2.2528781262405717,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044947144163054036,
      "loss": 5.055,
      "mean_token_accuracy": 0.18171602189540864,
      "num_tokens": 49725355.0,
      "step": 22700
    },
    {
      "entropy": 5.664501190185547,
      "epoch": 2.253374354902739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004494490667727451,
      "loss": 5.0568,
      "mean_token_accuracy": 0.1742839515209198,
      "num_tokens": 49735912.0,
      "step": 22705
    },
    {
      "entropy": 5.646450996398926,
      "epoch": 2.2538705835649067,
      "grad_norm": 1.125,
      "learning_rate": 0.00044942668758898873,
      "loss": 5.0996,
      "mean_token_accuracy": 0.18638602644205093,
      "num_tokens": 49746807.0,
      "step": 22710
    },
    {
      "entropy": 5.658433341979981,
      "epoch": 2.254366812227074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004494043040798265,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17436965107917785,
      "num_tokens": 49757505.0,
      "step": 22715
    },
    {
      "entropy": 5.683465003967285,
      "epoch": 2.2548630408892416,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004493819162458133,
      "loss": 5.1288,
      "mean_token_accuracy": 0.17816347777843475,
      "num_tokens": 49768689.0,
      "step": 22720
    },
    {
      "entropy": 5.709576272964478,
      "epoch": 2.255359269551409,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493595240875043,
      "loss": 5.1076,
      "mean_token_accuracy": 0.17020599246025087,
      "num_tokens": 49781069.0,
      "step": 22725
    },
    {
      "entropy": 5.660921573638916,
      "epoch": 2.255855498213577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044933712760545486,
      "loss": 5.0453,
      "mean_token_accuracy": 0.18135172128677368,
      "num_tokens": 49792820.0,
      "step": 22730
    },
    {
      "entropy": 5.630942916870117,
      "epoch": 2.2563517268757445,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004493147268002203,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18361211568117142,
      "num_tokens": 49802640.0,
      "step": 22735
    },
    {
      "entropy": 5.635715579986572,
      "epoch": 2.256847955537912,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044929232167235613,
      "loss": 5.0965,
      "mean_token_accuracy": 0.1782770723104477,
      "num_tokens": 49813578.0,
      "step": 22740
    },
    {
      "entropy": 5.6044210433959964,
      "epoch": 2.2573441842000794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004492699122224179,
      "loss": 4.9961,
      "mean_token_accuracy": 0.1870172455906868,
      "num_tokens": 49824289.0,
      "step": 22745
    },
    {
      "entropy": 5.66929988861084,
      "epoch": 2.257840412862247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044924749845096144,
      "loss": 4.9992,
      "mean_token_accuracy": 0.1828078880906105,
      "num_tokens": 49835235.0,
      "step": 22750
    },
    {
      "entropy": 5.631571197509766,
      "epoch": 2.2583366415244144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004492250803585425,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17662792801856994,
      "num_tokens": 49846266.0,
      "step": 22755
    },
    {
      "entropy": 5.5911863327026365,
      "epoch": 2.258832870186582,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004492026579457169,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18364065885543823,
      "num_tokens": 49856642.0,
      "step": 22760
    },
    {
      "entropy": 5.738644075393677,
      "epoch": 2.2593290988487493,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004491802312130408,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16899939626455307,
      "num_tokens": 49868170.0,
      "step": 22765
    },
    {
      "entropy": 5.752908945083618,
      "epoch": 2.2598253275109172,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004491578001610703,
      "loss": 5.0779,
      "mean_token_accuracy": 0.18114236295223235,
      "num_tokens": 49878485.0,
      "step": 22770
    },
    {
      "entropy": 5.674929237365722,
      "epoch": 2.2603215561730847,
      "grad_norm": 1.0,
      "learning_rate": 0.00044913536479036163,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1797547996044159,
      "num_tokens": 49889564.0,
      "step": 22775
    },
    {
      "entropy": 5.6196213722229,
      "epoch": 2.260817784835252,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044911292510147117,
      "loss": 5.0818,
      "mean_token_accuracy": 0.1787710055708885,
      "num_tokens": 49901081.0,
      "step": 22780
    },
    {
      "entropy": 5.639882373809814,
      "epoch": 2.2613140134974197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004490904810949553,
      "loss": 5.0367,
      "mean_token_accuracy": 0.17453390657901763,
      "num_tokens": 49910917.0,
      "step": 22785
    },
    {
      "entropy": 5.713014793395996,
      "epoch": 2.261810242159587,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044906803277137064,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17233809679746628,
      "num_tokens": 49923793.0,
      "step": 22790
    },
    {
      "entropy": 5.691460418701172,
      "epoch": 2.2623064708217546,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004490455801312738,
      "loss": 5.0698,
      "mean_token_accuracy": 0.1805021956562996,
      "num_tokens": 49934989.0,
      "step": 22795
    },
    {
      "entropy": 5.666776180267334,
      "epoch": 2.262802699483922,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004490231231752217,
      "loss": 5.1718,
      "mean_token_accuracy": 0.16782320886850358,
      "num_tokens": 49945961.0,
      "step": 22800
    },
    {
      "entropy": 5.7051331996917725,
      "epoch": 2.2632989281460896,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004490006619037711,
      "loss": 5.1005,
      "mean_token_accuracy": 0.17745893597602844,
      "num_tokens": 49958200.0,
      "step": 22805
    },
    {
      "entropy": 5.695048141479492,
      "epoch": 2.263795156808257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044897819631747903,
      "loss": 5.0994,
      "mean_token_accuracy": 0.17970169335603714,
      "num_tokens": 49969962.0,
      "step": 22810
    },
    {
      "entropy": 5.5964111328125,
      "epoch": 2.2642913854704245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044895572641690256,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1795100674033165,
      "num_tokens": 49979855.0,
      "step": 22815
    },
    {
      "entropy": 5.677940511703492,
      "epoch": 2.2647876141325924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000448933252202599,
      "loss": 5.1652,
      "mean_token_accuracy": 0.17826617360115052,
      "num_tokens": 49990905.0,
      "step": 22820
    },
    {
      "entropy": 5.671911334991455,
      "epoch": 2.26528384279476,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044891077367512544,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1785399302840233,
      "num_tokens": 50003188.0,
      "step": 22825
    },
    {
      "entropy": 5.648052406311035,
      "epoch": 2.2657800714569274,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004488882908350396,
      "loss": 5.0053,
      "mean_token_accuracy": 0.17793249785900117,
      "num_tokens": 50014538.0,
      "step": 22830
    },
    {
      "entropy": 5.631344366073608,
      "epoch": 2.266276300119095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004488658036828987,
      "loss": 5.0426,
      "mean_token_accuracy": 0.17478182166814804,
      "num_tokens": 50025899.0,
      "step": 22835
    },
    {
      "entropy": 5.595417642593384,
      "epoch": 2.2667725287812623,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004488433122192606,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1749619036912918,
      "num_tokens": 50036828.0,
      "step": 22840
    },
    {
      "entropy": 5.61637282371521,
      "epoch": 2.26726875744343,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044882081644468286,
      "loss": 5.0713,
      "mean_token_accuracy": 0.18536552786827087,
      "num_tokens": 50047166.0,
      "step": 22845
    },
    {
      "entropy": 5.615932941436768,
      "epoch": 2.2677649861055973,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044879831635972345,
      "loss": 5.0287,
      "mean_token_accuracy": 0.18579777032136918,
      "num_tokens": 50058635.0,
      "step": 22850
    },
    {
      "entropy": 5.607107734680175,
      "epoch": 2.268261214767765,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004487758119649403,
      "loss": 4.9365,
      "mean_token_accuracy": 0.18514069467782973,
      "num_tokens": 50068304.0,
      "step": 22855
    },
    {
      "entropy": 5.591134309768677,
      "epoch": 2.2687574434299327,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044875330326089153,
      "loss": 5.0138,
      "mean_token_accuracy": 0.18633233159780502,
      "num_tokens": 50079785.0,
      "step": 22860
    },
    {
      "entropy": 5.658198070526123,
      "epoch": 2.2692536720921,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004487307902481352,
      "loss": 5.0649,
      "mean_token_accuracy": 0.18379786163568496,
      "num_tokens": 50089637.0,
      "step": 22865
    },
    {
      "entropy": 5.628876686096191,
      "epoch": 2.2697499007542676,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044870827292722967,
      "loss": 5.0052,
      "mean_token_accuracy": 0.1810275912284851,
      "num_tokens": 50101447.0,
      "step": 22870
    },
    {
      "entropy": 5.655264520645142,
      "epoch": 2.270246129416435,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004486857512987333,
      "loss": 5.0513,
      "mean_token_accuracy": 0.17544603496789932,
      "num_tokens": 50111488.0,
      "step": 22875
    },
    {
      "entropy": 5.605441045761109,
      "epoch": 2.2707423580786026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044866322536320447,
      "loss": 5.0345,
      "mean_token_accuracy": 0.183256658911705,
      "num_tokens": 50121696.0,
      "step": 22880
    },
    {
      "entropy": 5.702871990203858,
      "epoch": 2.27123858674077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044864069512120185,
      "loss": 5.2143,
      "mean_token_accuracy": 0.16851796060800553,
      "num_tokens": 50133106.0,
      "step": 22885
    },
    {
      "entropy": 5.678193140029907,
      "epoch": 2.2717348154029375,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044861816057328424,
      "loss": 5.0198,
      "mean_token_accuracy": 0.18079519867897034,
      "num_tokens": 50143670.0,
      "step": 22890
    },
    {
      "entropy": 5.6064098358154295,
      "epoch": 2.272231044065105,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044859562172001036,
      "loss": 5.0076,
      "mean_token_accuracy": 0.18824035972356795,
      "num_tokens": 50155475.0,
      "step": 22895
    },
    {
      "entropy": 5.580458068847657,
      "epoch": 2.2727272727272725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004485730785619391,
      "loss": 4.9734,
      "mean_token_accuracy": 0.17780425101518632,
      "num_tokens": 50165830.0,
      "step": 22900
    },
    {
      "entropy": 5.5814472198486325,
      "epoch": 2.2732235013894404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004485505310996295,
      "loss": 4.967,
      "mean_token_accuracy": 0.1912178471684456,
      "num_tokens": 50176101.0,
      "step": 22905
    },
    {
      "entropy": 5.61358609199524,
      "epoch": 2.273719730051608,
      "grad_norm": 1.0,
      "learning_rate": 0.0004485279793336406,
      "loss": 5.0278,
      "mean_token_accuracy": 0.18315733969211578,
      "num_tokens": 50187021.0,
      "step": 22910
    },
    {
      "entropy": 5.581608819961548,
      "epoch": 2.2742159587137754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004485054232645319,
      "loss": 5.0139,
      "mean_token_accuracy": 0.18178930282592773,
      "num_tokens": 50198646.0,
      "step": 22915
    },
    {
      "entropy": 5.600338554382324,
      "epoch": 2.274712187375943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004484828628928625,
      "loss": 5.0014,
      "mean_token_accuracy": 0.18466790169477462,
      "num_tokens": 50210088.0,
      "step": 22920
    },
    {
      "entropy": 5.5951892852783205,
      "epoch": 2.2752084160381103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044846029821919193,
      "loss": 4.9787,
      "mean_token_accuracy": 0.18616720736026765,
      "num_tokens": 50221188.0,
      "step": 22925
    },
    {
      "entropy": 5.5767316818237305,
      "epoch": 2.275704644700278,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004484377292440797,
      "loss": 5.0198,
      "mean_token_accuracy": 0.17823280692100524,
      "num_tokens": 50233225.0,
      "step": 22930
    },
    {
      "entropy": 5.639154767990112,
      "epoch": 2.2762008733624453,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044841515596808553,
      "loss": 4.9967,
      "mean_token_accuracy": 0.179875847697258,
      "num_tokens": 50244324.0,
      "step": 22935
    },
    {
      "entropy": 5.653833436965942,
      "epoch": 2.276697102024613,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044839257839176914,
      "loss": 4.895,
      "mean_token_accuracy": 0.1890880808234215,
      "num_tokens": 50254060.0,
      "step": 22940
    },
    {
      "entropy": 5.68988528251648,
      "epoch": 2.2771933306867806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044836999651569055,
      "loss": 5.1599,
      "mean_token_accuracy": 0.17315447330474854,
      "num_tokens": 50264165.0,
      "step": 22945
    },
    {
      "entropy": 5.619876527786255,
      "epoch": 2.277689559348948,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004483474103404095,
      "loss": 5.058,
      "mean_token_accuracy": 0.1803364247083664,
      "num_tokens": 50274751.0,
      "step": 22950
    },
    {
      "entropy": 5.670982599258423,
      "epoch": 2.2781857880111156,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044832481986648633,
      "loss": 5.1543,
      "mean_token_accuracy": 0.17853597700595855,
      "num_tokens": 50286992.0,
      "step": 22955
    },
    {
      "entropy": 5.753789663314819,
      "epoch": 2.278682016673283,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000448302225094481,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1715932160615921,
      "num_tokens": 50298462.0,
      "step": 22960
    },
    {
      "entropy": 5.674284219741821,
      "epoch": 2.2791782453354505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044827962602495393,
      "loss": 5.0291,
      "mean_token_accuracy": 0.17296806573867798,
      "num_tokens": 50309717.0,
      "step": 22965
    },
    {
      "entropy": 5.647585916519165,
      "epoch": 2.279674473997618,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044825702265846557,
      "loss": 4.9625,
      "mean_token_accuracy": 0.178065924346447,
      "num_tokens": 50320334.0,
      "step": 22970
    },
    {
      "entropy": 5.61231255531311,
      "epoch": 2.2801707026597855,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004482344149955763,
      "loss": 5.0368,
      "mean_token_accuracy": 0.18500587046146394,
      "num_tokens": 50332028.0,
      "step": 22975
    },
    {
      "entropy": 5.582992124557495,
      "epoch": 2.280666931321953,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004482118030368469,
      "loss": 4.9493,
      "mean_token_accuracy": 0.18492230474948884,
      "num_tokens": 50343673.0,
      "step": 22980
    },
    {
      "entropy": 5.642216491699219,
      "epoch": 2.2811631599841204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000448189186782838,
      "loss": 5.0746,
      "mean_token_accuracy": 0.16886264383792876,
      "num_tokens": 50353542.0,
      "step": 22985
    },
    {
      "entropy": 5.616021537780762,
      "epoch": 2.2816593886462884,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044816656623411044,
      "loss": 4.9933,
      "mean_token_accuracy": 0.18061961680650712,
      "num_tokens": 50365317.0,
      "step": 22990
    },
    {
      "entropy": 5.632167530059815,
      "epoch": 2.282155617308456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004481439413912252,
      "loss": 4.9868,
      "mean_token_accuracy": 0.1899981364607811,
      "num_tokens": 50376105.0,
      "step": 22995
    },
    {
      "entropy": 5.611277151107788,
      "epoch": 2.2826518459706233,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044812131225474333,
      "loss": 5.0463,
      "mean_token_accuracy": 0.17848762422800063,
      "num_tokens": 50385854.0,
      "step": 23000
    },
    {
      "entropy": 5.611650371551514,
      "epoch": 2.283148074632791,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004480986788252259,
      "loss": 5.0446,
      "mean_token_accuracy": 0.17528014928102492,
      "num_tokens": 50396869.0,
      "step": 23005
    },
    {
      "entropy": 5.622700929641724,
      "epoch": 2.2836443032949583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004480760411032343,
      "loss": 4.977,
      "mean_token_accuracy": 0.18916340917348862,
      "num_tokens": 50407620.0,
      "step": 23010
    },
    {
      "entropy": 5.62038950920105,
      "epoch": 2.2841405319571257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004480533990893298,
      "loss": 4.9887,
      "mean_token_accuracy": 0.18542582392692566,
      "num_tokens": 50418401.0,
      "step": 23015
    },
    {
      "entropy": 5.610281324386596,
      "epoch": 2.284636760619293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004480307527840739,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1821703553199768,
      "num_tokens": 50429835.0,
      "step": 23020
    },
    {
      "entropy": 5.692525911331177,
      "epoch": 2.285132989281461,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480081021880282,
      "loss": 5.1029,
      "mean_token_accuracy": 0.1811493754386902,
      "num_tokens": 50440072.0,
      "step": 23025
    },
    {
      "entropy": 5.6768876075744625,
      "epoch": 2.2856292179436286,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004479854473017543,
      "loss": 5.0578,
      "mean_token_accuracy": 0.17643370032310485,
      "num_tokens": 50450829.0,
      "step": 23030
    },
    {
      "entropy": 5.656885385513306,
      "epoch": 2.286125446605796,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044796278812581417,
      "loss": 5.0071,
      "mean_token_accuracy": 0.18661701083183288,
      "num_tokens": 50460913.0,
      "step": 23035
    },
    {
      "entropy": 5.561117267608642,
      "epoch": 2.2866216752679636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004479401246607695,
      "loss": 4.9626,
      "mean_token_accuracy": 0.18416867554187774,
      "num_tokens": 50473140.0,
      "step": 23040
    },
    {
      "entropy": 5.671007061004639,
      "epoch": 2.287117903930131,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004479174569071823,
      "loss": 5.0879,
      "mean_token_accuracy": 0.18275498002767562,
      "num_tokens": 50483147.0,
      "step": 23045
    },
    {
      "entropy": 5.610121154785157,
      "epoch": 2.2876141325922985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004478947848656149,
      "loss": 4.9373,
      "mean_token_accuracy": 0.18959422111511232,
      "num_tokens": 50493955.0,
      "step": 23050
    },
    {
      "entropy": 5.6570597171783445,
      "epoch": 2.288110361254466,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044787210853662943,
      "loss": 5.0428,
      "mean_token_accuracy": 0.1744605779647827,
      "num_tokens": 50505143.0,
      "step": 23055
    },
    {
      "entropy": 5.689381074905396,
      "epoch": 2.2886065899166335,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004478494279207881,
      "loss": 5.0546,
      "mean_token_accuracy": 0.18348232805728912,
      "num_tokens": 50515515.0,
      "step": 23060
    },
    {
      "entropy": 5.67322268486023,
      "epoch": 2.289102818578801,
      "grad_norm": 1.0,
      "learning_rate": 0.00044782674301865346,
      "loss": 5.0498,
      "mean_token_accuracy": 0.17504405379295349,
      "num_tokens": 50525046.0,
      "step": 23065
    },
    {
      "entropy": 5.574195146560669,
      "epoch": 2.2895990472409684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004478040538307879,
      "loss": 4.9501,
      "mean_token_accuracy": 0.18937128335237502,
      "num_tokens": 50535906.0,
      "step": 23070
    },
    {
      "entropy": 5.632741832733155,
      "epoch": 2.2900952759031363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004477813603577543,
      "loss": 5.0637,
      "mean_token_accuracy": 0.17864498496055603,
      "num_tokens": 50546791.0,
      "step": 23075
    },
    {
      "entropy": 5.688953638076782,
      "epoch": 2.290591504565304,
      "grad_norm": 1.0,
      "learning_rate": 0.00044775866260011516,
      "loss": 5.0265,
      "mean_token_accuracy": 0.1786270782351494,
      "num_tokens": 50558150.0,
      "step": 23080
    },
    {
      "entropy": 5.687194299697876,
      "epoch": 2.2910877332274713,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004477359605584335,
      "loss": 5.0845,
      "mean_token_accuracy": 0.1761659249663353,
      "num_tokens": 50569030.0,
      "step": 23085
    },
    {
      "entropy": 5.63923749923706,
      "epoch": 2.2915839618896388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044771325423327227,
      "loss": 5.016,
      "mean_token_accuracy": 0.1913376748561859,
      "num_tokens": 50579872.0,
      "step": 23090
    },
    {
      "entropy": 5.691687154769897,
      "epoch": 2.2920801905518062,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004476905436251945,
      "loss": 5.0741,
      "mean_token_accuracy": 0.17739618122577666,
      "num_tokens": 50591389.0,
      "step": 23095
    },
    {
      "entropy": 5.653619956970215,
      "epoch": 2.2925764192139737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044766782873476333,
      "loss": 5.1091,
      "mean_token_accuracy": 0.18109946548938752,
      "num_tokens": 50603732.0,
      "step": 23100
    },
    {
      "entropy": 5.6425727844238285,
      "epoch": 2.293072647876141,
      "grad_norm": 1.03125,
      "learning_rate": 0.000447645109562542,
      "loss": 5.014,
      "mean_token_accuracy": 0.18398984670639038,
      "num_tokens": 50614950.0,
      "step": 23105
    },
    {
      "entropy": 5.605583000183105,
      "epoch": 2.293568876538309,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044762238610909414,
      "loss": 5.0135,
      "mean_token_accuracy": 0.17965713888406754,
      "num_tokens": 50625387.0,
      "step": 23110
    },
    {
      "entropy": 5.709899950027466,
      "epoch": 2.2940651052004766,
      "grad_norm": 1.0,
      "learning_rate": 0.0004475996583749829,
      "loss": 5.0807,
      "mean_token_accuracy": 0.1764488473534584,
      "num_tokens": 50637017.0,
      "step": 23115
    },
    {
      "entropy": 5.621318244934082,
      "epoch": 2.294561333862644,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044757692636077215,
      "loss": 5.0147,
      "mean_token_accuracy": 0.1825342133641243,
      "num_tokens": 50647291.0,
      "step": 23120
    },
    {
      "entropy": 5.751184892654419,
      "epoch": 2.2950575625248115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044755419006702545,
      "loss": 5.1312,
      "mean_token_accuracy": 0.1774738311767578,
      "num_tokens": 50659750.0,
      "step": 23125
    },
    {
      "entropy": 5.740788507461548,
      "epoch": 2.295553791186979,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044753144949430667,
      "loss": 5.0607,
      "mean_token_accuracy": 0.18036912977695466,
      "num_tokens": 50671489.0,
      "step": 23130
    },
    {
      "entropy": 5.638572025299072,
      "epoch": 2.2960500198491465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004475087046431798,
      "loss": 5.0221,
      "mean_token_accuracy": 0.1900410383939743,
      "num_tokens": 50682377.0,
      "step": 23135
    },
    {
      "entropy": 5.618973588943481,
      "epoch": 2.296546248511314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004474859555142087,
      "loss": 5.0078,
      "mean_token_accuracy": 0.1848379224538803,
      "num_tokens": 50693502.0,
      "step": 23140
    },
    {
      "entropy": 5.646398973464966,
      "epoch": 2.2970424771734814,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044746320210795753,
      "loss": 5.032,
      "mean_token_accuracy": 0.17861706614494324,
      "num_tokens": 50704182.0,
      "step": 23145
    },
    {
      "entropy": 5.640265226364136,
      "epoch": 2.297538705835649,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004474404444249906,
      "loss": 4.9595,
      "mean_token_accuracy": 0.1842410832643509,
      "num_tokens": 50714209.0,
      "step": 23150
    },
    {
      "entropy": 5.59399299621582,
      "epoch": 2.2980349344978164,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044741768246587226,
      "loss": 4.9864,
      "mean_token_accuracy": 0.1839508518576622,
      "num_tokens": 50724862.0,
      "step": 23155
    },
    {
      "entropy": 5.683650016784668,
      "epoch": 2.2985311631599843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004473949162311669,
      "loss": 5.0843,
      "mean_token_accuracy": 0.17836710810661316,
      "num_tokens": 50735009.0,
      "step": 23160
    },
    {
      "entropy": 5.582537889480591,
      "epoch": 2.2990273918221518,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004473721457214391,
      "loss": 4.976,
      "mean_token_accuracy": 0.1829600602388382,
      "num_tokens": 50745883.0,
      "step": 23165
    },
    {
      "entropy": 5.6242499351501465,
      "epoch": 2.2995236204843192,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004473493709372535,
      "loss": 5.0455,
      "mean_token_accuracy": 0.1826542943716049,
      "num_tokens": 50755341.0,
      "step": 23170
    },
    {
      "entropy": 5.608227014541626,
      "epoch": 2.3000198491464867,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473265918791749,
      "loss": 4.9661,
      "mean_token_accuracy": 0.18440122306346893,
      "num_tokens": 50765891.0,
      "step": 23175
    },
    {
      "entropy": 5.693029928207397,
      "epoch": 2.300516077808654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473038085477682,
      "loss": 5.0619,
      "mean_token_accuracy": 0.18068478405475616,
      "num_tokens": 50777228.0,
      "step": 23180
    },
    {
      "entropy": 5.6406303405761715,
      "epoch": 2.3010123064708217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044728102094359826,
      "loss": 5.1074,
      "mean_token_accuracy": 0.17495658099651337,
      "num_tokens": 50787334.0,
      "step": 23185
    },
    {
      "entropy": 5.63738226890564,
      "epoch": 2.301508535132989,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004472582290672303,
      "loss": 5.0192,
      "mean_token_accuracy": 0.17729806900024414,
      "num_tokens": 50798433.0,
      "step": 23190
    },
    {
      "entropy": 5.623672199249268,
      "epoch": 2.3020047637951566,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044723543291922936,
      "loss": 4.9628,
      "mean_token_accuracy": 0.18933759033679962,
      "num_tokens": 50808560.0,
      "step": 23195
    },
    {
      "entropy": 5.583974838256836,
      "epoch": 2.3025009924573245,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044721263250016086,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18357377201318742,
      "num_tokens": 50819328.0,
      "step": 23200
    },
    {
      "entropy": 5.67671275138855,
      "epoch": 2.302997221119492,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044718982781059017,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17829752266407012,
      "num_tokens": 50830066.0,
      "step": 23205
    },
    {
      "entropy": 5.725508213043213,
      "epoch": 2.3034934497816595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044716701885108275,
      "loss": 5.1505,
      "mean_token_accuracy": 0.17625298649072646,
      "num_tokens": 50840775.0,
      "step": 23210
    },
    {
      "entropy": 5.655249166488647,
      "epoch": 2.303989678443827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044714420562220423,
      "loss": 5.0969,
      "mean_token_accuracy": 0.17684097737073898,
      "num_tokens": 50850936.0,
      "step": 23215
    },
    {
      "entropy": 5.624174642562866,
      "epoch": 2.3044859071059944,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004471213881245204,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18041082322597504,
      "num_tokens": 50862800.0,
      "step": 23220
    },
    {
      "entropy": 5.694290351867676,
      "epoch": 2.304982135768162,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000447098566358597,
      "loss": 5.035,
      "mean_token_accuracy": 0.18797155022621154,
      "num_tokens": 50873450.0,
      "step": 23225
    },
    {
      "entropy": 5.61688003540039,
      "epoch": 2.3054783644303294,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044707574032500005,
      "loss": 5.0007,
      "mean_token_accuracy": 0.17770280092954635,
      "num_tokens": 50885624.0,
      "step": 23230
    },
    {
      "entropy": 5.668830537796021,
      "epoch": 2.305974593092497,
      "grad_norm": 1.25,
      "learning_rate": 0.0004470529100242955,
      "loss": 4.9951,
      "mean_token_accuracy": 0.18588711470365524,
      "num_tokens": 50896464.0,
      "step": 23235
    },
    {
      "entropy": 5.687212800979614,
      "epoch": 2.3064708217546643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004470300754570495,
      "loss": 5.0332,
      "mean_token_accuracy": 0.18281919360160828,
      "num_tokens": 50907062.0,
      "step": 23240
    },
    {
      "entropy": 5.553774452209472,
      "epoch": 2.3069670504168323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004470072366238284,
      "loss": 4.9953,
      "mean_token_accuracy": 0.18823616802692414,
      "num_tokens": 50916637.0,
      "step": 23245
    },
    {
      "entropy": 5.578912544250488,
      "epoch": 2.3074632790789997,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044698439352519836,
      "loss": 4.9333,
      "mean_token_accuracy": 0.1869349420070648,
      "num_tokens": 50926573.0,
      "step": 23250
    },
    {
      "entropy": 5.649013185501099,
      "epoch": 2.307959507741167,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044696154616172604,
      "loss": 5.0275,
      "mean_token_accuracy": 0.18704343736171722,
      "num_tokens": 50936767.0,
      "step": 23255
    },
    {
      "entropy": 5.665967702865601,
      "epoch": 2.3084557364033347,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004469386945339779,
      "loss": 5.1413,
      "mean_token_accuracy": 0.1792658492922783,
      "num_tokens": 50948884.0,
      "step": 23260
    },
    {
      "entropy": 5.544977331161499,
      "epoch": 2.308951965065502,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004469158386425206,
      "loss": 4.9639,
      "mean_token_accuracy": 0.1782665252685547,
      "num_tokens": 50959634.0,
      "step": 23265
    },
    {
      "entropy": 5.633510160446167,
      "epoch": 2.3094481937276696,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000446892978487921,
      "loss": 5.0632,
      "mean_token_accuracy": 0.16926396489143372,
      "num_tokens": 50970420.0,
      "step": 23270
    },
    {
      "entropy": 5.6637688159942625,
      "epoch": 2.309944422389837,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044687011407074583,
      "loss": 5.062,
      "mean_token_accuracy": 0.18073338717222215,
      "num_tokens": 50981564.0,
      "step": 23275
    },
    {
      "entropy": 5.651223278045654,
      "epoch": 2.3104406510520046,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004468472453915623,
      "loss": 4.9678,
      "mean_token_accuracy": 0.18373632580041885,
      "num_tokens": 50991523.0,
      "step": 23280
    },
    {
      "entropy": 5.60380277633667,
      "epoch": 2.3109368797141725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004468243724509373,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18350670784711837,
      "num_tokens": 51002339.0,
      "step": 23285
    },
    {
      "entropy": 5.657422828674316,
      "epoch": 2.31143310837634,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004468014952494382,
      "loss": 5.1059,
      "mean_token_accuracy": 0.18121177703142166,
      "num_tokens": 51013973.0,
      "step": 23290
    },
    {
      "entropy": 5.68475308418274,
      "epoch": 2.3119293370385074,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044677861378763205,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18108393400907516,
      "num_tokens": 51024450.0,
      "step": 23295
    },
    {
      "entropy": 5.624780941009521,
      "epoch": 2.312425565700675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044675572806608655,
      "loss": 5.0,
      "mean_token_accuracy": 0.18304375261068345,
      "num_tokens": 51035098.0,
      "step": 23300
    },
    {
      "entropy": 5.627146816253662,
      "epoch": 2.3129217943628424,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044673283808536915,
      "loss": 5.017,
      "mean_token_accuracy": 0.18252273499965668,
      "num_tokens": 51045402.0,
      "step": 23305
    },
    {
      "entropy": 5.674162530899048,
      "epoch": 2.31341802302501,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004467099438460474,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18435097187757493,
      "num_tokens": 51056939.0,
      "step": 23310
    },
    {
      "entropy": 5.651375198364258,
      "epoch": 2.3139142516871773,
      "grad_norm": 1.25,
      "learning_rate": 0.00044668704534868886,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18816355466842652,
      "num_tokens": 51066952.0,
      "step": 23315
    },
    {
      "entropy": 5.660790586471558,
      "epoch": 2.314410480349345,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044666414259386173,
      "loss": 4.9812,
      "mean_token_accuracy": 0.19218651205301285,
      "num_tokens": 51076907.0,
      "step": 23320
    },
    {
      "entropy": 5.663710451126098,
      "epoch": 2.3149067090115123,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044664123558213367,
      "loss": 5.1268,
      "mean_token_accuracy": 0.1731318563222885,
      "num_tokens": 51087089.0,
      "step": 23325
    },
    {
      "entropy": 5.6761617183685305,
      "epoch": 2.31540293767368,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004466183243140729,
      "loss": 5.0337,
      "mean_token_accuracy": 0.1823024943470955,
      "num_tokens": 51097977.0,
      "step": 23330
    },
    {
      "entropy": 5.733677101135254,
      "epoch": 2.3158991663358477,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044659540879024746,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17660926282405853,
      "num_tokens": 51108929.0,
      "step": 23335
    },
    {
      "entropy": 5.678241539001465,
      "epoch": 2.316395394998015,
      "grad_norm": 1.125,
      "learning_rate": 0.0004465724890112256,
      "loss": 5.0125,
      "mean_token_accuracy": 0.18342423141002656,
      "num_tokens": 51118860.0,
      "step": 23340
    },
    {
      "entropy": 5.617818784713745,
      "epoch": 2.3168916236601826,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044654956497757576,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1783423274755478,
      "num_tokens": 51130321.0,
      "step": 23345
    },
    {
      "entropy": 5.573489427566528,
      "epoch": 2.31738785232235,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044652663668986634,
      "loss": 4.9759,
      "mean_token_accuracy": 0.18080510646104814,
      "num_tokens": 51140400.0,
      "step": 23350
    },
    {
      "entropy": 5.654998397827148,
      "epoch": 2.3178840809845176,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000446503704148666,
      "loss": 4.9884,
      "mean_token_accuracy": 0.18159496635198594,
      "num_tokens": 51149823.0,
      "step": 23355
    },
    {
      "entropy": 5.596245670318604,
      "epoch": 2.318380309646685,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004464807673545433,
      "loss": 4.9448,
      "mean_token_accuracy": 0.18443479090929032,
      "num_tokens": 51161193.0,
      "step": 23360
    },
    {
      "entropy": 5.628393507003784,
      "epoch": 2.3188765383088525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004464578263080671,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18094077557325364,
      "num_tokens": 51172566.0,
      "step": 23365
    },
    {
      "entropy": 5.620411062240601,
      "epoch": 2.3193727669710205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044643488100980623,
      "loss": 4.9652,
      "mean_token_accuracy": 0.1914927989244461,
      "num_tokens": 51182484.0,
      "step": 23370
    },
    {
      "entropy": 5.612241172790528,
      "epoch": 2.319868995633188,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004464119314603298,
      "loss": 4.9245,
      "mean_token_accuracy": 0.18481723219156265,
      "num_tokens": 51193081.0,
      "step": 23375
    },
    {
      "entropy": 5.628762483596802,
      "epoch": 2.3203652242953554,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044638897766020673,
      "loss": 5.0795,
      "mean_token_accuracy": 0.1868675708770752,
      "num_tokens": 51203540.0,
      "step": 23380
    },
    {
      "entropy": 5.601858186721802,
      "epoch": 2.320861452957523,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004463660196100064,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18504928052425385,
      "num_tokens": 51213482.0,
      "step": 23385
    },
    {
      "entropy": 5.642043828964233,
      "epoch": 2.3213576816196904,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044634305731029803,
      "loss": 5.007,
      "mean_token_accuracy": 0.18119369000196456,
      "num_tokens": 51223520.0,
      "step": 23390
    },
    {
      "entropy": 5.713339948654175,
      "epoch": 2.321853910281858,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044632009076165096,
      "loss": 5.0542,
      "mean_token_accuracy": 0.17941094785928727,
      "num_tokens": 51233133.0,
      "step": 23395
    },
    {
      "entropy": 5.656205129623413,
      "epoch": 2.3223501389440253,
      "grad_norm": 1.03125,
      "learning_rate": 0.000446297119964635,
      "loss": 5.0384,
      "mean_token_accuracy": 0.185980723798275,
      "num_tokens": 51245048.0,
      "step": 23400
    },
    {
      "entropy": 5.660390424728393,
      "epoch": 2.322846367606193,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004462741449198194,
      "loss": 5.0944,
      "mean_token_accuracy": 0.17624099552631378,
      "num_tokens": 51256394.0,
      "step": 23405
    },
    {
      "entropy": 5.7104692459106445,
      "epoch": 2.3233425962683603,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044625116562777413,
      "loss": 5.051,
      "mean_token_accuracy": 0.1867530345916748,
      "num_tokens": 51268311.0,
      "step": 23410
    },
    {
      "entropy": 5.625567054748535,
      "epoch": 2.323838824930528,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004462281820890689,
      "loss": 4.9954,
      "mean_token_accuracy": 0.177264404296875,
      "num_tokens": 51278386.0,
      "step": 23415
    },
    {
      "entropy": 5.6718535900115965,
      "epoch": 2.3243350535926957,
      "grad_norm": 1.0,
      "learning_rate": 0.00044620519430427376,
      "loss": 5.1351,
      "mean_token_accuracy": 0.179587784409523,
      "num_tokens": 51288544.0,
      "step": 23420
    },
    {
      "entropy": 5.685523223876953,
      "epoch": 2.324831282254863,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004461822022739588,
      "loss": 5.0777,
      "mean_token_accuracy": 0.1814781591296196,
      "num_tokens": 51300006.0,
      "step": 23425
    },
    {
      "entropy": 5.6649613857269285,
      "epoch": 2.3253275109170306,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000446159205998694,
      "loss": 5.0102,
      "mean_token_accuracy": 0.18712896406650542,
      "num_tokens": 51311206.0,
      "step": 23430
    },
    {
      "entropy": 5.6176246166229244,
      "epoch": 2.325823739579198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004461362054790497,
      "loss": 5.0049,
      "mean_token_accuracy": 0.18671022355556488,
      "num_tokens": 51321540.0,
      "step": 23435
    },
    {
      "entropy": 5.644129943847656,
      "epoch": 2.3263199682413656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044611320071559637,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17710921168327332,
      "num_tokens": 51333058.0,
      "step": 23440
    },
    {
      "entropy": 5.645361089706421,
      "epoch": 2.326816196903533,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044609019170890424,
      "loss": 5.0566,
      "mean_token_accuracy": 0.1817330464720726,
      "num_tokens": 51344200.0,
      "step": 23445
    },
    {
      "entropy": 5.611101293563843,
      "epoch": 2.3273124255657005,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044606717845954415,
      "loss": 5.0328,
      "mean_token_accuracy": 0.17921260893344879,
      "num_tokens": 51355365.0,
      "step": 23450
    },
    {
      "entropy": 5.644687843322754,
      "epoch": 2.3278086542278684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044604416096808656,
      "loss": 5.012,
      "mean_token_accuracy": 0.1841609463095665,
      "num_tokens": 51365328.0,
      "step": 23455
    },
    {
      "entropy": 5.733993196487427,
      "epoch": 2.328304882890036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044602113923510235,
      "loss": 5.1368,
      "mean_token_accuracy": 0.17378316670656205,
      "num_tokens": 51376922.0,
      "step": 23460
    },
    {
      "entropy": 5.698174238204956,
      "epoch": 2.3288011115522034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044599811326116246,
      "loss": 5.0831,
      "mean_token_accuracy": 0.18278326243162155,
      "num_tokens": 51388183.0,
      "step": 23465
    },
    {
      "entropy": 5.635926055908203,
      "epoch": 2.329297340214371,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044597508304683784,
      "loss": 5.0275,
      "mean_token_accuracy": 0.17707591950893403,
      "num_tokens": 51399323.0,
      "step": 23470
    },
    {
      "entropy": 5.688930368423462,
      "epoch": 2.3297935688765383,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004459520485926995,
      "loss": 5.0408,
      "mean_token_accuracy": 0.18505235761404037,
      "num_tokens": 51409350.0,
      "step": 23475
    },
    {
      "entropy": 5.739190483093262,
      "epoch": 2.330289797538706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004459290098993188,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17717290222644805,
      "num_tokens": 51420064.0,
      "step": 23480
    },
    {
      "entropy": 5.617911624908447,
      "epoch": 2.3307860262008733,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044590596696726686,
      "loss": 5.0232,
      "mean_token_accuracy": 0.1861350119113922,
      "num_tokens": 51432129.0,
      "step": 23485
    },
    {
      "entropy": 5.632277202606201,
      "epoch": 2.3312822548630407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004458829197971153,
      "loss": 5.0872,
      "mean_token_accuracy": 0.1773812085390091,
      "num_tokens": 51442795.0,
      "step": 23490
    },
    {
      "entropy": 5.587222337722778,
      "epoch": 2.3317784835252082,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044585986838943556,
      "loss": 4.9565,
      "mean_token_accuracy": 0.18548147082328797,
      "num_tokens": 51453551.0,
      "step": 23495
    },
    {
      "entropy": 5.650817394256592,
      "epoch": 2.3322747121873757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044583681274479927,
      "loss": 5.0533,
      "mean_token_accuracy": 0.1735969826579094,
      "num_tokens": 51464997.0,
      "step": 23500
    },
    {
      "entropy": 5.6134754657745365,
      "epoch": 2.3327709408495436,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004458137528637781,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18466079384088516,
      "num_tokens": 51476750.0,
      "step": 23505
    },
    {
      "entropy": 5.697409534454346,
      "epoch": 2.333267169511711,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044579068874694386,
      "loss": 5.0968,
      "mean_token_accuracy": 0.181773841381073,
      "num_tokens": 51487348.0,
      "step": 23510
    },
    {
      "entropy": 5.691198921203613,
      "epoch": 2.3337633981738786,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004457676203948687,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18046712428331374,
      "num_tokens": 51498646.0,
      "step": 23515
    },
    {
      "entropy": 5.588406324386597,
      "epoch": 2.334259626836046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004457445478081244,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18476483970880508,
      "num_tokens": 51508881.0,
      "step": 23520
    },
    {
      "entropy": 5.65836009979248,
      "epoch": 2.3347558554982135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044572147098728335,
      "loss": 5.0753,
      "mean_token_accuracy": 0.1802268534898758,
      "num_tokens": 51518959.0,
      "step": 23525
    },
    {
      "entropy": 5.650944948196411,
      "epoch": 2.335252084160381,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044569838993291757,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18270763009786606,
      "num_tokens": 51530472.0,
      "step": 23530
    },
    {
      "entropy": 5.6662839412689205,
      "epoch": 2.3357483128225485,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044567530464559963,
      "loss": 5.0178,
      "mean_token_accuracy": 0.1847679689526558,
      "num_tokens": 51542639.0,
      "step": 23535
    },
    {
      "entropy": 5.659786939620972,
      "epoch": 2.3362445414847164,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004456522151259019,
      "loss": 5.0458,
      "mean_token_accuracy": 0.18462443798780442,
      "num_tokens": 51553341.0,
      "step": 23540
    },
    {
      "entropy": 5.628106737136841,
      "epoch": 2.336740770146884,
      "grad_norm": 0.96875,
      "learning_rate": 0.000445629121374397,
      "loss": 5.0058,
      "mean_token_accuracy": 0.18270948976278306,
      "num_tokens": 51564787.0,
      "step": 23545
    },
    {
      "entropy": 5.705596303939819,
      "epoch": 2.3372369988090513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044560602339165746,
      "loss": 5.1204,
      "mean_token_accuracy": 0.17987988293170928,
      "num_tokens": 51576497.0,
      "step": 23550
    },
    {
      "entropy": 5.693693161010742,
      "epoch": 2.337733227471219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004455829211782562,
      "loss": 5.013,
      "mean_token_accuracy": 0.18044076412916182,
      "num_tokens": 51587940.0,
      "step": 23555
    },
    {
      "entropy": 5.6353835105896,
      "epoch": 2.3382294561333863,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004455598147347661,
      "loss": 4.9924,
      "mean_token_accuracy": 0.1830785647034645,
      "num_tokens": 51599103.0,
      "step": 23560
    },
    {
      "entropy": 5.677118539810181,
      "epoch": 2.3387256847955538,
      "grad_norm": 1.125,
      "learning_rate": 0.0004455367040617601,
      "loss": 5.139,
      "mean_token_accuracy": 0.17233790904283525,
      "num_tokens": 51610257.0,
      "step": 23565
    },
    {
      "entropy": 5.66479172706604,
      "epoch": 2.3392219134577212,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044551358915981135,
      "loss": 5.0344,
      "mean_token_accuracy": 0.18411360085010528,
      "num_tokens": 51620846.0,
      "step": 23570
    },
    {
      "entropy": 5.7358848571777346,
      "epoch": 2.3397181421198887,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004454904700294929,
      "loss": 5.0121,
      "mean_token_accuracy": 0.18848506212234498,
      "num_tokens": 51632432.0,
      "step": 23575
    },
    {
      "entropy": 5.628352117538452,
      "epoch": 2.340214370782056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004454673466713783,
      "loss": 4.9991,
      "mean_token_accuracy": 0.18529493361711502,
      "num_tokens": 51644086.0,
      "step": 23580
    },
    {
      "entropy": 5.629423761367798,
      "epoch": 2.3407105994442237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044544421908604073,
      "loss": 5.0315,
      "mean_token_accuracy": 0.17891493141651155,
      "num_tokens": 51655660.0,
      "step": 23585
    },
    {
      "entropy": 5.6691450595855715,
      "epoch": 2.3412068281063916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044542108727405386,
      "loss": 5.1131,
      "mean_token_accuracy": 0.17287773936986922,
      "num_tokens": 51667628.0,
      "step": 23590
    },
    {
      "entropy": 5.615863943099976,
      "epoch": 2.341703056768559,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044539795123599117,
      "loss": 4.9795,
      "mean_token_accuracy": 0.1762861728668213,
      "num_tokens": 51677426.0,
      "step": 23595
    },
    {
      "entropy": 5.656557989120484,
      "epoch": 2.3421992854307265,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044537481097242654,
      "loss": 5.0752,
      "mean_token_accuracy": 0.1788928270339966,
      "num_tokens": 51688254.0,
      "step": 23600
    },
    {
      "entropy": 5.663504219055175,
      "epoch": 2.342695514092894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044535166648393377,
      "loss": 5.0502,
      "mean_token_accuracy": 0.18253463953733445,
      "num_tokens": 51699852.0,
      "step": 23605
    },
    {
      "entropy": 5.620824956893921,
      "epoch": 2.3431917427550615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044532851777108663,
      "loss": 5.0408,
      "mean_token_accuracy": 0.18622306287288665,
      "num_tokens": 51710758.0,
      "step": 23610
    },
    {
      "entropy": 5.528943109512329,
      "epoch": 2.343687971417229,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004453053648344593,
      "loss": 4.9271,
      "mean_token_accuracy": 0.18655824512243271,
      "num_tokens": 51721151.0,
      "step": 23615
    },
    {
      "entropy": 5.72039566040039,
      "epoch": 2.3441842000793964,
      "grad_norm": 0.890625,
      "learning_rate": 0.00044528220767462595,
      "loss": 5.1311,
      "mean_token_accuracy": 0.17608120292425156,
      "num_tokens": 51733160.0,
      "step": 23620
    },
    {
      "entropy": 5.68540849685669,
      "epoch": 2.3446804287415643,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004452590462921607,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18769578337669374,
      "num_tokens": 51743923.0,
      "step": 23625
    },
    {
      "entropy": 5.599461555480957,
      "epoch": 2.345176657403732,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044523588068763804,
      "loss": 5.008,
      "mean_token_accuracy": 0.18350127190351487,
      "num_tokens": 51754601.0,
      "step": 23630
    },
    {
      "entropy": 5.67606635093689,
      "epoch": 2.3456728860658993,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452127108616324,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17386944890022277,
      "num_tokens": 51766041.0,
      "step": 23635
    },
    {
      "entropy": 5.611720180511474,
      "epoch": 2.3461691147280668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044518953681471824,
      "loss": 4.9996,
      "mean_token_accuracy": 0.1921327605843544,
      "num_tokens": 51777541.0,
      "step": 23640
    },
    {
      "entropy": 5.696330881118774,
      "epoch": 2.3466653433902342,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004451663585474703,
      "loss": 5.06,
      "mean_token_accuracy": 0.17823253870010375,
      "num_tokens": 51789914.0,
      "step": 23645
    },
    {
      "entropy": 5.647289419174195,
      "epoch": 2.3471615720524017,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044514317606046336,
      "loss": 4.9853,
      "mean_token_accuracy": 0.19000324308872224,
      "num_tokens": 51801289.0,
      "step": 23650
    },
    {
      "entropy": 5.597973918914795,
      "epoch": 2.347657800714569,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004451199893542723,
      "loss": 4.9566,
      "mean_token_accuracy": 0.1875682592391968,
      "num_tokens": 51811781.0,
      "step": 23655
    },
    {
      "entropy": 5.616306447982788,
      "epoch": 2.3481540293767367,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004450967984294721,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1803790658712387,
      "num_tokens": 51822100.0,
      "step": 23660
    },
    {
      "entropy": 5.6573058605194095,
      "epoch": 2.348650258038904,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004450736032866377,
      "loss": 5.0117,
      "mean_token_accuracy": 0.18948930203914643,
      "num_tokens": 51832495.0,
      "step": 23665
    },
    {
      "entropy": 5.683979749679565,
      "epoch": 2.3491464867010716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044505040392634447,
      "loss": 5.082,
      "mean_token_accuracy": 0.1790887162089348,
      "num_tokens": 51842915.0,
      "step": 23670
    },
    {
      "entropy": 5.684537410736084,
      "epoch": 2.3496427153632395,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004450272003491677,
      "loss": 5.1642,
      "mean_token_accuracy": 0.17364989817142487,
      "num_tokens": 51855268.0,
      "step": 23675
    },
    {
      "entropy": 5.620048379898071,
      "epoch": 2.350138944025407,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044500399255568274,
      "loss": 4.997,
      "mean_token_accuracy": 0.18797430247068406,
      "num_tokens": 51865756.0,
      "step": 23680
    },
    {
      "entropy": 5.624314975738526,
      "epoch": 2.3506351726875745,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044498078054646507,
      "loss": 4.9344,
      "mean_token_accuracy": 0.18846224695444108,
      "num_tokens": 51875079.0,
      "step": 23685
    },
    {
      "entropy": 5.645658540725708,
      "epoch": 2.351131401349742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044495756432209025,
      "loss": 5.1056,
      "mean_token_accuracy": 0.17901993691921234,
      "num_tokens": 51886607.0,
      "step": 23690
    },
    {
      "entropy": 5.688074827194214,
      "epoch": 2.3516276300119094,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044493434388313423,
      "loss": 5.0719,
      "mean_token_accuracy": 0.17960314452648163,
      "num_tokens": 51897553.0,
      "step": 23695
    },
    {
      "entropy": 5.630893850326538,
      "epoch": 2.352123858674077,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044491111923017254,
      "loss": 4.9756,
      "mean_token_accuracy": 0.18375248312950135,
      "num_tokens": 51908395.0,
      "step": 23700
    },
    {
      "entropy": 5.632253217697143,
      "epoch": 2.3526200873362444,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044488789036378117,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18283228874206542,
      "num_tokens": 51919764.0,
      "step": 23705
    },
    {
      "entropy": 5.557384586334228,
      "epoch": 2.3531163159984123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004448646572845363,
      "loss": 5.021,
      "mean_token_accuracy": 0.1826092317700386,
      "num_tokens": 51930904.0,
      "step": 23710
    },
    {
      "entropy": 5.633280181884766,
      "epoch": 2.35361254466058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004448414199930139,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18375150710344315,
      "num_tokens": 51942373.0,
      "step": 23715
    },
    {
      "entropy": 5.73573899269104,
      "epoch": 2.3541087733227473,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044481817848979026,
      "loss": 5.0772,
      "mean_token_accuracy": 0.19094143360853194,
      "num_tokens": 51953770.0,
      "step": 23720
    },
    {
      "entropy": 5.62532525062561,
      "epoch": 2.3546050019849147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004447949327754418,
      "loss": 5.0574,
      "mean_token_accuracy": 0.17318505942821502,
      "num_tokens": 51964047.0,
      "step": 23725
    },
    {
      "entropy": 5.62755274772644,
      "epoch": 2.355101230647082,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004447716828505447,
      "loss": 5.0622,
      "mean_token_accuracy": 0.1789298266172409,
      "num_tokens": 51974813.0,
      "step": 23730
    },
    {
      "entropy": 5.619158697128296,
      "epoch": 2.3555974593092497,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004447484287156758,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1807703211903572,
      "num_tokens": 51986321.0,
      "step": 23735
    },
    {
      "entropy": 5.663240146636963,
      "epoch": 2.356093687971417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004447251703714117,
      "loss": 5.051,
      "mean_token_accuracy": 0.17103521078824996,
      "num_tokens": 51996294.0,
      "step": 23740
    },
    {
      "entropy": 5.566661024093628,
      "epoch": 2.3565899166335846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000444701907818329,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18504777401685715,
      "num_tokens": 52008359.0,
      "step": 23745
    },
    {
      "entropy": 5.624560117721558,
      "epoch": 2.357086145295752,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004446786410570048,
      "loss": 4.9876,
      "mean_token_accuracy": 0.1875425547361374,
      "num_tokens": 52019604.0,
      "step": 23750
    },
    {
      "entropy": 5.648702621459961,
      "epoch": 2.3575823739579196,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044465537008801575,
      "loss": 4.9926,
      "mean_token_accuracy": 0.17866195738315582,
      "num_tokens": 52031238.0,
      "step": 23755
    },
    {
      "entropy": 5.615241670608521,
      "epoch": 2.3580786026200875,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004446320949119392,
      "loss": 5.0285,
      "mean_token_accuracy": 0.1817118525505066,
      "num_tokens": 52042108.0,
      "step": 23760
    },
    {
      "entropy": 5.6171690940856935,
      "epoch": 2.358574831282255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044460881552935215,
      "loss": 5.0584,
      "mean_token_accuracy": 0.18494725078344346,
      "num_tokens": 52052894.0,
      "step": 23765
    },
    {
      "entropy": 5.6314586162567135,
      "epoch": 2.3590710599444225,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044458553194083204,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18923577964305877,
      "num_tokens": 52063622.0,
      "step": 23770
    },
    {
      "entropy": 5.585512590408325,
      "epoch": 2.35956728860659,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004445622441469562,
      "loss": 4.9596,
      "mean_token_accuracy": 0.18312161415815353,
      "num_tokens": 52074457.0,
      "step": 23775
    },
    {
      "entropy": 5.6188633918762205,
      "epoch": 2.3600635172687574,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044453895214830195,
      "loss": 4.9792,
      "mean_token_accuracy": 0.18165361136198044,
      "num_tokens": 52086067.0,
      "step": 23780
    },
    {
      "entropy": 5.582472515106201,
      "epoch": 2.360559745930925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004445156559454471,
      "loss": 4.9826,
      "mean_token_accuracy": 0.1825300246477127,
      "num_tokens": 52096449.0,
      "step": 23785
    },
    {
      "entropy": 5.61797022819519,
      "epoch": 2.3610559745930924,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044449235553896914,
      "loss": 5.0626,
      "mean_token_accuracy": 0.17974215894937515,
      "num_tokens": 52107930.0,
      "step": 23790
    },
    {
      "entropy": 5.690414571762085,
      "epoch": 2.36155220325526,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004444690509294461,
      "loss": 5.1302,
      "mean_token_accuracy": 0.18122328519821168,
      "num_tokens": 52119132.0,
      "step": 23795
    },
    {
      "entropy": 5.680971670150757,
      "epoch": 2.3620484319174277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004444457421174558,
      "loss": 4.9614,
      "mean_token_accuracy": 0.1849889412522316,
      "num_tokens": 52129634.0,
      "step": 23800
    },
    {
      "entropy": 5.625244188308716,
      "epoch": 2.3625446605795952,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004444224291035761,
      "loss": 4.9695,
      "mean_token_accuracy": 0.1815538600087166,
      "num_tokens": 52140218.0,
      "step": 23805
    },
    {
      "entropy": 5.702409362792968,
      "epoch": 2.3630408892417627,
      "grad_norm": 1.0,
      "learning_rate": 0.00044439911188838533,
      "loss": 5.0843,
      "mean_token_accuracy": 0.18672626912593843,
      "num_tokens": 52150795.0,
      "step": 23810
    },
    {
      "entropy": 5.559905099868774,
      "epoch": 2.36353711790393,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004443757904724615,
      "loss": 4.9362,
      "mean_token_accuracy": 0.18530319780111312,
      "num_tokens": 52161240.0,
      "step": 23815
    },
    {
      "entropy": 5.617721033096314,
      "epoch": 2.3640333465660976,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044435246485638317,
      "loss": 5.0226,
      "mean_token_accuracy": 0.18787378519773484,
      "num_tokens": 52172389.0,
      "step": 23820
    },
    {
      "entropy": 5.646850538253784,
      "epoch": 2.364529575228265,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004443291350407285,
      "loss": 5.0231,
      "mean_token_accuracy": 0.18538450300693513,
      "num_tokens": 52183799.0,
      "step": 23825
    },
    {
      "entropy": 5.5912487506866455,
      "epoch": 2.3650258038904326,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004443058010260762,
      "loss": 4.9399,
      "mean_token_accuracy": 0.19167597144842147,
      "num_tokens": 52195291.0,
      "step": 23830
    },
    {
      "entropy": 5.6043590068817135,
      "epoch": 2.3655220325526,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044428246281300493,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1787293091416359,
      "num_tokens": 52205952.0,
      "step": 23835
    },
    {
      "entropy": 5.700117301940918,
      "epoch": 2.3660182612147675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044425912040209326,
      "loss": 5.0844,
      "mean_token_accuracy": 0.18185113072395326,
      "num_tokens": 52216995.0,
      "step": 23840
    },
    {
      "entropy": 5.697113513946533,
      "epoch": 2.3665144898769355,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044423577379392023,
      "loss": 5.0568,
      "mean_token_accuracy": 0.1759188562631607,
      "num_tokens": 52228297.0,
      "step": 23845
    },
    {
      "entropy": 5.686745500564575,
      "epoch": 2.367010718539103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004442124229890646,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18056197166442872,
      "num_tokens": 52238918.0,
      "step": 23850
    },
    {
      "entropy": 5.632084751129151,
      "epoch": 2.3675069472012704,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044418906798810554,
      "loss": 4.9983,
      "mean_token_accuracy": 0.18616099506616593,
      "num_tokens": 52249277.0,
      "step": 23855
    },
    {
      "entropy": 5.643439674377442,
      "epoch": 2.368003175863438,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044416570879162213,
      "loss": 5.0583,
      "mean_token_accuracy": 0.18216273188591003,
      "num_tokens": 52259404.0,
      "step": 23860
    },
    {
      "entropy": 5.655850982666015,
      "epoch": 2.3684994045256054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004441423454001936,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1840360179543495,
      "num_tokens": 52271368.0,
      "step": 23865
    },
    {
      "entropy": 5.617729043960571,
      "epoch": 2.368995633187773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044411897781439946,
      "loss": 5.026,
      "mean_token_accuracy": 0.1809672310948372,
      "num_tokens": 52282176.0,
      "step": 23870
    },
    {
      "entropy": 5.678512001037598,
      "epoch": 2.3694918618499403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004440956060348191,
      "loss": 5.1299,
      "mean_token_accuracy": 0.17272921055555343,
      "num_tokens": 52292593.0,
      "step": 23875
    },
    {
      "entropy": 5.638871765136718,
      "epoch": 2.369988090512108,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004440722300620321,
      "loss": 5.0369,
      "mean_token_accuracy": 0.183934223651886,
      "num_tokens": 52302509.0,
      "step": 23880
    },
    {
      "entropy": 5.674328136444092,
      "epoch": 2.3704843191742757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044404884989661804,
      "loss": 5.0587,
      "mean_token_accuracy": 0.18173269927501678,
      "num_tokens": 52314200.0,
      "step": 23885
    },
    {
      "entropy": 5.661617422103882,
      "epoch": 2.370980547836443,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004440254655391568,
      "loss": 5.0748,
      "mean_token_accuracy": 0.18369985818862916,
      "num_tokens": 52325506.0,
      "step": 23890
    },
    {
      "entropy": 5.55656418800354,
      "epoch": 2.3714767764986107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004440020769902282,
      "loss": 4.8808,
      "mean_token_accuracy": 0.19079311788082123,
      "num_tokens": 52336466.0,
      "step": 23895
    },
    {
      "entropy": 5.669423961639405,
      "epoch": 2.371973005160778,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044397868425041234,
      "loss": 5.0642,
      "mean_token_accuracy": 0.1831175997853279,
      "num_tokens": 52347320.0,
      "step": 23900
    },
    {
      "entropy": 5.7005432605743405,
      "epoch": 2.3724692338229456,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004439552873202891,
      "loss": 5.0271,
      "mean_token_accuracy": 0.18960791677236558,
      "num_tokens": 52358355.0,
      "step": 23905
    },
    {
      "entropy": 5.6285459995269775,
      "epoch": 2.372965462485113,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044393188620043893,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18724546283483506,
      "num_tokens": 52369639.0,
      "step": 23910
    },
    {
      "entropy": 5.639919567108154,
      "epoch": 2.3734616911472806,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004439084808914419,
      "loss": 4.9873,
      "mean_token_accuracy": 0.1872972697019577,
      "num_tokens": 52381691.0,
      "step": 23915
    },
    {
      "entropy": 5.652935361862182,
      "epoch": 2.373957919809448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044388507139387864,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1795426845550537,
      "num_tokens": 52392927.0,
      "step": 23920
    },
    {
      "entropy": 5.642256450653076,
      "epoch": 2.3744541484716155,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004438616577083293,
      "loss": 5.061,
      "mean_token_accuracy": 0.17954618036746978,
      "num_tokens": 52404812.0,
      "step": 23925
    },
    {
      "entropy": 5.75106406211853,
      "epoch": 2.3749503771337834,
      "grad_norm": 1.109375,
      "learning_rate": 0.000443838239835375,
      "loss": 5.114,
      "mean_token_accuracy": 0.18229379504919052,
      "num_tokens": 52417035.0,
      "step": 23930
    },
    {
      "entropy": 5.701602220535278,
      "epoch": 2.375446605795951,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004438148177755959,
      "loss": 5.0245,
      "mean_token_accuracy": 0.18705900609493256,
      "num_tokens": 52428482.0,
      "step": 23935
    },
    {
      "entropy": 5.663438749313355,
      "epoch": 2.3759428344581184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044379139152957327,
      "loss": 5.0955,
      "mean_token_accuracy": 0.18286439329385756,
      "num_tokens": 52438471.0,
      "step": 23940
    },
    {
      "entropy": 5.5900795459747314,
      "epoch": 2.376439063120286,
      "grad_norm": 1.125,
      "learning_rate": 0.00044376796109788777,
      "loss": 4.9769,
      "mean_token_accuracy": 0.17957217544317244,
      "num_tokens": 52450418.0,
      "step": 23945
    },
    {
      "entropy": 5.655994272232055,
      "epoch": 2.3769352917824533,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044374452648112046,
      "loss": 5.0649,
      "mean_token_accuracy": 0.18562849909067153,
      "num_tokens": 52461774.0,
      "step": 23950
    },
    {
      "entropy": 5.691488552093506,
      "epoch": 2.377431520444621,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044372108767985253,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17794844806194304,
      "num_tokens": 52473232.0,
      "step": 23955
    },
    {
      "entropy": 5.682729053497314,
      "epoch": 2.3779277491067883,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004436976446946652,
      "loss": 5.0578,
      "mean_token_accuracy": 0.18162949234247208,
      "num_tokens": 52484683.0,
      "step": 23960
    },
    {
      "entropy": 5.725747680664062,
      "epoch": 2.3784239777689558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044367419752613983,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1702961340546608,
      "num_tokens": 52496578.0,
      "step": 23965
    },
    {
      "entropy": 5.6152345657348635,
      "epoch": 2.3789202064311237,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004436507461748577,
      "loss": 4.9305,
      "mean_token_accuracy": 0.19125109314918518,
      "num_tokens": 52506658.0,
      "step": 23970
    },
    {
      "entropy": 5.714120101928711,
      "epoch": 2.379416435093291,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004436272906414007,
      "loss": 5.1232,
      "mean_token_accuracy": 0.17261904627084732,
      "num_tokens": 52516443.0,
      "step": 23975
    },
    {
      "entropy": 5.641940784454346,
      "epoch": 2.3799126637554586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004436038309263501,
      "loss": 5.083,
      "mean_token_accuracy": 0.17664094418287277,
      "num_tokens": 52526858.0,
      "step": 23980
    },
    {
      "entropy": 5.701242399215698,
      "epoch": 2.380408892417626,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044358036703028773,
      "loss": 5.076,
      "mean_token_accuracy": 0.17838872820138932,
      "num_tokens": 52538263.0,
      "step": 23985
    },
    {
      "entropy": 5.615535879135132,
      "epoch": 2.3809051210797936,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044355689895379563,
      "loss": 5.0192,
      "mean_token_accuracy": 0.18522707521915435,
      "num_tokens": 52548722.0,
      "step": 23990
    },
    {
      "entropy": 5.707331991195678,
      "epoch": 2.381401349741961,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044353342669745566,
      "loss": 5.1245,
      "mean_token_accuracy": 0.17393383681774138,
      "num_tokens": 52560189.0,
      "step": 23995
    },
    {
      "entropy": 5.607894992828369,
      "epoch": 2.3818975784041285,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044350995026184976,
      "loss": 5.0058,
      "mean_token_accuracy": 0.18343357145786285,
      "num_tokens": 52569505.0,
      "step": 24000
    },
    {
      "epoch": 2.3818975784041285,
      "eval_entropy": 5.403048668827629,
      "eval_loss": 5.162159442901611,
      "eval_mean_token_accuracy": 0.18089783211860672,
      "eval_num_tokens": 52569505.0,
      "eval_runtime": 20.2525,
      "eval_samples_per_second": 1662.461,
      "eval_steps_per_second": 207.826,
      "step": 24000
    },
    {
      "entropy": 5.669043397903442,
      "epoch": 2.382393807066296,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044348646964756036,
      "loss": 5.1113,
      "mean_token_accuracy": 0.17567050457000732,
      "num_tokens": 52580665.0,
      "step": 24005
    },
    {
      "entropy": 5.685612058639526,
      "epoch": 2.3828900357284635,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004434629848551694,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1786411225795746,
      "num_tokens": 52591506.0,
      "step": 24010
    },
    {
      "entropy": 5.691883850097656,
      "epoch": 2.3833862643906314,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004434394958852595,
      "loss": 5.1526,
      "mean_token_accuracy": 0.17271867841482164,
      "num_tokens": 52602004.0,
      "step": 24015
    },
    {
      "entropy": 5.645259714126587,
      "epoch": 2.383882493052799,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044341600273841304,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18516378849744797,
      "num_tokens": 52613035.0,
      "step": 24020
    },
    {
      "entropy": 5.627730464935302,
      "epoch": 2.3843787217149663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044339250541521264,
      "loss": 5.0633,
      "mean_token_accuracy": 0.18329248428344727,
      "num_tokens": 52623732.0,
      "step": 24025
    },
    {
      "entropy": 5.697849988937378,
      "epoch": 2.384874950377134,
      "grad_norm": 1.125,
      "learning_rate": 0.00044336900391624096,
      "loss": 5.1065,
      "mean_token_accuracy": 0.17491652965545654,
      "num_tokens": 52634325.0,
      "step": 24030
    },
    {
      "entropy": 5.656633377075195,
      "epoch": 2.3853711790393013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004433454982420808,
      "loss": 5.0012,
      "mean_token_accuracy": 0.19104858338832856,
      "num_tokens": 52645214.0,
      "step": 24035
    },
    {
      "entropy": 5.654068374633789,
      "epoch": 2.3858674077014688,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000443321988393315,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18309513628482818,
      "num_tokens": 52655726.0,
      "step": 24040
    },
    {
      "entropy": 5.658140420913696,
      "epoch": 2.3863636363636362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004432984743705265,
      "loss": 5.1676,
      "mean_token_accuracy": 0.17882040590047837,
      "num_tokens": 52667129.0,
      "step": 24045
    },
    {
      "entropy": 5.672997522354126,
      "epoch": 2.3868598650258037,
      "grad_norm": 1.125,
      "learning_rate": 0.00044327495617429866,
      "loss": 5.0955,
      "mean_token_accuracy": 0.176936274766922,
      "num_tokens": 52678707.0,
      "step": 24050
    },
    {
      "entropy": 5.713320350646972,
      "epoch": 2.3873560936879716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004432514338052144,
      "loss": 5.0389,
      "mean_token_accuracy": 0.18242161870002746,
      "num_tokens": 52688456.0,
      "step": 24055
    },
    {
      "entropy": 5.584195327758789,
      "epoch": 2.387852322350139,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004432279072638571,
      "loss": 4.9369,
      "mean_token_accuracy": 0.18988122045993805,
      "num_tokens": 52699591.0,
      "step": 24060
    },
    {
      "entropy": 5.61508731842041,
      "epoch": 2.3883485510123066,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044320437655081033,
      "loss": 5.05,
      "mean_token_accuracy": 0.18222817480564119,
      "num_tokens": 52710115.0,
      "step": 24065
    },
    {
      "entropy": 5.621645259857178,
      "epoch": 2.388844779674474,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044318084166665737,
      "loss": 5.0099,
      "mean_token_accuracy": 0.17807424664497376,
      "num_tokens": 52720214.0,
      "step": 24070
    },
    {
      "entropy": 5.653671360015869,
      "epoch": 2.3893410083366415,
      "grad_norm": 1.203125,
      "learning_rate": 0.000443157302611982,
      "loss": 4.9672,
      "mean_token_accuracy": 0.18673840910196304,
      "num_tokens": 52731523.0,
      "step": 24075
    },
    {
      "entropy": 5.641802453994751,
      "epoch": 2.389837236998809,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004431337593873678,
      "loss": 5.0233,
      "mean_token_accuracy": 0.18842461705207825,
      "num_tokens": 52742770.0,
      "step": 24080
    },
    {
      "entropy": 5.626781845092774,
      "epoch": 2.3903334656609765,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004431102119933987,
      "loss": 5.0093,
      "mean_token_accuracy": 0.18103839904069902,
      "num_tokens": 52753939.0,
      "step": 24085
    },
    {
      "entropy": 5.693289756774902,
      "epoch": 2.390829694323144,
      "grad_norm": 1.125,
      "learning_rate": 0.0004430866604306586,
      "loss": 5.0901,
      "mean_token_accuracy": 0.1772838279604912,
      "num_tokens": 52764446.0,
      "step": 24090
    },
    {
      "entropy": 5.602644062042236,
      "epoch": 2.3913259229853114,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044306310469973147,
      "loss": 5.027,
      "mean_token_accuracy": 0.18185321390628814,
      "num_tokens": 52773933.0,
      "step": 24095
    },
    {
      "entropy": 5.591819000244141,
      "epoch": 2.391822151647479,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044303954480120155,
      "loss": 4.9739,
      "mean_token_accuracy": 0.18670920431613922,
      "num_tokens": 52784481.0,
      "step": 24100
    },
    {
      "entropy": 5.622302675247193,
      "epoch": 2.392318380309647,
      "grad_norm": 1.125,
      "learning_rate": 0.00044301598073565295,
      "loss": 5.022,
      "mean_token_accuracy": 0.18448663055896758,
      "num_tokens": 52795866.0,
      "step": 24105
    },
    {
      "entropy": 5.5894921779632565,
      "epoch": 2.3928146089718143,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004429924125036702,
      "loss": 5.0269,
      "mean_token_accuracy": 0.17816588282585144,
      "num_tokens": 52806617.0,
      "step": 24110
    },
    {
      "entropy": 5.682108163833618,
      "epoch": 2.393310837633982,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004429688401058375,
      "loss": 5.1568,
      "mean_token_accuracy": 0.1802821323275566,
      "num_tokens": 52819246.0,
      "step": 24115
    },
    {
      "entropy": 5.6146331310272215,
      "epoch": 2.3938070662961493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004429452635427394,
      "loss": 4.9987,
      "mean_token_accuracy": 0.1895383581519127,
      "num_tokens": 52829730.0,
      "step": 24120
    },
    {
      "entropy": 5.664277362823486,
      "epoch": 2.3943032949583167,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044292168281496086,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17675748318433762,
      "num_tokens": 52840010.0,
      "step": 24125
    },
    {
      "entropy": 5.6633695602417,
      "epoch": 2.394799523620484,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004428980979230863,
      "loss": 5.0627,
      "mean_token_accuracy": 0.1798376202583313,
      "num_tokens": 52850619.0,
      "step": 24130
    },
    {
      "entropy": 5.704005193710327,
      "epoch": 2.3952957522826517,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044287450886770075,
      "loss": 5.1029,
      "mean_token_accuracy": 0.1787817195057869,
      "num_tokens": 52862560.0,
      "step": 24135
    },
    {
      "entropy": 5.679721736907959,
      "epoch": 2.3957919809448196,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004428509156493891,
      "loss": 5.021,
      "mean_token_accuracy": 0.18217352479696275,
      "num_tokens": 52872741.0,
      "step": 24140
    },
    {
      "entropy": 5.582712268829345,
      "epoch": 2.396288209606987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004428273182687364,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1837775468826294,
      "num_tokens": 52884026.0,
      "step": 24145
    },
    {
      "entropy": 5.64924726486206,
      "epoch": 2.3967844382691545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004428037167263279,
      "loss": 5.0072,
      "mean_token_accuracy": 0.18868427127599716,
      "num_tokens": 52894429.0,
      "step": 24150
    },
    {
      "entropy": 5.650578403472901,
      "epoch": 2.397280666931322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044278011102274887,
      "loss": 5.0833,
      "mean_token_accuracy": 0.17857176810503006,
      "num_tokens": 52907157.0,
      "step": 24155
    },
    {
      "entropy": 5.619512128829956,
      "epoch": 2.3977768955934895,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044275650115858445,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18887805044651032,
      "num_tokens": 52918563.0,
      "step": 24160
    },
    {
      "entropy": 5.617229080200195,
      "epoch": 2.398273124255657,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004427328871344204,
      "loss": 4.9783,
      "mean_token_accuracy": 0.18689391165971755,
      "num_tokens": 52929879.0,
      "step": 24165
    },
    {
      "entropy": 5.545815086364746,
      "epoch": 2.3987693529178244,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004427092689508422,
      "loss": 4.9739,
      "mean_token_accuracy": 0.183808034658432,
      "num_tokens": 52941474.0,
      "step": 24170
    },
    {
      "entropy": 5.609093570709229,
      "epoch": 2.399265581579992,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004426856466084354,
      "loss": 5.0274,
      "mean_token_accuracy": 0.17926309555768966,
      "num_tokens": 52952326.0,
      "step": 24175
    },
    {
      "entropy": 5.6382301330566404,
      "epoch": 2.3997618102421594,
      "grad_norm": 1.0,
      "learning_rate": 0.000442662020107786,
      "loss": 5.0529,
      "mean_token_accuracy": 0.18043892234563827,
      "num_tokens": 52962546.0,
      "step": 24180
    },
    {
      "entropy": 5.659893131256103,
      "epoch": 2.400258038904327,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004426383894494797,
      "loss": 5.0447,
      "mean_token_accuracy": 0.18511613607406616,
      "num_tokens": 52972619.0,
      "step": 24185
    },
    {
      "entropy": 5.667884254455567,
      "epoch": 2.400754267566495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004426147546341027,
      "loss": 5.0551,
      "mean_token_accuracy": 0.17983965426683426,
      "num_tokens": 52982536.0,
      "step": 24190
    },
    {
      "entropy": 5.595958566665649,
      "epoch": 2.4012504962286623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044259111566224084,
      "loss": 5.0862,
      "mean_token_accuracy": 0.180138236284256,
      "num_tokens": 52994390.0,
      "step": 24195
    },
    {
      "entropy": 5.654284286499023,
      "epoch": 2.4017467248908297,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044256747253448045,
      "loss": 5.0069,
      "mean_token_accuracy": 0.1808893233537674,
      "num_tokens": 53004204.0,
      "step": 24200
    },
    {
      "entropy": 5.694934844970703,
      "epoch": 2.402242953552997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044254382525140784,
      "loss": 5.0621,
      "mean_token_accuracy": 0.18730689734220504,
      "num_tokens": 53014045.0,
      "step": 24205
    },
    {
      "entropy": 5.67195234298706,
      "epoch": 2.4027391822151647,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044252017381360935,
      "loss": 5.0805,
      "mean_token_accuracy": 0.17557526528835296,
      "num_tokens": 53024945.0,
      "step": 24210
    },
    {
      "entropy": 5.65529317855835,
      "epoch": 2.403235410877332,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004424965182216716,
      "loss": 4.9674,
      "mean_token_accuracy": 0.18037308007478714,
      "num_tokens": 53035203.0,
      "step": 24215
    },
    {
      "entropy": 5.619567823410034,
      "epoch": 2.4037316395394996,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044247285847618105,
      "loss": 5.0019,
      "mean_token_accuracy": 0.184102101624012,
      "num_tokens": 53047359.0,
      "step": 24220
    },
    {
      "entropy": 5.689309883117676,
      "epoch": 2.4042278682016676,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004424491945777245,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17921091467142106,
      "num_tokens": 53057716.0,
      "step": 24225
    },
    {
      "entropy": 5.680348110198975,
      "epoch": 2.404724096863835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044242552652688885,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17212722599506378,
      "num_tokens": 53069087.0,
      "step": 24230
    },
    {
      "entropy": 5.625019598007202,
      "epoch": 2.4052203255260025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004424018543242608,
      "loss": 4.9862,
      "mean_token_accuracy": 0.18755027353763581,
      "num_tokens": 53079462.0,
      "step": 24235
    },
    {
      "entropy": 5.63356614112854,
      "epoch": 2.40571655418817,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004423781779704275,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18231211304664613,
      "num_tokens": 53090398.0,
      "step": 24240
    },
    {
      "entropy": 5.623479700088501,
      "epoch": 2.4062127828503375,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004423544974659761,
      "loss": 4.9455,
      "mean_token_accuracy": 0.18779552429914476,
      "num_tokens": 53101250.0,
      "step": 24245
    },
    {
      "entropy": 5.612404680252075,
      "epoch": 2.406709011512505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044233081281149373,
      "loss": 5.0224,
      "mean_token_accuracy": 0.18438326567411423,
      "num_tokens": 53112039.0,
      "step": 24250
    },
    {
      "entropy": 5.593623685836792,
      "epoch": 2.4072052401746724,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044230712400756774,
      "loss": 4.9793,
      "mean_token_accuracy": 0.19613129794597625,
      "num_tokens": 53123348.0,
      "step": 24255
    },
    {
      "entropy": 5.667486953735351,
      "epoch": 2.40770146883684,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044228343105478564,
      "loss": 5.0288,
      "mean_token_accuracy": 0.1798804819583893,
      "num_tokens": 53133878.0,
      "step": 24260
    },
    {
      "entropy": 5.654512929916382,
      "epoch": 2.4081976974990074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004422597339537349,
      "loss": 5.01,
      "mean_token_accuracy": 0.18174678236246108,
      "num_tokens": 53144763.0,
      "step": 24265
    },
    {
      "entropy": 5.675848436355591,
      "epoch": 2.408693926161175,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004422360327050032,
      "loss": 5.0335,
      "mean_token_accuracy": 0.1782678633928299,
      "num_tokens": 53155550.0,
      "step": 24270
    },
    {
      "entropy": 5.606454467773437,
      "epoch": 2.4091901548233428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044221232730917823,
      "loss": 4.9437,
      "mean_token_accuracy": 0.19321919232606888,
      "num_tokens": 53166719.0,
      "step": 24275
    },
    {
      "entropy": 5.619027900695801,
      "epoch": 2.4096863834855102,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044218861776684776,
      "loss": 5.0174,
      "mean_token_accuracy": 0.1854616269469261,
      "num_tokens": 53177923.0,
      "step": 24280
    },
    {
      "entropy": 5.649141836166382,
      "epoch": 2.4101826121476777,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044216490407859995,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1839844211935997,
      "num_tokens": 53188628.0,
      "step": 24285
    },
    {
      "entropy": 5.5947106838226315,
      "epoch": 2.410678840809845,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044214118624502267,
      "loss": 4.9657,
      "mean_token_accuracy": 0.19205175042152406,
      "num_tokens": 53198617.0,
      "step": 24290
    },
    {
      "entropy": 5.759106636047363,
      "epoch": 2.4111750694720127,
      "grad_norm": 1.0,
      "learning_rate": 0.0004421174642667042,
      "loss": 5.1365,
      "mean_token_accuracy": 0.17299814522266388,
      "num_tokens": 53210490.0,
      "step": 24295
    },
    {
      "entropy": 5.688277626037598,
      "epoch": 2.41167129813418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044209373814423265,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17670177519321442,
      "num_tokens": 53221309.0,
      "step": 24300
    },
    {
      "entropy": 5.731527328491211,
      "epoch": 2.4121675267963476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004420700078781965,
      "loss": 5.092,
      "mean_token_accuracy": 0.1775069296360016,
      "num_tokens": 53233342.0,
      "step": 24305
    },
    {
      "entropy": 5.634399127960205,
      "epoch": 2.4126637554585155,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044204627346918407,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18245672136545182,
      "num_tokens": 53243557.0,
      "step": 24310
    },
    {
      "entropy": 5.657622766494751,
      "epoch": 2.413159984120683,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044202253491778404,
      "loss": 4.9997,
      "mean_token_accuracy": 0.18269436359405516,
      "num_tokens": 53253274.0,
      "step": 24315
    },
    {
      "entropy": 5.6114586353302,
      "epoch": 2.4136562127828505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004419987922245852,
      "loss": 5.009,
      "mean_token_accuracy": 0.1840328425168991,
      "num_tokens": 53263955.0,
      "step": 24320
    },
    {
      "entropy": 5.696019458770752,
      "epoch": 2.414152441445018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000441975045390176,
      "loss": 5.1126,
      "mean_token_accuracy": 0.18315571248531343,
      "num_tokens": 53274538.0,
      "step": 24325
    },
    {
      "entropy": 5.690678215026855,
      "epoch": 2.4146486701071854,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044195129441514547,
      "loss": 5.0418,
      "mean_token_accuracy": 0.18040310740470886,
      "num_tokens": 53284260.0,
      "step": 24330
    },
    {
      "entropy": 5.6722441673278805,
      "epoch": 2.415144898769353,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044192753930008267,
      "loss": 5.0315,
      "mean_token_accuracy": 0.17950193732976913,
      "num_tokens": 53295818.0,
      "step": 24335
    },
    {
      "entropy": 5.666093111038208,
      "epoch": 2.4156411274315204,
      "grad_norm": 1.0,
      "learning_rate": 0.0004419037800455765,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18509826213121414,
      "num_tokens": 53305576.0,
      "step": 24340
    },
    {
      "entropy": 5.596221017837524,
      "epoch": 2.416137356093688,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004418800166522163,
      "loss": 5.0226,
      "mean_token_accuracy": 0.18519390672445296,
      "num_tokens": 53317218.0,
      "step": 24345
    },
    {
      "entropy": 5.6318745613098145,
      "epoch": 2.4166335847558553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044185624912059125,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1821242243051529,
      "num_tokens": 53328688.0,
      "step": 24350
    },
    {
      "entropy": 5.638358211517334,
      "epoch": 2.417129813418023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004418324774512908,
      "loss": 5.0509,
      "mean_token_accuracy": 0.18068932592868805,
      "num_tokens": 53339806.0,
      "step": 24355
    },
    {
      "entropy": 5.674987888336181,
      "epoch": 2.4176260420801907,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004418087016449043,
      "loss": 5.0,
      "mean_token_accuracy": 0.18818468749523162,
      "num_tokens": 53350088.0,
      "step": 24360
    },
    {
      "entropy": 5.620306777954101,
      "epoch": 2.418122270742358,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044178492170202153,
      "loss": 4.9572,
      "mean_token_accuracy": 0.19192886650562285,
      "num_tokens": 53360390.0,
      "step": 24365
    },
    {
      "entropy": 5.66402678489685,
      "epoch": 2.4186184994045257,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044176113762323213,
      "loss": 5.0476,
      "mean_token_accuracy": 0.18452104777097703,
      "num_tokens": 53371250.0,
      "step": 24370
    },
    {
      "entropy": 5.67554292678833,
      "epoch": 2.419114728066693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044173734940912586,
      "loss": 5.0338,
      "mean_token_accuracy": 0.1750217631459236,
      "num_tokens": 53382228.0,
      "step": 24375
    },
    {
      "entropy": 5.701251888275147,
      "epoch": 2.4196109567288606,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044171355706029247,
      "loss": 5.1736,
      "mean_token_accuracy": 0.17791528552770614,
      "num_tokens": 53393383.0,
      "step": 24380
    },
    {
      "entropy": 5.611854553222656,
      "epoch": 2.420107185391028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044168976057732225,
      "loss": 5.0541,
      "mean_token_accuracy": 0.17979892641305922,
      "num_tokens": 53404311.0,
      "step": 24385
    },
    {
      "entropy": 5.6642450332641605,
      "epoch": 2.4206034140531956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004416659599608051,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18580833524465562,
      "num_tokens": 53415730.0,
      "step": 24390
    },
    {
      "entropy": 5.649228858947754,
      "epoch": 2.4210996427153635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004416421552113312,
      "loss": 4.9925,
      "mean_token_accuracy": 0.1853975385427475,
      "num_tokens": 53426190.0,
      "step": 24395
    },
    {
      "entropy": 5.672480869293213,
      "epoch": 2.421595871377531,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000441618346329491,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18361989110708238,
      "num_tokens": 53437430.0,
      "step": 24400
    },
    {
      "entropy": 5.670510911941529,
      "epoch": 2.4220921000396984,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044159453331587493,
      "loss": 5.0299,
      "mean_token_accuracy": 0.18686124682426453,
      "num_tokens": 53446645.0,
      "step": 24405
    },
    {
      "entropy": 5.641959953308105,
      "epoch": 2.422588328701866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044157071617107325,
      "loss": 4.9867,
      "mean_token_accuracy": 0.1893565833568573,
      "num_tokens": 53456609.0,
      "step": 24410
    },
    {
      "entropy": 5.622661828994751,
      "epoch": 2.4230845573640334,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044154689489567684,
      "loss": 4.9906,
      "mean_token_accuracy": 0.18073417842388154,
      "num_tokens": 53467235.0,
      "step": 24415
    },
    {
      "entropy": 5.687197208404541,
      "epoch": 2.423580786026201,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044152306949027625,
      "loss": 5.0922,
      "mean_token_accuracy": 0.17960855066776277,
      "num_tokens": 53477447.0,
      "step": 24420
    },
    {
      "entropy": 5.711556386947632,
      "epoch": 2.4240770146883683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044149923995546244,
      "loss": 5.0823,
      "mean_token_accuracy": 0.18254781812429427,
      "num_tokens": 53491104.0,
      "step": 24425
    },
    {
      "entropy": 5.765154600143433,
      "epoch": 2.424573243350536,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004414754062918262,
      "loss": 5.1199,
      "mean_token_accuracy": 0.174993833899498,
      "num_tokens": 53501906.0,
      "step": 24430
    },
    {
      "entropy": 5.635650205612182,
      "epoch": 2.4250694720127033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004414515684999587,
      "loss": 5.102,
      "mean_token_accuracy": 0.18678320795297623,
      "num_tokens": 53512125.0,
      "step": 24435
    },
    {
      "entropy": 5.686468696594238,
      "epoch": 2.4255657006748708,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004414277265804508,
      "loss": 5.1604,
      "mean_token_accuracy": 0.17414432615041733,
      "num_tokens": 53524142.0,
      "step": 24440
    },
    {
      "entropy": 5.716044282913208,
      "epoch": 2.4260619293370387,
      "grad_norm": 1.046875,
      "learning_rate": 0.000441403880533894,
      "loss": 5.0188,
      "mean_token_accuracy": 0.17954546064138413,
      "num_tokens": 53534852.0,
      "step": 24445
    },
    {
      "entropy": 5.643224287033081,
      "epoch": 2.426558157999206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044138003036087956,
      "loss": 5.048,
      "mean_token_accuracy": 0.17883153557777404,
      "num_tokens": 53545452.0,
      "step": 24450
    },
    {
      "entropy": 5.625688076019287,
      "epoch": 2.4270543866613736,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004413561760619988,
      "loss": 5.0715,
      "mean_token_accuracy": 0.18045449554920195,
      "num_tokens": 53558101.0,
      "step": 24455
    },
    {
      "entropy": 5.686895561218262,
      "epoch": 2.427550615323541,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004413323176378434,
      "loss": 5.0924,
      "mean_token_accuracy": 0.18751779198646545,
      "num_tokens": 53568142.0,
      "step": 24460
    },
    {
      "entropy": 5.656569194793701,
      "epoch": 2.4280468439857086,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004413084550890049,
      "loss": 5.0477,
      "mean_token_accuracy": 0.17940704375505448,
      "num_tokens": 53579175.0,
      "step": 24465
    },
    {
      "entropy": 5.704124069213867,
      "epoch": 2.428543072647876,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004412845884160751,
      "loss": 5.0643,
      "mean_token_accuracy": 0.18062990307807922,
      "num_tokens": 53589861.0,
      "step": 24470
    },
    {
      "entropy": 5.7344746589660645,
      "epoch": 2.4290393013100435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004412607176196458,
      "loss": 5.1361,
      "mean_token_accuracy": 0.17469954192638398,
      "num_tokens": 53601385.0,
      "step": 24475
    },
    {
      "entropy": 5.708641624450683,
      "epoch": 2.429535529972211,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044123684270030907,
      "loss": 5.0381,
      "mean_token_accuracy": 0.18075545728206635,
      "num_tokens": 53611473.0,
      "step": 24480
    },
    {
      "entropy": 5.610617303848267,
      "epoch": 2.430031758634379,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044121296365865674,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1797149196267128,
      "num_tokens": 53623232.0,
      "step": 24485
    },
    {
      "entropy": 5.588866758346557,
      "epoch": 2.4305279872965464,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044118908049528105,
      "loss": 4.991,
      "mean_token_accuracy": 0.18917846977710723,
      "num_tokens": 53634127.0,
      "step": 24490
    },
    {
      "entropy": 5.6592811107635494,
      "epoch": 2.431024215958714,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044116519321077435,
      "loss": 4.9556,
      "mean_token_accuracy": 0.18878164142370224,
      "num_tokens": 53644471.0,
      "step": 24495
    },
    {
      "entropy": 5.649917459487915,
      "epoch": 2.4315204446208813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044114130180572883,
      "loss": 5.1008,
      "mean_token_accuracy": 0.1712856709957123,
      "num_tokens": 53656241.0,
      "step": 24500
    },
    {
      "entropy": 5.611505699157715,
      "epoch": 2.432016673283049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004411174062807371,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18064554929733276,
      "num_tokens": 53666916.0,
      "step": 24505
    },
    {
      "entropy": 5.669985914230347,
      "epoch": 2.4325129019452163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004410935066363916,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18317323327064514,
      "num_tokens": 53677473.0,
      "step": 24510
    },
    {
      "entropy": 5.626914024353027,
      "epoch": 2.4330091306073838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044106960287328503,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18211162239313125,
      "num_tokens": 53688424.0,
      "step": 24515
    },
    {
      "entropy": 5.659061527252197,
      "epoch": 2.4335053592695512,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004410456949920102,
      "loss": 5.0331,
      "mean_token_accuracy": 0.18500015288591384,
      "num_tokens": 53700390.0,
      "step": 24520
    },
    {
      "entropy": 5.583011245727539,
      "epoch": 2.4340015879317187,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044102178299315994,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18346976190805436,
      "num_tokens": 53711986.0,
      "step": 24525
    },
    {
      "entropy": 5.703500509262085,
      "epoch": 2.4344978165938866,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044099786687732713,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17991265505552292,
      "num_tokens": 53723657.0,
      "step": 24530
    },
    {
      "entropy": 5.7487555027008055,
      "epoch": 2.434994045256054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004409739466451049,
      "loss": 5.2015,
      "mean_token_accuracy": 0.17934693545103073,
      "num_tokens": 53736352.0,
      "step": 24535
    },
    {
      "entropy": 5.661588382720947,
      "epoch": 2.4354902739182216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044095002229708655,
      "loss": 5.0682,
      "mean_token_accuracy": 0.18419946283102034,
      "num_tokens": 53747799.0,
      "step": 24540
    },
    {
      "entropy": 5.651215028762818,
      "epoch": 2.435986502580389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044092609383386517,
      "loss": 5.0663,
      "mean_token_accuracy": 0.1763293594121933,
      "num_tokens": 53759263.0,
      "step": 24545
    },
    {
      "entropy": 5.67442479133606,
      "epoch": 2.4364827312425565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044090216125603426,
      "loss": 5.1254,
      "mean_token_accuracy": 0.1757407605648041,
      "num_tokens": 53770589.0,
      "step": 24550
    },
    {
      "entropy": 5.668635272979737,
      "epoch": 2.436978959904724,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004408782245641872,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18558980226516725,
      "num_tokens": 53782851.0,
      "step": 24555
    },
    {
      "entropy": 5.646641588211059,
      "epoch": 2.4374751885668915,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004408542837589176,
      "loss": 4.9927,
      "mean_token_accuracy": 0.18729879558086396,
      "num_tokens": 53793685.0,
      "step": 24560
    },
    {
      "entropy": 5.590137052536011,
      "epoch": 2.437971417229059,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044083033884081914,
      "loss": 5.0158,
      "mean_token_accuracy": 0.1776821255683899,
      "num_tokens": 53805263.0,
      "step": 24565
    },
    {
      "entropy": 5.560867738723755,
      "epoch": 2.438467645891227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004408063898104856,
      "loss": 4.9311,
      "mean_token_accuracy": 0.19269164502620698,
      "num_tokens": 53816146.0,
      "step": 24570
    },
    {
      "entropy": 5.661389541625977,
      "epoch": 2.4389638745533944,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044078243666851094,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18135936409235,
      "num_tokens": 53826823.0,
      "step": 24575
    },
    {
      "entropy": 5.659524583816529,
      "epoch": 2.439460103215562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044075847941548905,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1791445344686508,
      "num_tokens": 53837132.0,
      "step": 24580
    },
    {
      "entropy": 5.641719198226928,
      "epoch": 2.4399563318777293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044073451805201396,
      "loss": 5.0227,
      "mean_token_accuracy": 0.18049940466880798,
      "num_tokens": 53847817.0,
      "step": 24585
    },
    {
      "entropy": 5.563594055175781,
      "epoch": 2.440452560539897,
      "grad_norm": 1.0,
      "learning_rate": 0.00044071055257868,
      "loss": 4.9375,
      "mean_token_accuracy": 0.19262927174568176,
      "num_tokens": 53858678.0,
      "step": 24590
    },
    {
      "entropy": 5.595092630386352,
      "epoch": 2.4409487892020643,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044068658299608145,
      "loss": 5.0212,
      "mean_token_accuracy": 0.1751885548233986,
      "num_tokens": 53868289.0,
      "step": 24595
    },
    {
      "entropy": 5.62332706451416,
      "epoch": 2.4414450178642317,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044066260930481264,
      "loss": 5.0249,
      "mean_token_accuracy": 0.1888625979423523,
      "num_tokens": 53880167.0,
      "step": 24600
    },
    {
      "entropy": 5.652504301071167,
      "epoch": 2.441941246526399,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004406386315054681,
      "loss": 4.9929,
      "mean_token_accuracy": 0.18577285557985307,
      "num_tokens": 53890176.0,
      "step": 24605
    },
    {
      "entropy": 5.668496322631836,
      "epoch": 2.4424374751885667,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044061464959864243,
      "loss": 5.0366,
      "mean_token_accuracy": 0.18174187690019608,
      "num_tokens": 53900620.0,
      "step": 24610
    },
    {
      "entropy": 5.6065648078918455,
      "epoch": 2.4429337038507346,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004405906635849302,
      "loss": 4.9859,
      "mean_token_accuracy": 0.18841178566217423,
      "num_tokens": 53911591.0,
      "step": 24615
    },
    {
      "entropy": 5.624911642074585,
      "epoch": 2.443429932512902,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004405666734649265,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18482012748718263,
      "num_tokens": 53921425.0,
      "step": 24620
    },
    {
      "entropy": 5.710365104675293,
      "epoch": 2.4439261611750696,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044054267923922594,
      "loss": 5.0442,
      "mean_token_accuracy": 0.18216480910778046,
      "num_tokens": 53931150.0,
      "step": 24625
    },
    {
      "entropy": 5.595992612838745,
      "epoch": 2.444422389837237,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044051868090842376,
      "loss": 5.0204,
      "mean_token_accuracy": 0.183709055185318,
      "num_tokens": 53942162.0,
      "step": 24630
    },
    {
      "entropy": 5.592884969711304,
      "epoch": 2.4449186184994045,
      "grad_norm": 1.0,
      "learning_rate": 0.00044049467847311486,
      "loss": 4.9659,
      "mean_token_accuracy": 0.18416932076215745,
      "num_tokens": 53953681.0,
      "step": 24635
    },
    {
      "entropy": 5.678501558303833,
      "epoch": 2.445414847161572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004404706719338946,
      "loss": 5.0467,
      "mean_token_accuracy": 0.17912745773792266,
      "num_tokens": 53965146.0,
      "step": 24640
    },
    {
      "entropy": 5.625998163223267,
      "epoch": 2.4459110758237395,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044044666129135825,
      "loss": 4.9747,
      "mean_token_accuracy": 0.1774854615330696,
      "num_tokens": 53976445.0,
      "step": 24645
    },
    {
      "entropy": 5.670749521255493,
      "epoch": 2.446407304485907,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044042264654610115,
      "loss": 4.9864,
      "mean_token_accuracy": 0.19049571454524994,
      "num_tokens": 53987318.0,
      "step": 24650
    },
    {
      "entropy": 5.610795402526856,
      "epoch": 2.446903533148075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044039862769871884,
      "loss": 5.0078,
      "mean_token_accuracy": 0.18146617859601974,
      "num_tokens": 53998154.0,
      "step": 24655
    },
    {
      "entropy": 5.684588623046875,
      "epoch": 2.4473997618102423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044037460474980703,
      "loss": 5.0379,
      "mean_token_accuracy": 0.18823650479316711,
      "num_tokens": 54009488.0,
      "step": 24660
    },
    {
      "entropy": 5.630157613754273,
      "epoch": 2.44789599047241,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004403505776999614,
      "loss": 5.0499,
      "mean_token_accuracy": 0.18426185995340347,
      "num_tokens": 54020199.0,
      "step": 24665
    },
    {
      "entropy": 5.672284650802612,
      "epoch": 2.4483922191345773,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044032654654977766,
      "loss": 5.0194,
      "mean_token_accuracy": 0.18323244154453278,
      "num_tokens": 54031204.0,
      "step": 24670
    },
    {
      "entropy": 5.603989887237549,
      "epoch": 2.4488884477967447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004403025112998519,
      "loss": 4.9328,
      "mean_token_accuracy": 0.19293049573898316,
      "num_tokens": 54042223.0,
      "step": 24675
    },
    {
      "entropy": 5.572174453735352,
      "epoch": 2.4493846764589122,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004402784719507799,
      "loss": 4.9458,
      "mean_token_accuracy": 0.19012868255376816,
      "num_tokens": 54053307.0,
      "step": 24680
    },
    {
      "entropy": 5.659886837005615,
      "epoch": 2.4498809051210797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004402544285031581,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17935463041067123,
      "num_tokens": 54062821.0,
      "step": 24685
    },
    {
      "entropy": 5.771626377105713,
      "epoch": 2.450377133783247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004402303809575825,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1687604382634163,
      "num_tokens": 54074070.0,
      "step": 24690
    },
    {
      "entropy": 5.646166133880615,
      "epoch": 2.4508733624454146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004402063293146495,
      "loss": 4.9905,
      "mean_token_accuracy": 0.18463281095027922,
      "num_tokens": 54085399.0,
      "step": 24695
    },
    {
      "entropy": 5.692306327819824,
      "epoch": 2.451369591107582,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004401822735749555,
      "loss": 5.054,
      "mean_token_accuracy": 0.1803164705634117,
      "num_tokens": 54096102.0,
      "step": 24700
    },
    {
      "entropy": 5.6693792819976805,
      "epoch": 2.45186581976975,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044015821373909707,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18091591596603393,
      "num_tokens": 54106878.0,
      "step": 24705
    },
    {
      "entropy": 5.721264410018921,
      "epoch": 2.4523620484319175,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044013414980767084,
      "loss": 5.1261,
      "mean_token_accuracy": 0.17893434911966324,
      "num_tokens": 54119164.0,
      "step": 24710
    },
    {
      "entropy": 5.663975524902344,
      "epoch": 2.452858277094085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004401100817812735,
      "loss": 5.0103,
      "mean_token_accuracy": 0.18791523277759553,
      "num_tokens": 54129863.0,
      "step": 24715
    },
    {
      "entropy": 5.686690187454223,
      "epoch": 2.4533545057562525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044008600966050196,
      "loss": 5.1553,
      "mean_token_accuracy": 0.1753321424126625,
      "num_tokens": 54141042.0,
      "step": 24720
    },
    {
      "entropy": 5.687054014205932,
      "epoch": 2.45385073441842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004400619334459531,
      "loss": 5.0027,
      "mean_token_accuracy": 0.17502734959125518,
      "num_tokens": 54151720.0,
      "step": 24725
    },
    {
      "entropy": 5.605967998504639,
      "epoch": 2.4543469630805874,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044003785313822386,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18596750497817993,
      "num_tokens": 54162711.0,
      "step": 24730
    },
    {
      "entropy": 5.6526178359985355,
      "epoch": 2.454843191742755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044001376873791166,
      "loss": 5.0267,
      "mean_token_accuracy": 0.17904089838266374,
      "num_tokens": 54173735.0,
      "step": 24735
    },
    {
      "entropy": 5.662525987625122,
      "epoch": 2.455339420404923,
      "grad_norm": 1.0,
      "learning_rate": 0.00043998968024561337,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17332074493169786,
      "num_tokens": 54185251.0,
      "step": 24740
    },
    {
      "entropy": 5.678590917587281,
      "epoch": 2.4558356490670903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004399655876619267,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18698774129152299,
      "num_tokens": 54195915.0,
      "step": 24745
    },
    {
      "entropy": 5.730702543258667,
      "epoch": 2.4563318777292578,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043994149098744884,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17546362578868865,
      "num_tokens": 54207553.0,
      "step": 24750
    },
    {
      "entropy": 5.650018501281738,
      "epoch": 2.4568281063914252,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004399173902227775,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1798814579844475,
      "num_tokens": 54218327.0,
      "step": 24755
    },
    {
      "entropy": 5.5853735446929935,
      "epoch": 2.4573243350535927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043989328536851014,
      "loss": 4.9629,
      "mean_token_accuracy": 0.18981994241476058,
      "num_tokens": 54229256.0,
      "step": 24760
    },
    {
      "entropy": 5.6767998218536375,
      "epoch": 2.45782056371576,
      "grad_norm": 1.125,
      "learning_rate": 0.00043986917642524465,
      "loss": 5.0162,
      "mean_token_accuracy": 0.1779901221394539,
      "num_tokens": 54241241.0,
      "step": 24765
    },
    {
      "entropy": 5.674370384216308,
      "epoch": 2.4583167923779277,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043984506339357887,
      "loss": 5.0912,
      "mean_token_accuracy": 0.1775071293115616,
      "num_tokens": 54252392.0,
      "step": 24770
    },
    {
      "entropy": 5.656294488906861,
      "epoch": 2.458813021040095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004398209462741107,
      "loss": 5.0026,
      "mean_token_accuracy": 0.1835629552602768,
      "num_tokens": 54262885.0,
      "step": 24775
    },
    {
      "entropy": 5.683870697021485,
      "epoch": 2.4593092497022626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004397968250674383,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17564603835344314,
      "num_tokens": 54273963.0,
      "step": 24780
    },
    {
      "entropy": 5.653402042388916,
      "epoch": 2.45980547836443,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397726997741597,
      "loss": 4.9995,
      "mean_token_accuracy": 0.18362840563058852,
      "num_tokens": 54283837.0,
      "step": 24785
    },
    {
      "entropy": 5.635627174377442,
      "epoch": 2.460301707026598,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397485703948732,
      "loss": 4.9791,
      "mean_token_accuracy": 0.18411055356264114,
      "num_tokens": 54295752.0,
      "step": 24790
    },
    {
      "entropy": 5.7475502490997314,
      "epoch": 2.4607979356887655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004397244369301771,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1789920896291733,
      "num_tokens": 54306853.0,
      "step": 24795
    },
    {
      "entropy": 5.617394590377808,
      "epoch": 2.461294164350933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043970029938066995,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18363140523433685,
      "num_tokens": 54318624.0,
      "step": 24800
    },
    {
      "entropy": 5.594711637496948,
      "epoch": 2.4617903930131004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004396761577469503,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1818070963025093,
      "num_tokens": 54330417.0,
      "step": 24805
    },
    {
      "entropy": 5.6219470500946045,
      "epoch": 2.462286621675268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043965201202961673,
      "loss": 4.975,
      "mean_token_accuracy": 0.1890327587723732,
      "num_tokens": 54340559.0,
      "step": 24810
    },
    {
      "entropy": 5.63855471611023,
      "epoch": 2.4627828503374354,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043962786222926807,
      "loss": 4.9249,
      "mean_token_accuracy": 0.18735036849975586,
      "num_tokens": 54350907.0,
      "step": 24815
    },
    {
      "entropy": 5.613633298873902,
      "epoch": 2.463279078999603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043960370834650325,
      "loss": 4.9014,
      "mean_token_accuracy": 0.19002406746149064,
      "num_tokens": 54361742.0,
      "step": 24820
    },
    {
      "entropy": 5.5782393455505375,
      "epoch": 2.4637753076617708,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004395795503819211,
      "loss": 5.0136,
      "mean_token_accuracy": 0.17908640950918198,
      "num_tokens": 54372979.0,
      "step": 24825
    },
    {
      "entropy": 5.590333366394043,
      "epoch": 2.4642715363239382,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004395553883361207,
      "loss": 5.0341,
      "mean_token_accuracy": 0.1814708024263382,
      "num_tokens": 54383557.0,
      "step": 24830
    },
    {
      "entropy": 5.646741914749145,
      "epoch": 2.4647677649861057,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043953122220970127,
      "loss": 4.9692,
      "mean_token_accuracy": 0.1898205265402794,
      "num_tokens": 54395542.0,
      "step": 24835
    },
    {
      "entropy": 5.645487117767334,
      "epoch": 2.465263993648273,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043950705200326205,
      "loss": 5.1535,
      "mean_token_accuracy": 0.17582110166549683,
      "num_tokens": 54407540.0,
      "step": 24840
    },
    {
      "entropy": 5.707405614852905,
      "epoch": 2.4657602223104407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004394828777174025,
      "loss": 5.1043,
      "mean_token_accuracy": 0.1803954005241394,
      "num_tokens": 54419308.0,
      "step": 24845
    },
    {
      "entropy": 5.6430868148803714,
      "epoch": 2.466256450972608,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043945869935272194,
      "loss": 4.9787,
      "mean_token_accuracy": 0.18480141460895538,
      "num_tokens": 54431104.0,
      "step": 24850
    },
    {
      "entropy": 5.661783218383789,
      "epoch": 2.4667526796347756,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043943451690982006,
      "loss": 4.9837,
      "mean_token_accuracy": 0.1824508175253868,
      "num_tokens": 54441501.0,
      "step": 24855
    },
    {
      "entropy": 5.656828165054321,
      "epoch": 2.467248908296943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004394103303892964,
      "loss": 5.0029,
      "mean_token_accuracy": 0.1895984783768654,
      "num_tokens": 54452370.0,
      "step": 24860
    },
    {
      "entropy": 5.616696119308472,
      "epoch": 2.4677451369591106,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043938613979175087,
      "loss": 5.0531,
      "mean_token_accuracy": 0.17872534990310668,
      "num_tokens": 54463378.0,
      "step": 24865
    },
    {
      "entropy": 5.654262542724609,
      "epoch": 2.468241365621278,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004393619451177834,
      "loss": 5.1013,
      "mean_token_accuracy": 0.17939613461494447,
      "num_tokens": 54474878.0,
      "step": 24870
    },
    {
      "entropy": 5.658661603927612,
      "epoch": 2.468737594283446,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043933774636799365,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18381153643131257,
      "num_tokens": 54487173.0,
      "step": 24875
    },
    {
      "entropy": 5.646792984008789,
      "epoch": 2.4692338229456134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043931354354298204,
      "loss": 5.055,
      "mean_token_accuracy": 0.18258477002382278,
      "num_tokens": 54498856.0,
      "step": 24880
    },
    {
      "entropy": 5.568074560165405,
      "epoch": 2.469730051607781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004392893366433486,
      "loss": 5.0013,
      "mean_token_accuracy": 0.1882135584950447,
      "num_tokens": 54508988.0,
      "step": 24885
    },
    {
      "entropy": 5.648411464691162,
      "epoch": 2.4702262802699484,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043926512566969366,
      "loss": 5.1403,
      "mean_token_accuracy": 0.1701885551214218,
      "num_tokens": 54521500.0,
      "step": 24890
    },
    {
      "entropy": 5.671198225021362,
      "epoch": 2.470722508932116,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043924091062261743,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17271308302879335,
      "num_tokens": 54532656.0,
      "step": 24895
    },
    {
      "entropy": 5.705259513854981,
      "epoch": 2.4712187375942833,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004392166915027206,
      "loss": 5.0706,
      "mean_token_accuracy": 0.18223993331193925,
      "num_tokens": 54542621.0,
      "step": 24900
    },
    {
      "entropy": 5.627662897109985,
      "epoch": 2.471714966256451,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004391924683106037,
      "loss": 4.9959,
      "mean_token_accuracy": 0.18598848581314087,
      "num_tokens": 54554585.0,
      "step": 24905
    },
    {
      "entropy": 5.697924852371216,
      "epoch": 2.4722111949186187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004391682410468674,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16784354895353318,
      "num_tokens": 54566247.0,
      "step": 24910
    },
    {
      "entropy": 5.723388671875,
      "epoch": 2.472707423580786,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043914400971211255,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18050018399953843,
      "num_tokens": 54577619.0,
      "step": 24915
    },
    {
      "entropy": 5.664799070358276,
      "epoch": 2.4732036522429537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004391197743069398,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1818472996354103,
      "num_tokens": 54587514.0,
      "step": 24920
    },
    {
      "entropy": 5.662533235549927,
      "epoch": 2.473699880905121,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004390955348319503,
      "loss": 5.0625,
      "mean_token_accuracy": 0.1748153507709503,
      "num_tokens": 54598822.0,
      "step": 24925
    },
    {
      "entropy": 5.630574464797974,
      "epoch": 2.4741961095672886,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043907129128774525,
      "loss": 4.9791,
      "mean_token_accuracy": 0.18877365738153457,
      "num_tokens": 54608724.0,
      "step": 24930
    },
    {
      "entropy": 5.658486032485962,
      "epoch": 2.474692338229456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043904704367492563,
      "loss": 5.02,
      "mean_token_accuracy": 0.18366604894399643,
      "num_tokens": 54618241.0,
      "step": 24935
    },
    {
      "entropy": 5.679964351654053,
      "epoch": 2.4751885668916236,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004390227919940928,
      "loss": 5.0347,
      "mean_token_accuracy": 0.18342637419700622,
      "num_tokens": 54628998.0,
      "step": 24940
    },
    {
      "entropy": 5.714731073379516,
      "epoch": 2.475684795553791,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043899853624584825,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17739275395870208,
      "num_tokens": 54639198.0,
      "step": 24945
    },
    {
      "entropy": 5.7302772998809814,
      "epoch": 2.4761810242159585,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043897427643079336,
      "loss": 5.0823,
      "mean_token_accuracy": 0.1843975231051445,
      "num_tokens": 54650308.0,
      "step": 24950
    },
    {
      "entropy": 5.758327960968018,
      "epoch": 2.476677252878126,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043895001254952963,
      "loss": 5.1525,
      "mean_token_accuracy": 0.17396498769521712,
      "num_tokens": 54662566.0,
      "step": 24955
    },
    {
      "entropy": 5.706288528442383,
      "epoch": 2.477173481540294,
      "grad_norm": 1.21875,
      "learning_rate": 0.000438925744602659,
      "loss": 5.0986,
      "mean_token_accuracy": 0.1819269061088562,
      "num_tokens": 54672624.0,
      "step": 24960
    },
    {
      "entropy": 5.670654964447022,
      "epoch": 2.4776697102024614,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043890147259078314,
      "loss": 5.0645,
      "mean_token_accuracy": 0.18253753185272217,
      "num_tokens": 54684506.0,
      "step": 24965
    },
    {
      "entropy": 5.754410076141357,
      "epoch": 2.478165938864629,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043887719651450384,
      "loss": 5.0986,
      "mean_token_accuracy": 0.1715438187122345,
      "num_tokens": 54694923.0,
      "step": 24970
    },
    {
      "entropy": 5.706668043136597,
      "epoch": 2.4786621675267964,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004388529163744233,
      "loss": 5.0879,
      "mean_token_accuracy": 0.17793770432472228,
      "num_tokens": 54706032.0,
      "step": 24975
    },
    {
      "entropy": 5.665470266342163,
      "epoch": 2.479158396188964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004388286321711435,
      "loss": 5.0165,
      "mean_token_accuracy": 0.18673141896724701,
      "num_tokens": 54717236.0,
      "step": 24980
    },
    {
      "entropy": 5.698895692825317,
      "epoch": 2.4796546248511313,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004388043439052665,
      "loss": 5.0729,
      "mean_token_accuracy": 0.1732577860355377,
      "num_tokens": 54727178.0,
      "step": 24985
    },
    {
      "entropy": 5.667464685440064,
      "epoch": 2.480150853513299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043878005157739486,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17712474316358567,
      "num_tokens": 54738060.0,
      "step": 24990
    },
    {
      "entropy": 5.674487447738647,
      "epoch": 2.4806470821754667,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004387557551881308,
      "loss": 4.9903,
      "mean_token_accuracy": 0.18518027663230896,
      "num_tokens": 54749487.0,
      "step": 24995
    },
    {
      "entropy": 5.6833597183227536,
      "epoch": 2.481143310837634,
      "grad_norm": 1.078125,
      "learning_rate": 0.000438731454738077,
      "loss": 5.0543,
      "mean_token_accuracy": 0.17680521458387374,
      "num_tokens": 54760681.0,
      "step": 25000
    },
    {
      "entropy": 5.645579195022583,
      "epoch": 2.4816395394998016,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043870715022783575,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18283092826604844,
      "num_tokens": 54774064.0,
      "step": 25005
    },
    {
      "entropy": 5.684694051742554,
      "epoch": 2.482135768161969,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004386828416580101,
      "loss": 5.0728,
      "mean_token_accuracy": 0.17949719429016114,
      "num_tokens": 54786368.0,
      "step": 25010
    },
    {
      "entropy": 5.6876781463623045,
      "epoch": 2.4826319968241366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004386585290292026,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17881639152765275,
      "num_tokens": 54796831.0,
      "step": 25015
    },
    {
      "entropy": 5.630071306228638,
      "epoch": 2.483128225486304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004386342123420161,
      "loss": 5.0966,
      "mean_token_accuracy": 0.1790160447359085,
      "num_tokens": 54806897.0,
      "step": 25020
    },
    {
      "entropy": 5.6190978527069095,
      "epoch": 2.4836244541484715,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043860989159705396,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1935818925499916,
      "num_tokens": 54818877.0,
      "step": 25025
    },
    {
      "entropy": 5.71186933517456,
      "epoch": 2.484120682810639,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043858556679491896,
      "loss": 5.0709,
      "mean_token_accuracy": 0.18580865859985352,
      "num_tokens": 54830581.0,
      "step": 25030
    },
    {
      "entropy": 5.619103717803955,
      "epoch": 2.4846169114728065,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043856123793621437,
      "loss": 4.9619,
      "mean_token_accuracy": 0.18389064818620682,
      "num_tokens": 54841344.0,
      "step": 25035
    },
    {
      "entropy": 5.593067169189453,
      "epoch": 2.485113140134974,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043853690502154355,
      "loss": 5.0141,
      "mean_token_accuracy": 0.1809844568371773,
      "num_tokens": 54853333.0,
      "step": 25040
    },
    {
      "entropy": 5.689177751541138,
      "epoch": 2.485609368797142,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004385125680515099,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18669671714305877,
      "num_tokens": 54864461.0,
      "step": 25045
    },
    {
      "entropy": 5.593482828140258,
      "epoch": 2.4861055974593094,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004384882270267168,
      "loss": 4.908,
      "mean_token_accuracy": 0.19126821607351302,
      "num_tokens": 54874766.0,
      "step": 25050
    },
    {
      "entropy": 5.588213491439819,
      "epoch": 2.486601826121477,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004384638819477681,
      "loss": 5.0033,
      "mean_token_accuracy": 0.1845671057701111,
      "num_tokens": 54885771.0,
      "step": 25055
    },
    {
      "entropy": 5.664577579498291,
      "epoch": 2.4870980547836443,
      "grad_norm": 1.125,
      "learning_rate": 0.00043843953281526727,
      "loss": 5.0288,
      "mean_token_accuracy": 0.1820132002234459,
      "num_tokens": 54897233.0,
      "step": 25060
    },
    {
      "entropy": 5.669129371643066,
      "epoch": 2.487594283445812,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043841517962981823,
      "loss": 5.0317,
      "mean_token_accuracy": 0.18504746705293657,
      "num_tokens": 54907767.0,
      "step": 25065
    },
    {
      "entropy": 5.678694581985473,
      "epoch": 2.4880905121079793,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043839082239202496,
      "loss": 5.0124,
      "mean_token_accuracy": 0.1836460903286934,
      "num_tokens": 54918279.0,
      "step": 25070
    },
    {
      "entropy": 5.656321430206299,
      "epoch": 2.4885867407701467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004383664611024912,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18660361766815187,
      "num_tokens": 54928762.0,
      "step": 25075
    },
    {
      "entropy": 5.612805271148682,
      "epoch": 2.489082969432314,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004383420957618213,
      "loss": 5.0068,
      "mean_token_accuracy": 0.18839598298072815,
      "num_tokens": 54940471.0,
      "step": 25080
    },
    {
      "entropy": 5.663314294815064,
      "epoch": 2.489579198094482,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043831772637061944,
      "loss": 5.0743,
      "mean_token_accuracy": 0.17877398580312728,
      "num_tokens": 54951359.0,
      "step": 25085
    },
    {
      "entropy": 5.7372753620147705,
      "epoch": 2.4900754267566496,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043829335292948986,
      "loss": 5.0988,
      "mean_token_accuracy": 0.18309610784053804,
      "num_tokens": 54963062.0,
      "step": 25090
    },
    {
      "entropy": 5.7597826480865475,
      "epoch": 2.490571655418817,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043826897543903696,
      "loss": 5.1024,
      "mean_token_accuracy": 0.18859989941120148,
      "num_tokens": 54974074.0,
      "step": 25095
    },
    {
      "entropy": 5.646685171127319,
      "epoch": 2.4910678840809846,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043824459389986534,
      "loss": 5.0798,
      "mean_token_accuracy": 0.17983355671167373,
      "num_tokens": 54985543.0,
      "step": 25100
    },
    {
      "entropy": 5.644846200942993,
      "epoch": 2.491564112743152,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004382202083125795,
      "loss": 5.0733,
      "mean_token_accuracy": 0.1711309149861336,
      "num_tokens": 54995939.0,
      "step": 25105
    },
    {
      "entropy": 5.655153703689575,
      "epoch": 2.4920603414053195,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004381958186777842,
      "loss": 5.089,
      "mean_token_accuracy": 0.17902976870536805,
      "num_tokens": 55006627.0,
      "step": 25110
    },
    {
      "entropy": 5.683953809738159,
      "epoch": 2.492556570067487,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004381714249960843,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1842508941888809,
      "num_tokens": 55016821.0,
      "step": 25115
    },
    {
      "entropy": 5.601080703735351,
      "epoch": 2.4930527987296545,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004381470272680846,
      "loss": 4.9932,
      "mean_token_accuracy": 0.18019702434539794,
      "num_tokens": 55026938.0,
      "step": 25120
    },
    {
      "entropy": 5.644254541397094,
      "epoch": 2.493549027391822,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004381226254943902,
      "loss": 5.0157,
      "mean_token_accuracy": 0.17956891506910325,
      "num_tokens": 55039257.0,
      "step": 25125
    },
    {
      "entropy": 5.726852178573608,
      "epoch": 2.49404525605399,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004380982196756062,
      "loss": 5.1118,
      "mean_token_accuracy": 0.18183813542127608,
      "num_tokens": 55049199.0,
      "step": 25130
    },
    {
      "entropy": 5.70598669052124,
      "epoch": 2.4945414847161573,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043807380981233777,
      "loss": 5.078,
      "mean_token_accuracy": 0.18143346458673476,
      "num_tokens": 55060471.0,
      "step": 25135
    },
    {
      "entropy": 5.577668523788452,
      "epoch": 2.495037713378325,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004380493959051903,
      "loss": 4.9881,
      "mean_token_accuracy": 0.187234503030777,
      "num_tokens": 55071332.0,
      "step": 25140
    },
    {
      "entropy": 5.667330741882324,
      "epoch": 2.4955339420404923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043802497795476916,
      "loss": 5.0753,
      "mean_token_accuracy": 0.1799657791852951,
      "num_tokens": 55082823.0,
      "step": 25145
    },
    {
      "entropy": 5.660717630386353,
      "epoch": 2.4960301707026598,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004380005559616798,
      "loss": 5.0573,
      "mean_token_accuracy": 0.1795022413134575,
      "num_tokens": 55094130.0,
      "step": 25150
    },
    {
      "entropy": 5.665185785293579,
      "epoch": 2.4965263993648272,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004379761299265279,
      "loss": 5.0399,
      "mean_token_accuracy": 0.18143978267908095,
      "num_tokens": 55104984.0,
      "step": 25155
    },
    {
      "entropy": 5.678394794464111,
      "epoch": 2.4970226280269947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004379516998499192,
      "loss": 5.0651,
      "mean_token_accuracy": 0.17720877379179,
      "num_tokens": 55116691.0,
      "step": 25160
    },
    {
      "entropy": 5.649364900588989,
      "epoch": 2.497518856689162,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004379272657324594,
      "loss": 4.9934,
      "mean_token_accuracy": 0.1883747711777687,
      "num_tokens": 55126605.0,
      "step": 25165
    },
    {
      "entropy": 5.67249813079834,
      "epoch": 2.49801508535133,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004379028275747545,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18573211133480072,
      "num_tokens": 55137110.0,
      "step": 25170
    },
    {
      "entropy": 5.692864847183228,
      "epoch": 2.4985113140134976,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043787838537741054,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18546133041381835,
      "num_tokens": 55147416.0,
      "step": 25175
    },
    {
      "entropy": 5.654462099075317,
      "epoch": 2.499007542675665,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004378539391410336,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18523830473423003,
      "num_tokens": 55159643.0,
      "step": 25180
    },
    {
      "entropy": 5.603913831710815,
      "epoch": 2.4995037713378325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043782948886622987,
      "loss": 4.9856,
      "mean_token_accuracy": 0.18738296777009963,
      "num_tokens": 55170196.0,
      "step": 25185
    },
    {
      "entropy": 5.633758687973023,
      "epoch": 2.5,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004378050345536057,
      "loss": 5.0355,
      "mean_token_accuracy": 0.1833956852555275,
      "num_tokens": 55181272.0,
      "step": 25190
    },
    {
      "entropy": 5.71118311882019,
      "epoch": 2.5004962286621675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043778057620376745,
      "loss": 5.0566,
      "mean_token_accuracy": 0.1829541102051735,
      "num_tokens": 55191126.0,
      "step": 25195
    },
    {
      "entropy": 5.636980104446411,
      "epoch": 2.500992457324335,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043775611381732164,
      "loss": 4.9652,
      "mean_token_accuracy": 0.19063429832458495,
      "num_tokens": 55201000.0,
      "step": 25200
    },
    {
      "entropy": 5.718400287628174,
      "epoch": 2.5014886859865024,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000437731647394875,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1783851459622383,
      "num_tokens": 55211479.0,
      "step": 25205
    },
    {
      "entropy": 5.694533252716065,
      "epoch": 2.50198491464867,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004377071769370341,
      "loss": 4.9975,
      "mean_token_accuracy": 0.18743504136800765,
      "num_tokens": 55220824.0,
      "step": 25210
    },
    {
      "entropy": 5.619998264312744,
      "epoch": 2.5024811433108374,
      "grad_norm": 1.125,
      "learning_rate": 0.00043768270244440587,
      "loss": 5.001,
      "mean_token_accuracy": 0.18323302417993545,
      "num_tokens": 55230857.0,
      "step": 25215
    },
    {
      "entropy": 5.622949647903442,
      "epoch": 2.5029773719730053,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043765822391759714,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18415959626436235,
      "num_tokens": 55241611.0,
      "step": 25220
    },
    {
      "entropy": 5.67461085319519,
      "epoch": 2.5034736006351728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004376337413572149,
      "loss": 5.0968,
      "mean_token_accuracy": 0.1807187393307686,
      "num_tokens": 55252410.0,
      "step": 25225
    },
    {
      "entropy": 5.66388463973999,
      "epoch": 2.5039698292973402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043760925476386635,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1836072325706482,
      "num_tokens": 55262426.0,
      "step": 25230
    },
    {
      "entropy": 5.711055183410645,
      "epoch": 2.5044660579595077,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004375847641381587,
      "loss": 5.1161,
      "mean_token_accuracy": 0.17919648438692093,
      "num_tokens": 55273240.0,
      "step": 25235
    },
    {
      "entropy": 5.665049457550049,
      "epoch": 2.504962286621675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004375602694806993,
      "loss": 5.0469,
      "mean_token_accuracy": 0.1839945748448372,
      "num_tokens": 55284210.0,
      "step": 25240
    },
    {
      "entropy": 5.667417716979981,
      "epoch": 2.5054585152838427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043753577079209537,
      "loss": 4.9537,
      "mean_token_accuracy": 0.19455762058496476,
      "num_tokens": 55293752.0,
      "step": 25245
    },
    {
      "entropy": 5.678518486022949,
      "epoch": 2.5059547439460106,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043751126807295464,
      "loss": 5.1203,
      "mean_token_accuracy": 0.18698941469192504,
      "num_tokens": 55305260.0,
      "step": 25250
    },
    {
      "entropy": 5.759114027023315,
      "epoch": 2.506450972608178,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043748676132388464,
      "loss": 5.1336,
      "mean_token_accuracy": 0.18146060705184935,
      "num_tokens": 55316283.0,
      "step": 25255
    },
    {
      "entropy": 5.708069896697998,
      "epoch": 2.5069472012703455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043746225054549316,
      "loss": 5.1457,
      "mean_token_accuracy": 0.17224950939416886,
      "num_tokens": 55327817.0,
      "step": 25260
    },
    {
      "entropy": 5.661601305007935,
      "epoch": 2.507443429932513,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043743773573838785,
      "loss": 5.0832,
      "mean_token_accuracy": 0.18342555165290833,
      "num_tokens": 55338406.0,
      "step": 25265
    },
    {
      "entropy": 5.690545177459716,
      "epoch": 2.5079396585946805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043741321690317673,
      "loss": 5.1222,
      "mean_token_accuracy": 0.17292115539312364,
      "num_tokens": 55349297.0,
      "step": 25270
    },
    {
      "entropy": 5.684371995925903,
      "epoch": 2.508435887256848,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004373886940404677,
      "loss": 5.0271,
      "mean_token_accuracy": 0.17735394686460496,
      "num_tokens": 55360229.0,
      "step": 25275
    },
    {
      "entropy": 5.640713310241699,
      "epoch": 2.5089321159190154,
      "grad_norm": 1.0,
      "learning_rate": 0.0004373641671508691,
      "loss": 5.0051,
      "mean_token_accuracy": 0.1890163838863373,
      "num_tokens": 55371570.0,
      "step": 25280
    },
    {
      "entropy": 5.6147760391235355,
      "epoch": 2.509428344581183,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000437339636234989,
      "loss": 4.9793,
      "mean_token_accuracy": 0.18692427426576613,
      "num_tokens": 55382781.0,
      "step": 25285
    },
    {
      "entropy": 5.663873672485352,
      "epoch": 2.5099245732433504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043731510129343574,
      "loss": 5.0115,
      "mean_token_accuracy": 0.17802907079458236,
      "num_tokens": 55394631.0,
      "step": 25290
    },
    {
      "entropy": 5.683970785140991,
      "epoch": 2.510420801905518,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043729056232681766,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19016171246767044,
      "num_tokens": 55406057.0,
      "step": 25295
    },
    {
      "entropy": 5.630306386947632,
      "epoch": 2.5109170305676853,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004372660193357434,
      "loss": 5.0032,
      "mean_token_accuracy": 0.18268244117498397,
      "num_tokens": 55416096.0,
      "step": 25300
    },
    {
      "entropy": 5.660558271408081,
      "epoch": 2.5114132592298533,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004372414723208215,
      "loss": 5.0078,
      "mean_token_accuracy": 0.1854439601302147,
      "num_tokens": 55427522.0,
      "step": 25305
    },
    {
      "entropy": 5.6349945068359375,
      "epoch": 2.5119094878920207,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004372169212826606,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18022428900003434,
      "num_tokens": 55438762.0,
      "step": 25310
    },
    {
      "entropy": 5.661237287521362,
      "epoch": 2.512405716554188,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043719236622186976,
      "loss": 5.0334,
      "mean_token_accuracy": 0.18107508718967438,
      "num_tokens": 55449552.0,
      "step": 25315
    },
    {
      "entropy": 5.638271331787109,
      "epoch": 2.5129019452163557,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004371678071390576,
      "loss": 5.0772,
      "mean_token_accuracy": 0.17720600813627244,
      "num_tokens": 55459610.0,
      "step": 25320
    },
    {
      "entropy": 5.643656873703003,
      "epoch": 2.513398173878523,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043714324403483333,
      "loss": 4.9971,
      "mean_token_accuracy": 0.1832373097538948,
      "num_tokens": 55470212.0,
      "step": 25325
    },
    {
      "entropy": 5.586993646621704,
      "epoch": 2.5138944025406906,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043711867690980605,
      "loss": 4.9466,
      "mean_token_accuracy": 0.19346727132797242,
      "num_tokens": 55482639.0,
      "step": 25330
    },
    {
      "entropy": 5.612970352172852,
      "epoch": 2.5143906312028586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043709410576458477,
      "loss": 4.9293,
      "mean_token_accuracy": 0.18825414627790452,
      "num_tokens": 55493088.0,
      "step": 25335
    },
    {
      "entropy": 5.654910469055176,
      "epoch": 2.514886859865026,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043706953059977907,
      "loss": 5.0605,
      "mean_token_accuracy": 0.1875940352678299,
      "num_tokens": 55503641.0,
      "step": 25340
    },
    {
      "entropy": 5.582834911346436,
      "epoch": 2.5153830885271935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043704495141599816,
      "loss": 4.8904,
      "mean_token_accuracy": 0.18542776107788086,
      "num_tokens": 55514433.0,
      "step": 25345
    },
    {
      "entropy": 5.577947473526001,
      "epoch": 2.515879317189361,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004370203682138517,
      "loss": 4.9326,
      "mean_token_accuracy": 0.18672901391983032,
      "num_tokens": 55524646.0,
      "step": 25350
    },
    {
      "entropy": 5.650780868530274,
      "epoch": 2.5163755458515285,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004369957809939492,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1808781549334526,
      "num_tokens": 55535587.0,
      "step": 25355
    },
    {
      "entropy": 5.674287176132202,
      "epoch": 2.516871774513696,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004369711897569004,
      "loss": 5.007,
      "mean_token_accuracy": 0.17933329641819,
      "num_tokens": 55545861.0,
      "step": 25360
    },
    {
      "entropy": 5.674876165390015,
      "epoch": 2.5173680031758634,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004369465945033151,
      "loss": 5.064,
      "mean_token_accuracy": 0.18010154366493225,
      "num_tokens": 55556930.0,
      "step": 25365
    },
    {
      "entropy": 5.684883308410645,
      "epoch": 2.517864231838031,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043692199523380325,
      "loss": 5.09,
      "mean_token_accuracy": 0.1827279269695282,
      "num_tokens": 55570783.0,
      "step": 25370
    },
    {
      "entropy": 5.594750595092774,
      "epoch": 2.5183604605001984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004368973919489748,
      "loss": 5.0032,
      "mean_token_accuracy": 0.18594694137573242,
      "num_tokens": 55582123.0,
      "step": 25375
    },
    {
      "entropy": 5.6418296813964846,
      "epoch": 2.518856689162366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043687278464943986,
      "loss": 4.9977,
      "mean_token_accuracy": 0.18400147557258606,
      "num_tokens": 55592977.0,
      "step": 25380
    },
    {
      "entropy": 5.645345640182495,
      "epoch": 2.5193529178245333,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004368481733358088,
      "loss": 4.9953,
      "mean_token_accuracy": 0.18896136432886124,
      "num_tokens": 55603179.0,
      "step": 25385
    },
    {
      "entropy": 5.650879716873169,
      "epoch": 2.519849146486701,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004368235580086917,
      "loss": 5.0472,
      "mean_token_accuracy": 0.17799067944288255,
      "num_tokens": 55616144.0,
      "step": 25390
    },
    {
      "entropy": 5.67500114440918,
      "epoch": 2.5203453751488687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043679893866869904,
      "loss": 5.129,
      "mean_token_accuracy": 0.17890845239162445,
      "num_tokens": 55626862.0,
      "step": 25395
    },
    {
      "entropy": 5.652491092681885,
      "epoch": 2.520841603811036,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004367743153164414,
      "loss": 4.9892,
      "mean_token_accuracy": 0.18198488503694535,
      "num_tokens": 55637564.0,
      "step": 25400
    },
    {
      "entropy": 5.6125462532043455,
      "epoch": 2.5213378324732036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043674968795252926,
      "loss": 4.9615,
      "mean_token_accuracy": 0.19039101749658585,
      "num_tokens": 55647642.0,
      "step": 25405
    },
    {
      "entropy": 5.684770917892456,
      "epoch": 2.521834061135371,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004367250565775734,
      "loss": 5.1318,
      "mean_token_accuracy": 0.1782332107424736,
      "num_tokens": 55659126.0,
      "step": 25410
    },
    {
      "entropy": 5.670426321029663,
      "epoch": 2.5223302897975386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004367004211921847,
      "loss": 4.9842,
      "mean_token_accuracy": 0.18452599346637727,
      "num_tokens": 55669270.0,
      "step": 25415
    },
    {
      "entropy": 5.600512599945068,
      "epoch": 2.522826518459706,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043667578179697387,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18643377274274825,
      "num_tokens": 55680478.0,
      "step": 25420
    },
    {
      "entropy": 5.67880482673645,
      "epoch": 2.523322747121874,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004366511383925521,
      "loss": 5.0518,
      "mean_token_accuracy": 0.179998679459095,
      "num_tokens": 55691228.0,
      "step": 25425
    },
    {
      "entropy": 5.731458568572998,
      "epoch": 2.5238189757840415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004366264909795305,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1773786410689354,
      "num_tokens": 55702676.0,
      "step": 25430
    },
    {
      "entropy": 5.656477308273315,
      "epoch": 2.524315204446209,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004366018395585201,
      "loss": 5.0522,
      "mean_token_accuracy": 0.18807232826948167,
      "num_tokens": 55711643.0,
      "step": 25435
    },
    {
      "entropy": 5.692053413391113,
      "epoch": 2.5248114331083764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004365771841301323,
      "loss": 5.101,
      "mean_token_accuracy": 0.18204228281974794,
      "num_tokens": 55722478.0,
      "step": 25440
    },
    {
      "entropy": 5.671212863922119,
      "epoch": 2.525307661770544,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004365525246949786,
      "loss": 5.0107,
      "mean_token_accuracy": 0.18533118516206742,
      "num_tokens": 55733489.0,
      "step": 25445
    },
    {
      "entropy": 5.646311283111572,
      "epoch": 2.5258038904327114,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043652786125367035,
      "loss": 5.0141,
      "mean_token_accuracy": 0.1832287237048149,
      "num_tokens": 55745824.0,
      "step": 25450
    },
    {
      "entropy": 5.610022735595703,
      "epoch": 2.526300119094879,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043650319380681915,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18612879067659377,
      "num_tokens": 55757024.0,
      "step": 25455
    },
    {
      "entropy": 5.677986288070679,
      "epoch": 2.5267963477570463,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004364785223550368,
      "loss": 5.1048,
      "mean_token_accuracy": 0.17867791056632995,
      "num_tokens": 55768674.0,
      "step": 25460
    },
    {
      "entropy": 5.685248184204101,
      "epoch": 2.527292576419214,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043645384689893504,
      "loss": 5.067,
      "mean_token_accuracy": 0.1826762467622757,
      "num_tokens": 55779226.0,
      "step": 25465
    },
    {
      "entropy": 5.737293529510498,
      "epoch": 2.5277888050813813,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004364291674391258,
      "loss": 5.1312,
      "mean_token_accuracy": 0.1748372122645378,
      "num_tokens": 55790470.0,
      "step": 25470
    },
    {
      "entropy": 5.740451002120972,
      "epoch": 2.528285033743549,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000436404483976221,
      "loss": 5.1159,
      "mean_token_accuracy": 0.18109044134616853,
      "num_tokens": 55801215.0,
      "step": 25475
    },
    {
      "entropy": 5.6935444355010985,
      "epoch": 2.5287812624057167,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043637979651083285,
      "loss": 5.1153,
      "mean_token_accuracy": 0.18102347999811172,
      "num_tokens": 55812663.0,
      "step": 25480
    },
    {
      "entropy": 5.62253623008728,
      "epoch": 2.529277491067884,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043635510504357343,
      "loss": 4.9618,
      "mean_token_accuracy": 0.19322903007268905,
      "num_tokens": 55822337.0,
      "step": 25485
    },
    {
      "entropy": 5.659409141540527,
      "epoch": 2.5297737197300516,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043633040957505515,
      "loss": 5.1139,
      "mean_token_accuracy": 0.1769360825419426,
      "num_tokens": 55833578.0,
      "step": 25490
    },
    {
      "entropy": 5.681674528121948,
      "epoch": 2.530269948392219,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004363057101058903,
      "loss": 4.9732,
      "mean_token_accuracy": 0.18965887725353242,
      "num_tokens": 55843768.0,
      "step": 25495
    },
    {
      "entropy": 5.708784008026123,
      "epoch": 2.5307661770543866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004362810066366915,
      "loss": 5.0238,
      "mean_token_accuracy": 0.17806546241045,
      "num_tokens": 55853879.0,
      "step": 25500
    },
    {
      "entropy": 5.583788299560547,
      "epoch": 2.531262405716554,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043625629916807117,
      "loss": 5.0486,
      "mean_token_accuracy": 0.18131475299596786,
      "num_tokens": 55864606.0,
      "step": 25505
    },
    {
      "entropy": 5.613346576690674,
      "epoch": 2.531758634378722,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043623158770064214,
      "loss": 4.9434,
      "mean_token_accuracy": 0.18963708430528642,
      "num_tokens": 55875036.0,
      "step": 25510
    },
    {
      "entropy": 5.623200130462647,
      "epoch": 2.5322548630408894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004362068722350172,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18050894886255264,
      "num_tokens": 55884924.0,
      "step": 25515
    },
    {
      "entropy": 5.645705699920654,
      "epoch": 2.532751091703057,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043618215277180917,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18138808608055115,
      "num_tokens": 55894975.0,
      "step": 25520
    },
    {
      "entropy": 5.640945529937744,
      "epoch": 2.5332473203652244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043615742931163106,
      "loss": 4.9861,
      "mean_token_accuracy": 0.1882346510887146,
      "num_tokens": 55905325.0,
      "step": 25525
    },
    {
      "entropy": 5.6875128746032715,
      "epoch": 2.533743549027392,
      "grad_norm": 1.0625,
      "learning_rate": 0.000436132701855096,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18376206308603288,
      "num_tokens": 55916816.0,
      "step": 25530
    },
    {
      "entropy": 5.574372053146362,
      "epoch": 2.5342397776895593,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043610797040281705,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18286829590797424,
      "num_tokens": 55928725.0,
      "step": 25535
    },
    {
      "entropy": 5.694140625,
      "epoch": 2.534736006351727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004360832349554077,
      "loss": 5.0817,
      "mean_token_accuracy": 0.18617503345012665,
      "num_tokens": 55939555.0,
      "step": 25540
    },
    {
      "entropy": 5.664679098129272,
      "epoch": 2.5352322350138943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004360584955134811,
      "loss": 4.932,
      "mean_token_accuracy": 0.18666796684265136,
      "num_tokens": 55949944.0,
      "step": 25545
    },
    {
      "entropy": 5.669880390167236,
      "epoch": 2.5357284636760618,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043603375207765103,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1855031281709671,
      "num_tokens": 55959549.0,
      "step": 25550
    },
    {
      "entropy": 5.601847362518311,
      "epoch": 2.5362246923382292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004360090046485308,
      "loss": 5.0114,
      "mean_token_accuracy": 0.1877799302339554,
      "num_tokens": 55970088.0,
      "step": 25555
    },
    {
      "entropy": 5.704870986938476,
      "epoch": 2.536720921000397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004359842532267342,
      "loss": 5.0339,
      "mean_token_accuracy": 0.19221299588680268,
      "num_tokens": 55980485.0,
      "step": 25560
    },
    {
      "entropy": 5.716379642486572,
      "epoch": 2.5372171496625646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043595949781287505,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18533414602279663,
      "num_tokens": 55991577.0,
      "step": 25565
    },
    {
      "entropy": 5.643041658401489,
      "epoch": 2.537713378324732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043593473840756715,
      "loss": 4.9967,
      "mean_token_accuracy": 0.18206588923931122,
      "num_tokens": 56002813.0,
      "step": 25570
    },
    {
      "entropy": 5.760383224487304,
      "epoch": 2.5382096069868996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004359099750114245,
      "loss": 5.1578,
      "mean_token_accuracy": 0.18257914781570433,
      "num_tokens": 56013646.0,
      "step": 25575
    },
    {
      "entropy": 5.715669918060303,
      "epoch": 2.538705835649067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004358852076250613,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18118494302034377,
      "num_tokens": 56024777.0,
      "step": 25580
    },
    {
      "entropy": 5.670663547515869,
      "epoch": 2.5392020643112345,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043586043624909153,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18317019790410996,
      "num_tokens": 56035550.0,
      "step": 25585
    },
    {
      "entropy": 5.620978689193725,
      "epoch": 2.539698292973402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004358356608841296,
      "loss": 5.0101,
      "mean_token_accuracy": 0.17996984869241714,
      "num_tokens": 56046835.0,
      "step": 25590
    },
    {
      "entropy": 5.601159477233887,
      "epoch": 2.54019452163557,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043581088153078984,
      "loss": 4.9999,
      "mean_token_accuracy": 0.18884214758872986,
      "num_tokens": 56056404.0,
      "step": 25595
    },
    {
      "entropy": 5.682380199432373,
      "epoch": 2.5406907502977374,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043578609818968676,
      "loss": 5.0469,
      "mean_token_accuracy": 0.18129748553037645,
      "num_tokens": 56066690.0,
      "step": 25600
    },
    {
      "entropy": 5.735978603363037,
      "epoch": 2.541186978959905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043576131086143496,
      "loss": 5.1277,
      "mean_token_accuracy": 0.17694250196218492,
      "num_tokens": 56076051.0,
      "step": 25605
    },
    {
      "entropy": 5.687660980224609,
      "epoch": 2.5416832076220723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043573651954664897,
      "loss": 5.002,
      "mean_token_accuracy": 0.18550646007061006,
      "num_tokens": 56086601.0,
      "step": 25610
    },
    {
      "entropy": 5.604254293441772,
      "epoch": 2.54217943628424,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043571172424594363,
      "loss": 5.0089,
      "mean_token_accuracy": 0.18162765651941298,
      "num_tokens": 56097954.0,
      "step": 25615
    },
    {
      "entropy": 5.580146980285645,
      "epoch": 2.5426756649464073,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004356869249599339,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18654119223356247,
      "num_tokens": 56109729.0,
      "step": 25620
    },
    {
      "entropy": 5.68748106956482,
      "epoch": 2.5431718936085748,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004356621216892347,
      "loss": 5.0172,
      "mean_token_accuracy": 0.18518364876508714,
      "num_tokens": 56120231.0,
      "step": 25625
    },
    {
      "entropy": 5.671230411529541,
      "epoch": 2.5436681222707422,
      "grad_norm": 1.125,
      "learning_rate": 0.00043563731443446095,
      "loss": 5.037,
      "mean_token_accuracy": 0.17789935916662217,
      "num_tokens": 56130558.0,
      "step": 25630
    },
    {
      "entropy": 5.785156631469727,
      "epoch": 2.5441643509329097,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043561250319622806,
      "loss": 5.1269,
      "mean_token_accuracy": 0.17575464993715287,
      "num_tokens": 56142540.0,
      "step": 25635
    },
    {
      "entropy": 5.671662712097168,
      "epoch": 2.544660579595077,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043558768797515116,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17934560775756836,
      "num_tokens": 56153802.0,
      "step": 25640
    },
    {
      "entropy": 5.722948026657105,
      "epoch": 2.5451568082572447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043556286877184566,
      "loss": 5.073,
      "mean_token_accuracy": 0.18161123543977736,
      "num_tokens": 56164990.0,
      "step": 25645
    },
    {
      "entropy": 5.658890438079834,
      "epoch": 2.5456530369194126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004355380455869269,
      "loss": 5.1009,
      "mean_token_accuracy": 0.17856732606887818,
      "num_tokens": 56176472.0,
      "step": 25650
    },
    {
      "entropy": 5.644011688232422,
      "epoch": 2.54614926558158,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004355132184210106,
      "loss": 5.0485,
      "mean_token_accuracy": 0.18162305206060408,
      "num_tokens": 56187504.0,
      "step": 25655
    },
    {
      "entropy": 5.637327337265015,
      "epoch": 2.5466454942437475,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004354883872747124,
      "loss": 5.0311,
      "mean_token_accuracy": 0.1794920966029167,
      "num_tokens": 56198214.0,
      "step": 25660
    },
    {
      "entropy": 5.702619552612305,
      "epoch": 2.547141722905915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043546355214864796,
      "loss": 5.0562,
      "mean_token_accuracy": 0.17969711869955063,
      "num_tokens": 56209885.0,
      "step": 25665
    },
    {
      "entropy": 5.640521001815796,
      "epoch": 2.5476379515680825,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004354387130434332,
      "loss": 5.025,
      "mean_token_accuracy": 0.1811523348093033,
      "num_tokens": 56221281.0,
      "step": 25670
    },
    {
      "entropy": 5.633408927917481,
      "epoch": 2.54813418023025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004354138699596841,
      "loss": 5.0308,
      "mean_token_accuracy": 0.1860833376646042,
      "num_tokens": 56231914.0,
      "step": 25675
    },
    {
      "entropy": 5.639281797409057,
      "epoch": 2.548630408892418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004353890228980167,
      "loss": 4.9918,
      "mean_token_accuracy": 0.17956048399209976,
      "num_tokens": 56242308.0,
      "step": 25680
    },
    {
      "entropy": 5.631120777130127,
      "epoch": 2.5491266375545854,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004353641718590471,
      "loss": 5.0083,
      "mean_token_accuracy": 0.17523447275161744,
      "num_tokens": 56252570.0,
      "step": 25685
    },
    {
      "entropy": 5.719456720352173,
      "epoch": 2.549622866216753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043533931684339156,
      "loss": 5.0364,
      "mean_token_accuracy": 0.18597699254751204,
      "num_tokens": 56263669.0,
      "step": 25690
    },
    {
      "entropy": 5.677528381347656,
      "epoch": 2.5501190948789203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004353144578516665,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18025257885456086,
      "num_tokens": 56273869.0,
      "step": 25695
    },
    {
      "entropy": 5.65227313041687,
      "epoch": 2.5506153235410878,
      "grad_norm": 1.0,
      "learning_rate": 0.00043528959488448835,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18967438042163848,
      "num_tokens": 56285670.0,
      "step": 25700
    },
    {
      "entropy": 5.748519372940064,
      "epoch": 2.5511115522032553,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043526472794247365,
      "loss": 5.0834,
      "mean_token_accuracy": 0.18124571889638902,
      "num_tokens": 56296579.0,
      "step": 25705
    },
    {
      "entropy": 5.725867652893067,
      "epoch": 2.5516077808654227,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043523985702623896,
      "loss": 5.0849,
      "mean_token_accuracy": 0.1834245055913925,
      "num_tokens": 56307879.0,
      "step": 25710
    },
    {
      "entropy": 5.683178520202636,
      "epoch": 2.55210400952759,
      "grad_norm": 1.25,
      "learning_rate": 0.00043521498213640113,
      "loss": 5.0268,
      "mean_token_accuracy": 0.18063794821500778,
      "num_tokens": 56318848.0,
      "step": 25715
    },
    {
      "entropy": 5.707958698272705,
      "epoch": 2.5526002381897577,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043519010327357704,
      "loss": 5.1089,
      "mean_token_accuracy": 0.18326107859611512,
      "num_tokens": 56332002.0,
      "step": 25720
    },
    {
      "entropy": 5.661526250839233,
      "epoch": 2.553096466851925,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004351652204383836,
      "loss": 5.084,
      "mean_token_accuracy": 0.1779334530234337,
      "num_tokens": 56342323.0,
      "step": 25725
    },
    {
      "entropy": 5.74315299987793,
      "epoch": 2.5535926955140926,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004351403336314377,
      "loss": 5.0679,
      "mean_token_accuracy": 0.17721349447965623,
      "num_tokens": 56354925.0,
      "step": 25730
    },
    {
      "entropy": 5.653876733779907,
      "epoch": 2.5540889241762605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004351154428533566,
      "loss": 4.9721,
      "mean_token_accuracy": 0.18177264630794526,
      "num_tokens": 56366987.0,
      "step": 25735
    },
    {
      "entropy": 5.6108269691467285,
      "epoch": 2.554585152838428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004350905481047576,
      "loss": 5.0344,
      "mean_token_accuracy": 0.183129945397377,
      "num_tokens": 56379496.0,
      "step": 25740
    },
    {
      "entropy": 5.629112815856933,
      "epoch": 2.5550813815005955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004350656493862579,
      "loss": 4.8854,
      "mean_token_accuracy": 0.196013405919075,
      "num_tokens": 56391132.0,
      "step": 25745
    },
    {
      "entropy": 5.751727342605591,
      "epoch": 2.555577610162763,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043504074669847517,
      "loss": 5.1569,
      "mean_token_accuracy": 0.17282678782939911,
      "num_tokens": 56402687.0,
      "step": 25750
    },
    {
      "entropy": 5.55720100402832,
      "epoch": 2.5560738388249304,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004350158400420266,
      "loss": 4.9634,
      "mean_token_accuracy": 0.18183744549751282,
      "num_tokens": 56414202.0,
      "step": 25755
    },
    {
      "entropy": 5.661267518997192,
      "epoch": 2.556570067487098,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043499092941753003,
      "loss": 5.0614,
      "mean_token_accuracy": 0.1818167433142662,
      "num_tokens": 56424949.0,
      "step": 25760
    },
    {
      "entropy": 5.723602819442749,
      "epoch": 2.557066296149266,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043496601482560327,
      "loss": 5.1546,
      "mean_token_accuracy": 0.17481906563043595,
      "num_tokens": 56437185.0,
      "step": 25765
    },
    {
      "entropy": 5.744333982467651,
      "epoch": 2.5575625248114333,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043494109626686394,
      "loss": 5.0525,
      "mean_token_accuracy": 0.17697134464979172,
      "num_tokens": 56447646.0,
      "step": 25770
    },
    {
      "entropy": 5.653592300415039,
      "epoch": 2.558058753473601,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043491617374193017,
      "loss": 5.0646,
      "mean_token_accuracy": 0.18254243731498718,
      "num_tokens": 56458245.0,
      "step": 25775
    },
    {
      "entropy": 5.660891914367676,
      "epoch": 2.5585549821357683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004348912472514197,
      "loss": 4.9865,
      "mean_token_accuracy": 0.1835293471813202,
      "num_tokens": 56470186.0,
      "step": 25780
    },
    {
      "entropy": 5.675621843338012,
      "epoch": 2.5590512107979357,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043486631679595097,
      "loss": 5.089,
      "mean_token_accuracy": 0.18124013394117355,
      "num_tokens": 56481332.0,
      "step": 25785
    },
    {
      "entropy": 5.598939037322998,
      "epoch": 2.559547439460103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004348413823761421,
      "loss": 4.9671,
      "mean_token_accuracy": 0.18697798252105713,
      "num_tokens": 56491700.0,
      "step": 25790
    },
    {
      "entropy": 5.671095180511474,
      "epoch": 2.5600436681222707,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004348164439926112,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18817038387060164,
      "num_tokens": 56502780.0,
      "step": 25795
    },
    {
      "entropy": 5.6422138690948485,
      "epoch": 2.560539896784438,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043479150164597695,
      "loss": 5.0518,
      "mean_token_accuracy": 0.1856871947646141,
      "num_tokens": 56515039.0,
      "step": 25800
    },
    {
      "entropy": 5.687790775299073,
      "epoch": 2.5610361254466056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043476655533685783,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17510657012462616,
      "num_tokens": 56527705.0,
      "step": 25805
    },
    {
      "entropy": 5.6365080833435055,
      "epoch": 2.561532354108773,
      "grad_norm": 1.125,
      "learning_rate": 0.00043474160506587234,
      "loss": 4.9813,
      "mean_token_accuracy": 0.188120599091053,
      "num_tokens": 56537748.0,
      "step": 25810
    },
    {
      "entropy": 5.638648271560669,
      "epoch": 2.5620285827709406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004347166508336393,
      "loss": 4.9675,
      "mean_token_accuracy": 0.18231754750013351,
      "num_tokens": 56548804.0,
      "step": 25815
    },
    {
      "entropy": 5.660673332214356,
      "epoch": 2.5625248114331085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043469169264077754,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18314376175403596,
      "num_tokens": 56561245.0,
      "step": 25820
    },
    {
      "entropy": 5.634836673736572,
      "epoch": 2.563021040095276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004346667304879058,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18589188158512115,
      "num_tokens": 56570823.0,
      "step": 25825
    },
    {
      "entropy": 5.611022043228149,
      "epoch": 2.5635172687574435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043464176437564317,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18364454209804534,
      "num_tokens": 56581439.0,
      "step": 25830
    },
    {
      "entropy": 5.683574628829956,
      "epoch": 2.564013497419611,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004346167943046089,
      "loss": 5.065,
      "mean_token_accuracy": 0.18210402131080627,
      "num_tokens": 56593016.0,
      "step": 25835
    },
    {
      "entropy": 5.575800085067749,
      "epoch": 2.5645097260817784,
      "grad_norm": 0.96875,
      "learning_rate": 0.000434591820275422,
      "loss": 4.9403,
      "mean_token_accuracy": 0.1862596720457077,
      "num_tokens": 56603701.0,
      "step": 25840
    },
    {
      "entropy": 5.626632976531982,
      "epoch": 2.565005954743946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345668422887018,
      "loss": 5.0017,
      "mean_token_accuracy": 0.17568336576223373,
      "num_tokens": 56614781.0,
      "step": 25845
    },
    {
      "entropy": 5.598658561706543,
      "epoch": 2.565502183406114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004345418603450678,
      "loss": 4.9341,
      "mean_token_accuracy": 0.18475325554609298,
      "num_tokens": 56625477.0,
      "step": 25850
    },
    {
      "entropy": 5.62488751411438,
      "epoch": 2.5659984120682813,
      "grad_norm": 1.125,
      "learning_rate": 0.0004345168744451395,
      "loss": 4.998,
      "mean_token_accuracy": 0.18958120197057723,
      "num_tokens": 56635230.0,
      "step": 25855
    },
    {
      "entropy": 5.669651174545288,
      "epoch": 2.5664946407304488,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004344918845895364,
      "loss": 4.982,
      "mean_token_accuracy": 0.18734031766653061,
      "num_tokens": 56645777.0,
      "step": 25860
    },
    {
      "entropy": 5.65584397315979,
      "epoch": 2.5669908693926162,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004344668907788782,
      "loss": 5.1214,
      "mean_token_accuracy": 0.18161870688199996,
      "num_tokens": 56656790.0,
      "step": 25865
    },
    {
      "entropy": 5.627488279342652,
      "epoch": 2.5674870980547837,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004344418930137848,
      "loss": 5.042,
      "mean_token_accuracy": 0.18178726136684417,
      "num_tokens": 56668771.0,
      "step": 25870
    },
    {
      "entropy": 5.6444817066192625,
      "epoch": 2.567983326716951,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004344168912948759,
      "loss": 5.0334,
      "mean_token_accuracy": 0.1817369654774666,
      "num_tokens": 56680437.0,
      "step": 25875
    },
    {
      "entropy": 5.691593265533447,
      "epoch": 2.5684795553791187,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043439188562277174,
      "loss": 5.0182,
      "mean_token_accuracy": 0.1825493171811104,
      "num_tokens": 56691152.0,
      "step": 25880
    },
    {
      "entropy": 5.703454399108887,
      "epoch": 2.568975784041286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043436687599809226,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1807217612862587,
      "num_tokens": 56701677.0,
      "step": 25885
    },
    {
      "entropy": 5.616486883163452,
      "epoch": 2.5694720127034536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043434186242145763,
      "loss": 4.9763,
      "mean_token_accuracy": 0.1932992309331894,
      "num_tokens": 56712166.0,
      "step": 25890
    },
    {
      "entropy": 5.658292579650879,
      "epoch": 2.569968241365621,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004343168448934881,
      "loss": 5.0159,
      "mean_token_accuracy": 0.1889220505952835,
      "num_tokens": 56722010.0,
      "step": 25895
    },
    {
      "entropy": 5.611346960067749,
      "epoch": 2.5704644700277886,
      "grad_norm": 1.0,
      "learning_rate": 0.00043429182341480423,
      "loss": 4.9903,
      "mean_token_accuracy": 0.18231963515281677,
      "num_tokens": 56732368.0,
      "step": 25900
    },
    {
      "entropy": 5.624818944931031,
      "epoch": 2.5709606986899565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004342667979860263,
      "loss": 5.109,
      "mean_token_accuracy": 0.1762375384569168,
      "num_tokens": 56745084.0,
      "step": 25905
    },
    {
      "entropy": 5.682452821731568,
      "epoch": 2.571456927352124,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043424176860777495,
      "loss": 5.0194,
      "mean_token_accuracy": 0.18106290847063064,
      "num_tokens": 56756184.0,
      "step": 25910
    },
    {
      "entropy": 5.62312159538269,
      "epoch": 2.5719531560142914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043421673528067095,
      "loss": 4.9941,
      "mean_token_accuracy": 0.18956553041934968,
      "num_tokens": 56767505.0,
      "step": 25915
    },
    {
      "entropy": 5.620361280441284,
      "epoch": 2.572449384676459,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043419169800533496,
      "loss": 5.0147,
      "mean_token_accuracy": 0.17729537785053254,
      "num_tokens": 56778660.0,
      "step": 25920
    },
    {
      "entropy": 5.652346563339234,
      "epoch": 2.5729456133386264,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004341666567823879,
      "loss": 5.0593,
      "mean_token_accuracy": 0.1809963971376419,
      "num_tokens": 56787892.0,
      "step": 25925
    },
    {
      "entropy": 5.659970235824585,
      "epoch": 2.573441842000794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004341416116124508,
      "loss": 5.1004,
      "mean_token_accuracy": 0.1768239051103592,
      "num_tokens": 56797910.0,
      "step": 25930
    },
    {
      "entropy": 5.672247743606567,
      "epoch": 2.5739380706629618,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004341165624961445,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17954781353473664,
      "num_tokens": 56807975.0,
      "step": 25935
    },
    {
      "entropy": 5.652756881713867,
      "epoch": 2.5744342993251292,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043409150943409035,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18433890193700792,
      "num_tokens": 56817540.0,
      "step": 25940
    },
    {
      "entropy": 5.676446056365966,
      "epoch": 2.5749305279872967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043406645242690967,
      "loss": 5.0259,
      "mean_token_accuracy": 0.18350195288658142,
      "num_tokens": 56828917.0,
      "step": 25945
    },
    {
      "entropy": 5.584841299057007,
      "epoch": 2.575426756649464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340413914752237,
      "loss": 4.9426,
      "mean_token_accuracy": 0.1905482068657875,
      "num_tokens": 56839888.0,
      "step": 25950
    },
    {
      "entropy": 5.561910724639892,
      "epoch": 2.5759229853116317,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004340163265796539,
      "loss": 4.9579,
      "mean_token_accuracy": 0.18707791119813919,
      "num_tokens": 56850191.0,
      "step": 25955
    },
    {
      "entropy": 5.597445487976074,
      "epoch": 2.576419213973799,
      "grad_norm": 1.125,
      "learning_rate": 0.0004339912577408218,
      "loss": 5.0079,
      "mean_token_accuracy": 0.1862395092844963,
      "num_tokens": 56861020.0,
      "step": 25960
    },
    {
      "entropy": 5.60824728012085,
      "epoch": 2.5769154426359666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004339661849593492,
      "loss": 4.967,
      "mean_token_accuracy": 0.18432794362306595,
      "num_tokens": 56871184.0,
      "step": 25965
    },
    {
      "entropy": 5.706402063369751,
      "epoch": 2.577411671298134,
      "grad_norm": 1.125,
      "learning_rate": 0.0004339411082358577,
      "loss": 5.1425,
      "mean_token_accuracy": 0.17467774897813798,
      "num_tokens": 56882081.0,
      "step": 25970
    },
    {
      "entropy": 5.737648391723633,
      "epoch": 2.5779078999603016,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043391602757096923,
      "loss": 5.0895,
      "mean_token_accuracy": 0.18277254998683928,
      "num_tokens": 56891594.0,
      "step": 25975
    },
    {
      "entropy": 5.716228675842285,
      "epoch": 2.578404128622469,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043389094296530565,
      "loss": 5.0942,
      "mean_token_accuracy": 0.18420978933572768,
      "num_tokens": 56903636.0,
      "step": 25980
    },
    {
      "entropy": 5.606277370452881,
      "epoch": 2.5789003572846365,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004338658544194891,
      "loss": 5.0619,
      "mean_token_accuracy": 0.1836351439356804,
      "num_tokens": 56915776.0,
      "step": 25985
    },
    {
      "entropy": 5.681915140151977,
      "epoch": 2.5793965859468044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043384076193414176,
      "loss": 5.0734,
      "mean_token_accuracy": 0.18072013556957245,
      "num_tokens": 56926851.0,
      "step": 25990
    },
    {
      "entropy": 5.681267404556275,
      "epoch": 2.579892814608972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043381566550988564,
      "loss": 5.0748,
      "mean_token_accuracy": 0.18130657225847244,
      "num_tokens": 56938373.0,
      "step": 25995
    },
    {
      "entropy": 5.703997373580933,
      "epoch": 2.5803890432711394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043379056514734334,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18238714784383775,
      "num_tokens": 56948891.0,
      "step": 26000
    },
    {
      "entropy": 5.618568420410156,
      "epoch": 2.580885271933307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004337654608471372,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18255245983600615,
      "num_tokens": 56961091.0,
      "step": 26005
    },
    {
      "entropy": 5.647027015686035,
      "epoch": 2.5813815005954743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043374035260988964,
      "loss": 5.0724,
      "mean_token_accuracy": 0.1757749453186989,
      "num_tokens": 56972073.0,
      "step": 26010
    },
    {
      "entropy": 5.575292348861694,
      "epoch": 2.581877729257642,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043371524043622344,
      "loss": 4.9397,
      "mean_token_accuracy": 0.19045234322547913,
      "num_tokens": 56981591.0,
      "step": 26015
    },
    {
      "entropy": 5.6275838851928714,
      "epoch": 2.5823739579198093,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004336901243267614,
      "loss": 4.9964,
      "mean_token_accuracy": 0.18611325621604918,
      "num_tokens": 56992045.0,
      "step": 26020
    },
    {
      "entropy": 5.662784099578857,
      "epoch": 2.582870186581977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000433665004282126,
      "loss": 5.0394,
      "mean_token_accuracy": 0.18273120671510695,
      "num_tokens": 57003108.0,
      "step": 26025
    },
    {
      "entropy": 5.610775470733643,
      "epoch": 2.5833664152441447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043363988030294055,
      "loss": 4.9557,
      "mean_token_accuracy": 0.18760842680931092,
      "num_tokens": 57014446.0,
      "step": 26030
    },
    {
      "entropy": 5.674784803390503,
      "epoch": 2.583862643906312,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004336147523898279,
      "loss": 5.0163,
      "mean_token_accuracy": 0.18551266193389893,
      "num_tokens": 57025758.0,
      "step": 26035
    },
    {
      "entropy": 5.63719072341919,
      "epoch": 2.5843588725684796,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004335896205434111,
      "loss": 4.9763,
      "mean_token_accuracy": 0.184026899933815,
      "num_tokens": 57036060.0,
      "step": 26040
    },
    {
      "entropy": 5.622169494628906,
      "epoch": 2.584855101230647,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043356448476431354,
      "loss": 5.0189,
      "mean_token_accuracy": 0.1781878262758255,
      "num_tokens": 57045567.0,
      "step": 26045
    },
    {
      "entropy": 5.574440813064575,
      "epoch": 2.5853513298928146,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004335393450531583,
      "loss": 4.9668,
      "mean_token_accuracy": 0.19716715961694717,
      "num_tokens": 57056942.0,
      "step": 26050
    },
    {
      "entropy": 5.652566719055176,
      "epoch": 2.585847558554982,
      "grad_norm": 1.0625,
      "learning_rate": 0.000433514201410569,
      "loss": 4.979,
      "mean_token_accuracy": 0.18578595966100692,
      "num_tokens": 57068651.0,
      "step": 26055
    },
    {
      "entropy": 5.7379007816314695,
      "epoch": 2.5863437872171495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043348905383716905,
      "loss": 5.1351,
      "mean_token_accuracy": 0.178755921125412,
      "num_tokens": 57080503.0,
      "step": 26060
    },
    {
      "entropy": 5.626857423782349,
      "epoch": 2.586840015879317,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043346390233358213,
      "loss": 4.9929,
      "mean_token_accuracy": 0.1845652148127556,
      "num_tokens": 57090807.0,
      "step": 26065
    },
    {
      "entropy": 5.621971464157104,
      "epoch": 2.5873362445414845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334387469004319,
      "loss": 5.0542,
      "mean_token_accuracy": 0.18081264048814774,
      "num_tokens": 57102636.0,
      "step": 26070
    },
    {
      "entropy": 5.667103815078735,
      "epoch": 2.5878324732036524,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004334135875383421,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18297615945339202,
      "num_tokens": 57113124.0,
      "step": 26075
    },
    {
      "entropy": 5.640835571289062,
      "epoch": 2.58832870186582,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004333884242479367,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18019091039896012,
      "num_tokens": 57124286.0,
      "step": 26080
    },
    {
      "entropy": 5.625841903686523,
      "epoch": 2.5888249305279873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043336325702983966,
      "loss": 5.0361,
      "mean_token_accuracy": 0.1807262569665909,
      "num_tokens": 57137024.0,
      "step": 26085
    },
    {
      "entropy": 5.657196807861328,
      "epoch": 2.589321159190155,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043333808588467504,
      "loss": 5.016,
      "mean_token_accuracy": 0.186787511408329,
      "num_tokens": 57147720.0,
      "step": 26090
    },
    {
      "entropy": 5.594150972366333,
      "epoch": 2.5898173878523223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004333129108130671,
      "loss": 4.9612,
      "mean_token_accuracy": 0.1881217271089554,
      "num_tokens": 57159156.0,
      "step": 26095
    },
    {
      "entropy": 5.705224323272705,
      "epoch": 2.5903136165144898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004332877318156402,
      "loss": 5.0986,
      "mean_token_accuracy": 0.18393683135509492,
      "num_tokens": 57170737.0,
      "step": 26100
    },
    {
      "entropy": 5.653520298004151,
      "epoch": 2.5908098451766572,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043326254889301856,
      "loss": 5.0184,
      "mean_token_accuracy": 0.1836725652217865,
      "num_tokens": 57181548.0,
      "step": 26105
    },
    {
      "entropy": 5.708379125595092,
      "epoch": 2.591306073838825,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004332373620458268,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17821245193481444,
      "num_tokens": 57192661.0,
      "step": 26110
    },
    {
      "entropy": 5.694961452484131,
      "epoch": 2.5918023025009926,
      "grad_norm": 1.0,
      "learning_rate": 0.00043321217127468935,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18495347797870637,
      "num_tokens": 57205108.0,
      "step": 26115
    },
    {
      "entropy": 5.739652109146118,
      "epoch": 2.59229853116316,
      "grad_norm": 1.25,
      "learning_rate": 0.00043318697658023097,
      "loss": 5.1131,
      "mean_token_accuracy": 0.17817084789276122,
      "num_tokens": 57215995.0,
      "step": 26120
    },
    {
      "entropy": 5.648296594619751,
      "epoch": 2.5927947598253276,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004331617779630765,
      "loss": 4.9906,
      "mean_token_accuracy": 0.18678855746984482,
      "num_tokens": 57227008.0,
      "step": 26125
    },
    {
      "entropy": 5.632290363311768,
      "epoch": 2.593290988487495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004331365754238507,
      "loss": 4.9787,
      "mean_token_accuracy": 0.18365381956100463,
      "num_tokens": 57238279.0,
      "step": 26130
    },
    {
      "entropy": 5.629273557662964,
      "epoch": 2.5937872171496625,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043311136896317855,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18256262689828873,
      "num_tokens": 57247811.0,
      "step": 26135
    },
    {
      "entropy": 5.671795463562011,
      "epoch": 2.59428344581183,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043308615858168517,
      "loss": 4.9758,
      "mean_token_accuracy": 0.1858594611287117,
      "num_tokens": 57256850.0,
      "step": 26140
    },
    {
      "entropy": 5.5723412990570065,
      "epoch": 2.5947796744739975,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004330609442799957,
      "loss": 5.023,
      "mean_token_accuracy": 0.18616122603416443,
      "num_tokens": 57267079.0,
      "step": 26145
    },
    {
      "entropy": 5.60983715057373,
      "epoch": 2.595275903136165,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004330357260587355,
      "loss": 5.0646,
      "mean_token_accuracy": 0.18331596106290818,
      "num_tokens": 57279122.0,
      "step": 26150
    },
    {
      "entropy": 5.7035260677337645,
      "epoch": 2.5957721317983324,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004330105039185297,
      "loss": 5.0386,
      "mean_token_accuracy": 0.18164402842521668,
      "num_tokens": 57289601.0,
      "step": 26155
    },
    {
      "entropy": 5.737919092178345,
      "epoch": 2.5962683604605004,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043298527786000393,
      "loss": 5.0857,
      "mean_token_accuracy": 0.18162143975496292,
      "num_tokens": 57299942.0,
      "step": 26160
    },
    {
      "entropy": 5.702292108535767,
      "epoch": 2.596764589122668,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004329600478837837,
      "loss": 5.0644,
      "mean_token_accuracy": 0.18574406653642656,
      "num_tokens": 57310983.0,
      "step": 26165
    },
    {
      "entropy": 5.639982843399048,
      "epoch": 2.5972608177848353,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043293481399049474,
      "loss": 4.9931,
      "mean_token_accuracy": 0.18306512534618377,
      "num_tokens": 57321724.0,
      "step": 26170
    },
    {
      "entropy": 5.6489753246307375,
      "epoch": 2.597757046447003,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004329095761807626,
      "loss": 5.0321,
      "mean_token_accuracy": 0.17672781199216842,
      "num_tokens": 57332222.0,
      "step": 26175
    },
    {
      "entropy": 5.651493263244629,
      "epoch": 2.5982532751091703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004328843344552133,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1878826931118965,
      "num_tokens": 57342956.0,
      "step": 26180
    },
    {
      "entropy": 5.63323278427124,
      "epoch": 2.5987495037713377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004328590888144726,
      "loss": 5.0042,
      "mean_token_accuracy": 0.1851015105843544,
      "num_tokens": 57355328.0,
      "step": 26185
    },
    {
      "entropy": 5.61380352973938,
      "epoch": 2.599245732433505,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004328338392591668,
      "loss": 5.0218,
      "mean_token_accuracy": 0.18558930307626725,
      "num_tokens": 57366533.0,
      "step": 26190
    },
    {
      "entropy": 5.759679985046387,
      "epoch": 2.599741961095673,
      "grad_norm": 1.125,
      "learning_rate": 0.00043280858578992185,
      "loss": 5.1264,
      "mean_token_accuracy": 0.17373338341712952,
      "num_tokens": 57377150.0,
      "step": 26195
    },
    {
      "entropy": 5.692174911499023,
      "epoch": 2.6002381897578406,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043278332840736397,
      "loss": 5.0115,
      "mean_token_accuracy": 0.18026281595230104,
      "num_tokens": 57388372.0,
      "step": 26200
    },
    {
      "entropy": 5.67551474571228,
      "epoch": 2.600734418420008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043275806711211957,
      "loss": 5.0128,
      "mean_token_accuracy": 0.17928636521101,
      "num_tokens": 57399613.0,
      "step": 26205
    },
    {
      "entropy": 5.6208006858825685,
      "epoch": 2.6012306470821756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043273280190481503,
      "loss": 5.0128,
      "mean_token_accuracy": 0.18587885797023773,
      "num_tokens": 57410117.0,
      "step": 26210
    },
    {
      "entropy": 5.635821056365967,
      "epoch": 2.601726875744343,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043270753278607687,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18278530836105347,
      "num_tokens": 57421511.0,
      "step": 26215
    },
    {
      "entropy": 5.684484243392944,
      "epoch": 2.6022231044065105,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043268225975653176,
      "loss": 5.0998,
      "mean_token_accuracy": 0.17143428027629853,
      "num_tokens": 57432785.0,
      "step": 26220
    },
    {
      "entropy": 5.6766644477844235,
      "epoch": 2.602719333068678,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004326569828168064,
      "loss": 5.0855,
      "mean_token_accuracy": 0.1794573739171028,
      "num_tokens": 57443394.0,
      "step": 26225
    },
    {
      "entropy": 5.704372215270996,
      "epoch": 2.6032155617308455,
      "grad_norm": 1.0,
      "learning_rate": 0.0004326317019675275,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17290879786014557,
      "num_tokens": 57454245.0,
      "step": 26230
    },
    {
      "entropy": 5.65001654624939,
      "epoch": 2.603711790393013,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043260641720932214,
      "loss": 4.9786,
      "mean_token_accuracy": 0.185720057785511,
      "num_tokens": 57465954.0,
      "step": 26235
    },
    {
      "entropy": 5.669500398635864,
      "epoch": 2.6042080190551804,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043258112854281726,
      "loss": 5.0495,
      "mean_token_accuracy": 0.17718528509140014,
      "num_tokens": 57477599.0,
      "step": 26240
    },
    {
      "entropy": 5.636776638031006,
      "epoch": 2.6047042477173483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004325558359686399,
      "loss": 4.9978,
      "mean_token_accuracy": 0.18852779269218445,
      "num_tokens": 57489043.0,
      "step": 26245
    },
    {
      "entropy": 5.591740846633911,
      "epoch": 2.605200476379516,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043253053948741734,
      "loss": 4.9775,
      "mean_token_accuracy": 0.18724198192358016,
      "num_tokens": 57499349.0,
      "step": 26250
    },
    {
      "entropy": 5.602311134338379,
      "epoch": 2.6056967050416833,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043250523909977684,
      "loss": 5.0438,
      "mean_token_accuracy": 0.18310852646827697,
      "num_tokens": 57510873.0,
      "step": 26255
    },
    {
      "entropy": 5.637062215805054,
      "epoch": 2.6061929337038507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004324799348063457,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18151203244924546,
      "num_tokens": 57522282.0,
      "step": 26260
    },
    {
      "entropy": 5.674783658981323,
      "epoch": 2.606689162366018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004324546266077518,
      "loss": 4.9978,
      "mean_token_accuracy": 0.1801411598920822,
      "num_tokens": 57532760.0,
      "step": 26265
    },
    {
      "entropy": 5.682670974731446,
      "epoch": 2.6071853910281857,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004324293145046222,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18443471044301987,
      "num_tokens": 57543373.0,
      "step": 26270
    },
    {
      "entropy": 5.638040208816529,
      "epoch": 2.607681619690353,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043240399849758484,
      "loss": 4.9697,
      "mean_token_accuracy": 0.18625985234975814,
      "num_tokens": 57553221.0,
      "step": 26275
    },
    {
      "entropy": 5.6777490139007565,
      "epoch": 2.608177848352521,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004323786785872675,
      "loss": 5.0292,
      "mean_token_accuracy": 0.18025047183036805,
      "num_tokens": 57563080.0,
      "step": 26280
    },
    {
      "entropy": 5.685750579833984,
      "epoch": 2.6086740770146886,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004323533547742981,
      "loss": 5.1253,
      "mean_token_accuracy": 0.18031857758760453,
      "num_tokens": 57573556.0,
      "step": 26285
    },
    {
      "entropy": 5.7270674228668215,
      "epoch": 2.609170305676856,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004323280270593046,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17756412029266358,
      "num_tokens": 57584750.0,
      "step": 26290
    },
    {
      "entropy": 5.715113544464112,
      "epoch": 2.6096665343390235,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043230269544291503,
      "loss": 5.0288,
      "mean_token_accuracy": 0.18316345363855363,
      "num_tokens": 57594975.0,
      "step": 26295
    },
    {
      "entropy": 5.702619457244873,
      "epoch": 2.610162763001191,
      "grad_norm": 1.0,
      "learning_rate": 0.0004322773599257575,
      "loss": 5.1007,
      "mean_token_accuracy": 0.17667917013168336,
      "num_tokens": 57606534.0,
      "step": 26300
    },
    {
      "entropy": 5.637858963012695,
      "epoch": 2.6106589916633585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043225202050846043,
      "loss": 4.9841,
      "mean_token_accuracy": 0.19012836515903472,
      "num_tokens": 57616715.0,
      "step": 26305
    },
    {
      "entropy": 5.6444416522979735,
      "epoch": 2.611155220325526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000432226677191652,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18943019658327104,
      "num_tokens": 57626658.0,
      "step": 26310
    },
    {
      "entropy": 5.64442892074585,
      "epoch": 2.6116514489876934,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043220132997596084,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1820923388004303,
      "num_tokens": 57635620.0,
      "step": 26315
    },
    {
      "entropy": 5.667735576629639,
      "epoch": 2.612147677649861,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004321759788620155,
      "loss": 5.0713,
      "mean_token_accuracy": 0.18220453560352326,
      "num_tokens": 57646584.0,
      "step": 26320
    },
    {
      "entropy": 5.6507209777832035,
      "epoch": 2.6126439063120284,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043215062385044446,
      "loss": 5.0271,
      "mean_token_accuracy": 0.17712943702936174,
      "num_tokens": 57656560.0,
      "step": 26325
    },
    {
      "entropy": 5.674764633178711,
      "epoch": 2.613140134974196,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043212526494187665,
      "loss": 5.0599,
      "mean_token_accuracy": 0.180084690451622,
      "num_tokens": 57667555.0,
      "step": 26330
    },
    {
      "entropy": 5.658495330810547,
      "epoch": 2.6136363636363638,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004320999021369408,
      "loss": 5.0218,
      "mean_token_accuracy": 0.18267727345228196,
      "num_tokens": 57678106.0,
      "step": 26335
    },
    {
      "entropy": 5.630930042266845,
      "epoch": 2.6141325922985312,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043207453543626587,
      "loss": 4.9658,
      "mean_token_accuracy": 0.18548955470323564,
      "num_tokens": 57688429.0,
      "step": 26340
    },
    {
      "entropy": 5.63034234046936,
      "epoch": 2.6146288209606987,
      "grad_norm": 1.0,
      "learning_rate": 0.00043204916484048094,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18648772090673446,
      "num_tokens": 57699179.0,
      "step": 26345
    },
    {
      "entropy": 5.627146244049072,
      "epoch": 2.615125049622866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043202379035021515,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17898908406496047,
      "num_tokens": 57710472.0,
      "step": 26350
    },
    {
      "entropy": 5.6852391242980955,
      "epoch": 2.6156212782850337,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004319984119660977,
      "loss": 5.0674,
      "mean_token_accuracy": 0.17867037504911423,
      "num_tokens": 57723063.0,
      "step": 26355
    },
    {
      "entropy": 5.760134315490722,
      "epoch": 2.616117506947201,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043197302968875795,
      "loss": 5.1749,
      "mean_token_accuracy": 0.17698613852262496,
      "num_tokens": 57734664.0,
      "step": 26360
    },
    {
      "entropy": 5.73687047958374,
      "epoch": 2.616613735609369,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004319476435188254,
      "loss": 5.0708,
      "mean_token_accuracy": 0.18331825286149978,
      "num_tokens": 57745081.0,
      "step": 26365
    },
    {
      "entropy": 5.643301057815552,
      "epoch": 2.6171099642715365,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043192225345692933,
      "loss": 5.0385,
      "mean_token_accuracy": 0.17983562648296356,
      "num_tokens": 57755074.0,
      "step": 26370
    },
    {
      "entropy": 5.666999292373657,
      "epoch": 2.617606192933704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043189685950369963,
      "loss": 5.0861,
      "mean_token_accuracy": 0.17508919090032576,
      "num_tokens": 57765503.0,
      "step": 26375
    },
    {
      "entropy": 5.718939208984375,
      "epoch": 2.6181024215958715,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043187146165976577,
      "loss": 5.0841,
      "mean_token_accuracy": 0.17882403284311293,
      "num_tokens": 57776905.0,
      "step": 26380
    },
    {
      "entropy": 5.703448629379272,
      "epoch": 2.618598650258039,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004318460599257578,
      "loss": 5.0541,
      "mean_token_accuracy": 0.18688775151968,
      "num_tokens": 57786720.0,
      "step": 26385
    },
    {
      "entropy": 5.702106523513794,
      "epoch": 2.6190948789202064,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043182065430230545,
      "loss": 5.0114,
      "mean_token_accuracy": 0.18404021710157395,
      "num_tokens": 57797171.0,
      "step": 26390
    },
    {
      "entropy": 5.666737699508667,
      "epoch": 2.619591107582374,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004317952447900388,
      "loss": 4.9726,
      "mean_token_accuracy": 0.1873207375407219,
      "num_tokens": 57807543.0,
      "step": 26395
    },
    {
      "entropy": 5.624127578735352,
      "epoch": 2.6200873362445414,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431769831389588,
      "loss": 5.065,
      "mean_token_accuracy": 0.17778669148683549,
      "num_tokens": 57818136.0,
      "step": 26400
    },
    {
      "entropy": 5.642215633392334,
      "epoch": 2.620583564906709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004317444141015831,
      "loss": 4.9974,
      "mean_token_accuracy": 0.18189473450183868,
      "num_tokens": 57830159.0,
      "step": 26405
    },
    {
      "entropy": 5.715713167190552,
      "epoch": 2.6210797935688763,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043171899292665457,
      "loss": 5.1043,
      "mean_token_accuracy": 0.18095002174377442,
      "num_tokens": 57841867.0,
      "step": 26410
    },
    {
      "entropy": 5.636816358566284,
      "epoch": 2.621576022231044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004316935678654327,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18951793015003204,
      "num_tokens": 57853982.0,
      "step": 26415
    },
    {
      "entropy": 5.621701192855835,
      "epoch": 2.6220722508932117,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043166813891854796,
      "loss": 4.944,
      "mean_token_accuracy": 0.18975506275892257,
      "num_tokens": 57866005.0,
      "step": 26420
    },
    {
      "entropy": 5.631848526000977,
      "epoch": 2.622568479555379,
      "grad_norm": 1.015625,
      "learning_rate": 0.000431642706086631,
      "loss": 5.0408,
      "mean_token_accuracy": 0.18316258639097213,
      "num_tokens": 57876732.0,
      "step": 26425
    },
    {
      "entropy": 5.641178226470947,
      "epoch": 2.6230647082175467,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043161726937031243,
      "loss": 4.9865,
      "mean_token_accuracy": 0.18667463213205338,
      "num_tokens": 57887178.0,
      "step": 26430
    },
    {
      "entropy": 5.671472358703613,
      "epoch": 2.623560936879714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000431591828770223,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18159946352243422,
      "num_tokens": 57898383.0,
      "step": 26435
    },
    {
      "entropy": 5.692240285873413,
      "epoch": 2.6240571655418816,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004315663842869938,
      "loss": 5.067,
      "mean_token_accuracy": 0.17654408812522887,
      "num_tokens": 57909009.0,
      "step": 26440
    },
    {
      "entropy": 5.706992864608765,
      "epoch": 2.624553394204049,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043154093592125554,
      "loss": 5.1449,
      "mean_token_accuracy": 0.18336669057607652,
      "num_tokens": 57920814.0,
      "step": 26445
    },
    {
      "entropy": 5.677439594268799,
      "epoch": 2.625049622866217,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043151548367363943,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18220461308956146,
      "num_tokens": 57932207.0,
      "step": 26450
    },
    {
      "entropy": 5.636863470077515,
      "epoch": 2.6255458515283845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043149002754477655,
      "loss": 5.0207,
      "mean_token_accuracy": 0.18190498352050782,
      "num_tokens": 57944663.0,
      "step": 26455
    },
    {
      "entropy": 5.693083763122559,
      "epoch": 2.626042080190552,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004314645675352982,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17716428339481355,
      "num_tokens": 57956213.0,
      "step": 26460
    },
    {
      "entropy": 5.746985292434692,
      "epoch": 2.6265383088527194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004314391036458357,
      "loss": 5.1548,
      "mean_token_accuracy": 0.1789749428629875,
      "num_tokens": 57967149.0,
      "step": 26465
    },
    {
      "entropy": 5.744828128814698,
      "epoch": 2.627034537514887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043141363587702054,
      "loss": 5.0342,
      "mean_token_accuracy": 0.18348321616649627,
      "num_tokens": 57978358.0,
      "step": 26470
    },
    {
      "entropy": 5.605380344390869,
      "epoch": 2.6275307661770544,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004313881642294843,
      "loss": 4.9704,
      "mean_token_accuracy": 0.18589708507061004,
      "num_tokens": 57990001.0,
      "step": 26475
    },
    {
      "entropy": 5.63989782333374,
      "epoch": 2.628026994839222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004313626887038585,
      "loss": 5.1007,
      "mean_token_accuracy": 0.18556373119354247,
      "num_tokens": 58003140.0,
      "step": 26480
    },
    {
      "entropy": 5.705256700515747,
      "epoch": 2.6285232235013893,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000431337209300775,
      "loss": 5.0629,
      "mean_token_accuracy": 0.1839625358581543,
      "num_tokens": 58014303.0,
      "step": 26485
    },
    {
      "entropy": 5.719945430755615,
      "epoch": 2.629019452163557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004313117260208656,
      "loss": 4.9915,
      "mean_token_accuracy": 0.1891400545835495,
      "num_tokens": 58024704.0,
      "step": 26490
    },
    {
      "entropy": 5.647173357009888,
      "epoch": 2.6295156808257243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004312862388647622,
      "loss": 4.9981,
      "mean_token_accuracy": 0.1851984366774559,
      "num_tokens": 58036587.0,
      "step": 26495
    },
    {
      "entropy": 5.566500997543335,
      "epoch": 2.6300119094878918,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043126074783309687,
      "loss": 4.9371,
      "mean_token_accuracy": 0.19166166931390763,
      "num_tokens": 58047106.0,
      "step": 26500
    },
    {
      "entropy": 5.603510332107544,
      "epoch": 2.6305081381500597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004312352529265016,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18058904111385346,
      "num_tokens": 58058227.0,
      "step": 26505
    },
    {
      "entropy": 5.722596120834351,
      "epoch": 2.631004366812227,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043120975414560886,
      "loss": 5.0716,
      "mean_token_accuracy": 0.18188426345586778,
      "num_tokens": 58069443.0,
      "step": 26510
    },
    {
      "entropy": 5.704393529891968,
      "epoch": 2.6315005954743946,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043118425149105073,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1872277870774269,
      "num_tokens": 58079925.0,
      "step": 26515
    },
    {
      "entropy": 5.652018117904663,
      "epoch": 2.631996824136562,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004311587449634597,
      "loss": 5.016,
      "mean_token_accuracy": 0.18333606719970702,
      "num_tokens": 58090974.0,
      "step": 26520
    },
    {
      "entropy": 5.6477500915527346,
      "epoch": 2.6324930527987296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004311332345634683,
      "loss": 5.0757,
      "mean_token_accuracy": 0.18030119240283965,
      "num_tokens": 58103548.0,
      "step": 26525
    },
    {
      "entropy": 5.671437835693359,
      "epoch": 2.632989281460897,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043110772029170916,
      "loss": 4.9906,
      "mean_token_accuracy": 0.1827929824590683,
      "num_tokens": 58113777.0,
      "step": 26530
    },
    {
      "entropy": 5.665403556823731,
      "epoch": 2.633485510123065,
      "grad_norm": 1.0,
      "learning_rate": 0.00043108220214881494,
      "loss": 4.9673,
      "mean_token_accuracy": 0.18347300738096237,
      "num_tokens": 58125230.0,
      "step": 26535
    },
    {
      "entropy": 5.662135219573974,
      "epoch": 2.6339817387852325,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004310566801354184,
      "loss": 5.0675,
      "mean_token_accuracy": 0.18513678163290023,
      "num_tokens": 58137572.0,
      "step": 26540
    },
    {
      "entropy": 5.702667999267578,
      "epoch": 2.6344779674474,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043103115425215246,
      "loss": 5.1,
      "mean_token_accuracy": 0.1773438885807991,
      "num_tokens": 58148207.0,
      "step": 26545
    },
    {
      "entropy": 5.722030830383301,
      "epoch": 2.6349741961095674,
      "grad_norm": 1.125,
      "learning_rate": 0.00043100562449965016,
      "loss": 5.0769,
      "mean_token_accuracy": 0.18110962361097335,
      "num_tokens": 58159390.0,
      "step": 26550
    },
    {
      "entropy": 5.6443647861480715,
      "epoch": 2.635470424771735,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043098009087854446,
      "loss": 5.0223,
      "mean_token_accuracy": 0.17924070209264756,
      "num_tokens": 58170842.0,
      "step": 26555
    },
    {
      "entropy": 5.665083312988282,
      "epoch": 2.6359666534339024,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043095455338946863,
      "loss": 5.0483,
      "mean_token_accuracy": 0.18204191625118255,
      "num_tokens": 58181507.0,
      "step": 26560
    },
    {
      "entropy": 5.655328321456909,
      "epoch": 2.63646288209607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000430929012033056,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18667052388191224,
      "num_tokens": 58192515.0,
      "step": 26565
    },
    {
      "entropy": 5.695898389816284,
      "epoch": 2.6369591107582373,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043090346680993983,
      "loss": 5.0397,
      "mean_token_accuracy": 0.18474819362163544,
      "num_tokens": 58202661.0,
      "step": 26570
    },
    {
      "entropy": 5.753752994537353,
      "epoch": 2.6374553394204048,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004308779177207536,
      "loss": 5.1549,
      "mean_token_accuracy": 0.17743263244628907,
      "num_tokens": 58213982.0,
      "step": 26575
    },
    {
      "entropy": 5.727571153640747,
      "epoch": 2.6379515680825723,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004308523647661309,
      "loss": 5.077,
      "mean_token_accuracy": 0.1821557402610779,
      "num_tokens": 58225386.0,
      "step": 26580
    },
    {
      "entropy": 5.676223659515381,
      "epoch": 2.6384477967447397,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043082680794670535,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1789563149213791,
      "num_tokens": 58236866.0,
      "step": 26585
    },
    {
      "entropy": 5.646325302124024,
      "epoch": 2.6389440254069076,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043080124726311084,
      "loss": 5.0089,
      "mean_token_accuracy": 0.18495680540800094,
      "num_tokens": 58248054.0,
      "step": 26590
    },
    {
      "entropy": 5.702210998535156,
      "epoch": 2.639440254069075,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004307756827159811,
      "loss": 5.077,
      "mean_token_accuracy": 0.18112428188323976,
      "num_tokens": 58259495.0,
      "step": 26595
    },
    {
      "entropy": 5.593976640701294,
      "epoch": 2.6399364827312426,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043075011430594994,
      "loss": 5.019,
      "mean_token_accuracy": 0.1857811540365219,
      "num_tokens": 58270766.0,
      "step": 26600
    },
    {
      "entropy": 5.73773455619812,
      "epoch": 2.64043271139341,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043072454203365175,
      "loss": 5.1356,
      "mean_token_accuracy": 0.18291642218828202,
      "num_tokens": 58281736.0,
      "step": 26605
    },
    {
      "entropy": 5.72137303352356,
      "epoch": 2.6409289400555775,
      "grad_norm": 1.25,
      "learning_rate": 0.0004306989658997204,
      "loss": 5.0406,
      "mean_token_accuracy": 0.19085373282432555,
      "num_tokens": 58292663.0,
      "step": 26610
    },
    {
      "entropy": 5.7284948348999025,
      "epoch": 2.641425168717745,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043067338590479023,
      "loss": 5.0385,
      "mean_token_accuracy": 0.19269680827856064,
      "num_tokens": 58303327.0,
      "step": 26615
    },
    {
      "entropy": 5.723639297485351,
      "epoch": 2.641921397379913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043064780204949543,
      "loss": 5.0702,
      "mean_token_accuracy": 0.18813256025314332,
      "num_tokens": 58313109.0,
      "step": 26620
    },
    {
      "entropy": 5.748323011398315,
      "epoch": 2.6424176260420804,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004306222143344706,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18415423184633256,
      "num_tokens": 58323587.0,
      "step": 26625
    },
    {
      "entropy": 5.657026195526123,
      "epoch": 2.642913854704248,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004305966227603503,
      "loss": 5.0046,
      "mean_token_accuracy": 0.18534397929906846,
      "num_tokens": 58333590.0,
      "step": 26630
    },
    {
      "entropy": 5.731450366973877,
      "epoch": 2.6434100833664154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004305710273277689,
      "loss": 5.0726,
      "mean_token_accuracy": 0.17821716517210007,
      "num_tokens": 58346127.0,
      "step": 26635
    },
    {
      "entropy": 5.712369918823242,
      "epoch": 2.643906312028583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043054542803736126,
      "loss": 5.0021,
      "mean_token_accuracy": 0.1834295094013214,
      "num_tokens": 58355769.0,
      "step": 26640
    },
    {
      "entropy": 5.773117208480835,
      "epoch": 2.6444025406907503,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043051982488976216,
      "loss": 5.1304,
      "mean_token_accuracy": 0.1820629745721817,
      "num_tokens": 58366932.0,
      "step": 26645
    },
    {
      "entropy": 5.650259065628052,
      "epoch": 2.644898769352918,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004304942178856066,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18249261379241943,
      "num_tokens": 58378198.0,
      "step": 26650
    },
    {
      "entropy": 5.667751169204712,
      "epoch": 2.6453949980150853,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043046860702552944,
      "loss": 4.9094,
      "mean_token_accuracy": 0.200072717666626,
      "num_tokens": 58388627.0,
      "step": 26655
    },
    {
      "entropy": 5.665114212036133,
      "epoch": 2.6458912266772527,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004304429923101658,
      "loss": 5.0186,
      "mean_token_accuracy": 0.18900009095668793,
      "num_tokens": 58400481.0,
      "step": 26660
    },
    {
      "entropy": 5.685538578033447,
      "epoch": 2.64638745533942,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000430417373740151,
      "loss": 5.0697,
      "mean_token_accuracy": 0.1810999721288681,
      "num_tokens": 58410827.0,
      "step": 26665
    },
    {
      "entropy": 5.709468984603882,
      "epoch": 2.6468836840015877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004303917513161201,
      "loss": 5.1148,
      "mean_token_accuracy": 0.17145849615335465,
      "num_tokens": 58422632.0,
      "step": 26670
    },
    {
      "entropy": 5.7094560146331785,
      "epoch": 2.6473799126637556,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004303661250387088,
      "loss": 5.1058,
      "mean_token_accuracy": 0.17371309101581572,
      "num_tokens": 58433770.0,
      "step": 26675
    },
    {
      "entropy": 5.6424637794494625,
      "epoch": 2.647876141325923,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004303404949085522,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18953278362751008,
      "num_tokens": 58442999.0,
      "step": 26680
    },
    {
      "entropy": 5.687827396392822,
      "epoch": 2.6483723699880906,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004303148609262861,
      "loss": 5.0914,
      "mean_token_accuracy": 0.1762094810605049,
      "num_tokens": 58454633.0,
      "step": 26685
    },
    {
      "entropy": 5.6774486064910885,
      "epoch": 2.648868598650258,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004302892230925462,
      "loss": 5.0352,
      "mean_token_accuracy": 0.1793937161564827,
      "num_tokens": 58464949.0,
      "step": 26690
    },
    {
      "entropy": 5.728937578201294,
      "epoch": 2.6493648273124255,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043026358140796806,
      "loss": 5.0509,
      "mean_token_accuracy": 0.1860477790236473,
      "num_tokens": 58475915.0,
      "step": 26695
    },
    {
      "entropy": 5.681257247924805,
      "epoch": 2.649861055974593,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004302379358731878,
      "loss": 5.0103,
      "mean_token_accuracy": 0.18885524272918702,
      "num_tokens": 58486582.0,
      "step": 26700
    },
    {
      "entropy": 5.658552503585815,
      "epoch": 2.6503572846367605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004302122864888411,
      "loss": 4.9983,
      "mean_token_accuracy": 0.18055080771446227,
      "num_tokens": 58498571.0,
      "step": 26705
    },
    {
      "entropy": 5.689505195617675,
      "epoch": 2.6508535132989284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004301866332555643,
      "loss": 5.0787,
      "mean_token_accuracy": 0.18417902886867524,
      "num_tokens": 58510445.0,
      "step": 26710
    },
    {
      "entropy": 5.661177682876587,
      "epoch": 2.651349741961096,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043016097617399327,
      "loss": 5.008,
      "mean_token_accuracy": 0.18398357778787613,
      "num_tokens": 58520248.0,
      "step": 26715
    },
    {
      "entropy": 5.728216934204101,
      "epoch": 2.6518459706232633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004301353152447644,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18304379880428315,
      "num_tokens": 58531993.0,
      "step": 26720
    },
    {
      "entropy": 5.643049907684326,
      "epoch": 2.652342199285431,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043010965046851404,
      "loss": 4.9994,
      "mean_token_accuracy": 0.19099672585725785,
      "num_tokens": 58543126.0,
      "step": 26725
    },
    {
      "entropy": 5.642090606689453,
      "epoch": 2.6528384279475983,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043008398184587847,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1761105552315712,
      "num_tokens": 58554627.0,
      "step": 26730
    },
    {
      "entropy": 5.7053216934204105,
      "epoch": 2.6533346566097658,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043005830937749447,
      "loss": 5.0049,
      "mean_token_accuracy": 0.1866767078638077,
      "num_tokens": 58565177.0,
      "step": 26735
    },
    {
      "entropy": 5.656838893890381,
      "epoch": 2.6538308852719332,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004300326330639984,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18028347045183182,
      "num_tokens": 58576596.0,
      "step": 26740
    },
    {
      "entropy": 5.691961765289307,
      "epoch": 2.6543271139341007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004300069529060272,
      "loss": 5.0645,
      "mean_token_accuracy": 0.175116066634655,
      "num_tokens": 58586250.0,
      "step": 26745
    },
    {
      "entropy": 5.679466295242309,
      "epoch": 2.654823342596268,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042998126890421754,
      "loss": 5.0178,
      "mean_token_accuracy": 0.1796593636274338,
      "num_tokens": 58597284.0,
      "step": 26750
    },
    {
      "entropy": 5.676307296752929,
      "epoch": 2.6553195712584357,
      "grad_norm": 1.125,
      "learning_rate": 0.00042995558105920627,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18005151599645614,
      "num_tokens": 58608888.0,
      "step": 26755
    },
    {
      "entropy": 5.62510986328125,
      "epoch": 2.6558157999206036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042992988937163066,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18594143092632293,
      "num_tokens": 58620809.0,
      "step": 26760
    },
    {
      "entropy": 5.632134437561035,
      "epoch": 2.656312028582771,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004299041938421275,
      "loss": 4.9625,
      "mean_token_accuracy": 0.1917755886912346,
      "num_tokens": 58630981.0,
      "step": 26765
    },
    {
      "entropy": 5.720618104934692,
      "epoch": 2.6568082572449385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042987849447133423,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17217201888561248,
      "num_tokens": 58642110.0,
      "step": 26770
    },
    {
      "entropy": 5.679708051681518,
      "epoch": 2.657304485907106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004298527912598879,
      "loss": 5.0357,
      "mean_token_accuracy": 0.18933371901512147,
      "num_tokens": 58654680.0,
      "step": 26775
    },
    {
      "entropy": 5.655454015731811,
      "epoch": 2.6578007145692735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004298270842084261,
      "loss": 5.0188,
      "mean_token_accuracy": 0.18402379453182222,
      "num_tokens": 58666005.0,
      "step": 26780
    },
    {
      "entropy": 5.659585237503052,
      "epoch": 2.658296943231441,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042980137331758626,
      "loss": 5.0785,
      "mean_token_accuracy": 0.17669565081596375,
      "num_tokens": 58677245.0,
      "step": 26785
    },
    {
      "entropy": 5.653941679000854,
      "epoch": 2.6587931718936084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042977565858800593,
      "loss": 4.9918,
      "mean_token_accuracy": 0.18521961569786072,
      "num_tokens": 58688470.0,
      "step": 26790
    },
    {
      "entropy": 5.7478760242462155,
      "epoch": 2.6592894005557763,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042974994002032275,
      "loss": 5.0976,
      "mean_token_accuracy": 0.18060144186019897,
      "num_tokens": 58698681.0,
      "step": 26795
    },
    {
      "entropy": 5.6696123600006105,
      "epoch": 2.659785629217944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042972421761517456,
      "loss": 5.0802,
      "mean_token_accuracy": 0.17850829511880875,
      "num_tokens": 58710113.0,
      "step": 26800
    },
    {
      "entropy": 5.716897487640381,
      "epoch": 2.6602818578801113,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004296984913731991,
      "loss": 5.062,
      "mean_token_accuracy": 0.18161360025405884,
      "num_tokens": 58720744.0,
      "step": 26805
    },
    {
      "entropy": 5.609233474731445,
      "epoch": 2.6607780865422788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004296727612950344,
      "loss": 4.9651,
      "mean_token_accuracy": 0.18630707263946533,
      "num_tokens": 58730955.0,
      "step": 26810
    },
    {
      "entropy": 5.754172134399414,
      "epoch": 2.6612743152044462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004296470273813186,
      "loss": 5.1086,
      "mean_token_accuracy": 0.1784886360168457,
      "num_tokens": 58743423.0,
      "step": 26815
    },
    {
      "entropy": 5.762178802490235,
      "epoch": 2.6617705438666137,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004296212896326896,
      "loss": 5.0127,
      "mean_token_accuracy": 0.1903802737593651,
      "num_tokens": 58754076.0,
      "step": 26820
    },
    {
      "entropy": 5.6792138576507565,
      "epoch": 2.662266772528781,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042959554804978596,
      "loss": 4.9614,
      "mean_token_accuracy": 0.18499543219804765,
      "num_tokens": 58765064.0,
      "step": 26825
    },
    {
      "entropy": 5.611749124526978,
      "epoch": 2.6627630011909487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042956980263324576,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19028008729219437,
      "num_tokens": 58776585.0,
      "step": 26830
    },
    {
      "entropy": 5.65381178855896,
      "epoch": 2.663259229853116,
      "grad_norm": 1.125,
      "learning_rate": 0.0004295440533837075,
      "loss": 5.0197,
      "mean_token_accuracy": 0.19029093980789186,
      "num_tokens": 58787533.0,
      "step": 26835
    },
    {
      "entropy": 5.667411470413208,
      "epoch": 2.6637554585152836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042951830030180983,
      "loss": 4.9412,
      "mean_token_accuracy": 0.18890873342752457,
      "num_tokens": 58797543.0,
      "step": 26840
    },
    {
      "entropy": 5.607243537902832,
      "epoch": 2.6642516871774515,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042949254338819123,
      "loss": 5.0073,
      "mean_token_accuracy": 0.1788231074810028,
      "num_tokens": 58808582.0,
      "step": 26845
    },
    {
      "entropy": 5.6267773628234865,
      "epoch": 2.664747915839619,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042946678264349036,
      "loss": 5.0533,
      "mean_token_accuracy": 0.18093544244766235,
      "num_tokens": 58819570.0,
      "step": 26850
    },
    {
      "entropy": 5.719548559188842,
      "epoch": 2.6652441445017865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042944101806834627,
      "loss": 5.1225,
      "mean_token_accuracy": 0.18031805753707886,
      "num_tokens": 58830063.0,
      "step": 26855
    },
    {
      "entropy": 5.653439235687256,
      "epoch": 2.665740373163954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004294152496633977,
      "loss": 4.9922,
      "mean_token_accuracy": 0.1936102405190468,
      "num_tokens": 58842239.0,
      "step": 26860
    },
    {
      "entropy": 5.701949167251587,
      "epoch": 2.6662366018261214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042938947742928357,
      "loss": 5.0736,
      "mean_token_accuracy": 0.17969090789556502,
      "num_tokens": 58852570.0,
      "step": 26865
    },
    {
      "entropy": 5.6900848865509035,
      "epoch": 2.666732830488289,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042936370136664314,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1836899548768997,
      "num_tokens": 58862759.0,
      "step": 26870
    },
    {
      "entropy": 5.619800424575805,
      "epoch": 2.6672290591504564,
      "grad_norm": 1.125,
      "learning_rate": 0.0004293379214761156,
      "loss": 4.9825,
      "mean_token_accuracy": 0.18612193018198014,
      "num_tokens": 58873976.0,
      "step": 26875
    },
    {
      "entropy": 5.609964370727539,
      "epoch": 2.6677252878126243,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004293121377583401,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18497966825962067,
      "num_tokens": 58885440.0,
      "step": 26880
    },
    {
      "entropy": 5.609320735931396,
      "epoch": 2.6682215164747918,
      "grad_norm": 1.0,
      "learning_rate": 0.0004292863502139561,
      "loss": 5.0459,
      "mean_token_accuracy": 0.18399827778339387,
      "num_tokens": 58896301.0,
      "step": 26885
    },
    {
      "entropy": 5.695574235916138,
      "epoch": 2.6687177451369593,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004292605588436031,
      "loss": 4.9883,
      "mean_token_accuracy": 0.18839032649993898,
      "num_tokens": 58906888.0,
      "step": 26890
    },
    {
      "entropy": 5.717593240737915,
      "epoch": 2.6692139737991267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042923476364792063,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17561130672693254,
      "num_tokens": 58918025.0,
      "step": 26895
    },
    {
      "entropy": 5.721667051315308,
      "epoch": 2.669710202461294,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004292089646275484,
      "loss": 5.1189,
      "mean_token_accuracy": 0.17596112340688705,
      "num_tokens": 58929012.0,
      "step": 26900
    },
    {
      "entropy": 5.733981513977051,
      "epoch": 2.6702064311234617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004291831617831261,
      "loss": 5.0636,
      "mean_token_accuracy": 0.18130408525466918,
      "num_tokens": 58939133.0,
      "step": 26905
    },
    {
      "entropy": 5.723064231872558,
      "epoch": 2.670702659785629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042915735511529356,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1719641849398613,
      "num_tokens": 58950227.0,
      "step": 26910
    },
    {
      "entropy": 5.621247148513794,
      "epoch": 2.6711988884477966,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004291315446246909,
      "loss": 4.9662,
      "mean_token_accuracy": 0.1845977172255516,
      "num_tokens": 58961614.0,
      "step": 26915
    },
    {
      "entropy": 5.662209987640381,
      "epoch": 2.671695117109964,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042910573031195804,
      "loss": 5.0752,
      "mean_token_accuracy": 0.18388722836971283,
      "num_tokens": 58973252.0,
      "step": 26920
    },
    {
      "entropy": 5.685889148712159,
      "epoch": 2.6721913457721316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004290799121777351,
      "loss": 5.032,
      "mean_token_accuracy": 0.1797099620103836,
      "num_tokens": 58985223.0,
      "step": 26925
    },
    {
      "entropy": 5.606911897659302,
      "epoch": 2.672687574434299,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004290540902226624,
      "loss": 4.9573,
      "mean_token_accuracy": 0.1877106472849846,
      "num_tokens": 58995651.0,
      "step": 26930
    },
    {
      "entropy": 5.652631521224976,
      "epoch": 2.673183803096467,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004290282644473802,
      "loss": 5.0383,
      "mean_token_accuracy": 0.18521432727575302,
      "num_tokens": 59006171.0,
      "step": 26935
    },
    {
      "entropy": 5.675985288619995,
      "epoch": 2.6736800317586344,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004290024348525289,
      "loss": 5.0256,
      "mean_token_accuracy": 0.18650251030921935,
      "num_tokens": 59017595.0,
      "step": 26940
    },
    {
      "entropy": 5.7189483642578125,
      "epoch": 2.674176260420802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004289766014387492,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17800354957580566,
      "num_tokens": 59028896.0,
      "step": 26945
    },
    {
      "entropy": 5.651254320144654,
      "epoch": 2.6746724890829694,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004289507642066813,
      "loss": 4.9454,
      "mean_token_accuracy": 0.18322572112083435,
      "num_tokens": 59040239.0,
      "step": 26950
    },
    {
      "entropy": 5.680315113067627,
      "epoch": 2.675168717745137,
      "grad_norm": 1.125,
      "learning_rate": 0.0004289249231569664,
      "loss": 5.1118,
      "mean_token_accuracy": 0.1795588955283165,
      "num_tokens": 59050742.0,
      "step": 26955
    },
    {
      "entropy": 5.807846593856811,
      "epoch": 2.6756649464073043,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042889907829024506,
      "loss": 5.158,
      "mean_token_accuracy": 0.17825432419776915,
      "num_tokens": 59062985.0,
      "step": 26960
    },
    {
      "entropy": 5.773179769515991,
      "epoch": 2.6761611750694723,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042887322960715813,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18636412024497986,
      "num_tokens": 59074384.0,
      "step": 26965
    },
    {
      "entropy": 5.721178293228149,
      "epoch": 2.6766574037316397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004288473771083468,
      "loss": 5.1686,
      "mean_token_accuracy": 0.17661335170269013,
      "num_tokens": 59086332.0,
      "step": 26970
    },
    {
      "entropy": 5.644415473937988,
      "epoch": 2.677153632393807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004288215207944519,
      "loss": 5.0022,
      "mean_token_accuracy": 0.1856015831232071,
      "num_tokens": 59098300.0,
      "step": 26975
    },
    {
      "entropy": 5.746345329284668,
      "epoch": 2.6776498610559747,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042879566066611474,
      "loss": 5.1275,
      "mean_token_accuracy": 0.18180275708436966,
      "num_tokens": 59109596.0,
      "step": 26980
    },
    {
      "entropy": 5.740393257141113,
      "epoch": 2.678146089718142,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004287697967239767,
      "loss": 5.0479,
      "mean_token_accuracy": 0.18664272576570512,
      "num_tokens": 59119590.0,
      "step": 26985
    },
    {
      "entropy": 5.656682968139648,
      "epoch": 2.6786423183803096,
      "grad_norm": 0.953125,
      "learning_rate": 0.000428743928968679,
      "loss": 5.0699,
      "mean_token_accuracy": 0.1831554353237152,
      "num_tokens": 59132219.0,
      "step": 26990
    },
    {
      "entropy": 5.65801591873169,
      "epoch": 2.679138547042477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042871805740086317,
      "loss": 4.9882,
      "mean_token_accuracy": 0.18883044719696046,
      "num_tokens": 59142752.0,
      "step": 26995
    },
    {
      "entropy": 5.643645095825195,
      "epoch": 2.6796347757046446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004286921820211707,
      "loss": 4.9876,
      "mean_token_accuracy": 0.17763181328773497,
      "num_tokens": 59153458.0,
      "step": 27000
    },
    {
      "epoch": 2.6796347757046446,
      "eval_entropy": 5.409022269381512,
      "eval_loss": 5.146875858306885,
      "eval_mean_token_accuracy": 0.18302252645637282,
      "eval_num_tokens": 59153458.0,
      "eval_runtime": 20.4778,
      "eval_samples_per_second": 1644.17,
      "eval_steps_per_second": 205.54,
      "step": 27000
    },
    {
      "entropy": 5.660647773742676,
      "epoch": 2.680131004366812,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042866630283024327,
      "loss": 5.035,
      "mean_token_accuracy": 0.18509134352207185,
      "num_tokens": 59164755.0,
      "step": 27005
    },
    {
      "entropy": 5.571868515014648,
      "epoch": 2.6806272330289795,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004286404198287227,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18236615955829621,
      "num_tokens": 59174470.0,
      "step": 27010
    },
    {
      "entropy": 5.64816517829895,
      "epoch": 2.681123461691147,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004286145330172508,
      "loss": 4.9827,
      "mean_token_accuracy": 0.1834064841270447,
      "num_tokens": 59184839.0,
      "step": 27015
    },
    {
      "entropy": 5.614450454711914,
      "epoch": 2.681619690353315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004285886423964695,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18907615691423416,
      "num_tokens": 59194380.0,
      "step": 27020
    },
    {
      "entropy": 5.672947740554809,
      "epoch": 2.6821159190154824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004285627479670208,
      "loss": 5.068,
      "mean_token_accuracy": 0.1852304682135582,
      "num_tokens": 59205509.0,
      "step": 27025
    },
    {
      "entropy": 5.6614306449890135,
      "epoch": 2.68261214767765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004285368497295468,
      "loss": 4.9486,
      "mean_token_accuracy": 0.18812448531389236,
      "num_tokens": 59216342.0,
      "step": 27030
    },
    {
      "entropy": 5.683060646057129,
      "epoch": 2.6831083763398174,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004285109476846898,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1862706646323204,
      "num_tokens": 59227217.0,
      "step": 27035
    },
    {
      "entropy": 5.731919813156128,
      "epoch": 2.683604605001985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004284850418330921,
      "loss": 5.1549,
      "mean_token_accuracy": 0.17365396171808242,
      "num_tokens": 59239441.0,
      "step": 27040
    },
    {
      "entropy": 5.714769554138184,
      "epoch": 2.6841008336641523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042845913217539613,
      "loss": 5.076,
      "mean_token_accuracy": 0.17650984823703766,
      "num_tokens": 59250015.0,
      "step": 27045
    },
    {
      "entropy": 5.628409671783447,
      "epoch": 2.6845970623263202,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042843321871224426,
      "loss": 4.9231,
      "mean_token_accuracy": 0.19207071959972383,
      "num_tokens": 59259673.0,
      "step": 27050
    },
    {
      "entropy": 5.672825813293457,
      "epoch": 2.6850932909884877,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042840730144427927,
      "loss": 5.098,
      "mean_token_accuracy": 0.18372515588998795,
      "num_tokens": 59270927.0,
      "step": 27055
    },
    {
      "entropy": 5.602501392364502,
      "epoch": 2.685589519650655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042838138037214375,
      "loss": 4.9964,
      "mean_token_accuracy": 0.18681490272283555,
      "num_tokens": 59280831.0,
      "step": 27060
    },
    {
      "entropy": 5.768775415420532,
      "epoch": 2.6860857483128227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042835545549648054,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17606378644704818,
      "num_tokens": 59291728.0,
      "step": 27065
    },
    {
      "entropy": 5.642182350158691,
      "epoch": 2.68658197697499,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042832952681793235,
      "loss": 4.9878,
      "mean_token_accuracy": 0.1856672778725624,
      "num_tokens": 59302094.0,
      "step": 27070
    },
    {
      "entropy": 5.632256412506104,
      "epoch": 2.6870782056371576,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042830359433714243,
      "loss": 5.0352,
      "mean_token_accuracy": 0.18313378542661668,
      "num_tokens": 59312951.0,
      "step": 27075
    },
    {
      "entropy": 5.644662523269654,
      "epoch": 2.687574434299325,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042827765805475367,
      "loss": 5.0461,
      "mean_token_accuracy": 0.1779997631907463,
      "num_tokens": 59322858.0,
      "step": 27080
    },
    {
      "entropy": 5.6687417984008786,
      "epoch": 2.6880706629614926,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042825171797140926,
      "loss": 5.0129,
      "mean_token_accuracy": 0.1820945769548416,
      "num_tokens": 59332440.0,
      "step": 27085
    },
    {
      "entropy": 5.625169038772583,
      "epoch": 2.68856689162366,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004282257740877525,
      "loss": 4.96,
      "mean_token_accuracy": 0.18692585974931716,
      "num_tokens": 59343265.0,
      "step": 27090
    },
    {
      "entropy": 5.578926086425781,
      "epoch": 2.6890631202858275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004281998264044267,
      "loss": 4.9686,
      "mean_token_accuracy": 0.18320464193820954,
      "num_tokens": 59354212.0,
      "step": 27095
    },
    {
      "entropy": 5.63108868598938,
      "epoch": 2.689559348947995,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004281738749220753,
      "loss": 4.9869,
      "mean_token_accuracy": 0.18863331973552705,
      "num_tokens": 59365674.0,
      "step": 27100
    },
    {
      "entropy": 5.673368978500366,
      "epoch": 2.690055577610163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042814791964134184,
      "loss": 4.97,
      "mean_token_accuracy": 0.1889665648341179,
      "num_tokens": 59376329.0,
      "step": 27105
    },
    {
      "entropy": 5.713589286804199,
      "epoch": 2.6905518062723304,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042812196056287,
      "loss": 5.0494,
      "mean_token_accuracy": 0.18530997037887573,
      "num_tokens": 59387689.0,
      "step": 27110
    },
    {
      "entropy": 5.700909280776978,
      "epoch": 2.691048034934498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042809599768730357,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1796042114496231,
      "num_tokens": 59397285.0,
      "step": 27115
    },
    {
      "entropy": 5.620287704467773,
      "epoch": 2.6915442635966653,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004280700310152861,
      "loss": 5.006,
      "mean_token_accuracy": 0.18353141695261002,
      "num_tokens": 59409284.0,
      "step": 27120
    },
    {
      "entropy": 5.639403057098389,
      "epoch": 2.692040492258833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004280440605474619,
      "loss": 4.9988,
      "mean_token_accuracy": 0.18692222386598586,
      "num_tokens": 59421086.0,
      "step": 27125
    },
    {
      "entropy": 5.729607009887696,
      "epoch": 2.6925367209210003,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004280180862844747,
      "loss": 5.0821,
      "mean_token_accuracy": 0.17680520564317703,
      "num_tokens": 59433226.0,
      "step": 27130
    },
    {
      "entropy": 5.641674995422363,
      "epoch": 2.693032949583168,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042799210822696865,
      "loss": 4.997,
      "mean_token_accuracy": 0.1876170441508293,
      "num_tokens": 59442613.0,
      "step": 27135
    },
    {
      "entropy": 5.631798315048218,
      "epoch": 2.6935291782453357,
      "grad_norm": 1.15625,
      "learning_rate": 0.000427966126375588,
      "loss": 5.0328,
      "mean_token_accuracy": 0.1782523959875107,
      "num_tokens": 59453473.0,
      "step": 27140
    },
    {
      "entropy": 5.6162491798400875,
      "epoch": 2.694025406907503,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004279401407309771,
      "loss": 4.9843,
      "mean_token_accuracy": 0.1897462159395218,
      "num_tokens": 59464495.0,
      "step": 27145
    },
    {
      "entropy": 5.661838388442993,
      "epoch": 2.6945216355696706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004279141512937802,
      "loss": 5.0522,
      "mean_token_accuracy": 0.17944473922252654,
      "num_tokens": 59475575.0,
      "step": 27150
    },
    {
      "entropy": 5.608452796936035,
      "epoch": 2.695017864231838,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004278881580646418,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18489157557487487,
      "num_tokens": 59487432.0,
      "step": 27155
    },
    {
      "entropy": 5.660338306427002,
      "epoch": 2.6955140928940056,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004278621610442066,
      "loss": 4.9881,
      "mean_token_accuracy": 0.19015263617038727,
      "num_tokens": 59498694.0,
      "step": 27160
    },
    {
      "entropy": 5.713187503814697,
      "epoch": 2.696010321556173,
      "grad_norm": 1.125,
      "learning_rate": 0.0004278361602331192,
      "loss": 5.1162,
      "mean_token_accuracy": 0.18084917813539506,
      "num_tokens": 59509303.0,
      "step": 27165
    },
    {
      "entropy": 5.61648120880127,
      "epoch": 2.6965065502183405,
      "grad_norm": 1.234375,
      "learning_rate": 0.00042781015563202434,
      "loss": 4.9976,
      "mean_token_accuracy": 0.18594128638505936,
      "num_tokens": 59519924.0,
      "step": 27170
    },
    {
      "entropy": 5.676542520523071,
      "epoch": 2.697002778880508,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004277841472415669,
      "loss": 5.0443,
      "mean_token_accuracy": 0.18603798002004623,
      "num_tokens": 59530880.0,
      "step": 27175
    },
    {
      "entropy": 5.703761100769043,
      "epoch": 2.6974990075426755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004277581350623918,
      "loss": 5.0652,
      "mean_token_accuracy": 0.17881901264190675,
      "num_tokens": 59542314.0,
      "step": 27180
    },
    {
      "entropy": 5.680798959732056,
      "epoch": 2.697995236204843,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004277321190951441,
      "loss": 5.029,
      "mean_token_accuracy": 0.1777633711695671,
      "num_tokens": 59552776.0,
      "step": 27185
    },
    {
      "entropy": 5.705180835723877,
      "epoch": 2.698491464867011,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042770609934046904,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18719606399536132,
      "num_tokens": 59563302.0,
      "step": 27190
    },
    {
      "entropy": 5.62778959274292,
      "epoch": 2.6989876935291783,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042768007579901176,
      "loss": 4.9978,
      "mean_token_accuracy": 0.17915077060461043,
      "num_tokens": 59574151.0,
      "step": 27195
    },
    {
      "entropy": 5.684006929397583,
      "epoch": 2.699483922191346,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042765404847141755,
      "loss": 5.0503,
      "mean_token_accuracy": 0.17885158807039261,
      "num_tokens": 59585093.0,
      "step": 27200
    },
    {
      "entropy": 5.638724613189697,
      "epoch": 2.6999801508535133,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004276280173583319,
      "loss": 4.9571,
      "mean_token_accuracy": 0.19368771761655806,
      "num_tokens": 59594908.0,
      "step": 27205
    },
    {
      "entropy": 5.6071265697479244,
      "epoch": 2.7004763795156808,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042760198246040025,
      "loss": 4.9723,
      "mean_token_accuracy": 0.19135363548994064,
      "num_tokens": 59607015.0,
      "step": 27210
    },
    {
      "entropy": 5.720832633972168,
      "epoch": 2.7009726081778482,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004275759437782683,
      "loss": 5.1152,
      "mean_token_accuracy": 0.17637474089860916,
      "num_tokens": 59617597.0,
      "step": 27215
    },
    {
      "entropy": 5.6585280418396,
      "epoch": 2.701468836840016,
      "grad_norm": 1.125,
      "learning_rate": 0.00042754990131258175,
      "loss": 4.97,
      "mean_token_accuracy": 0.18458838313817977,
      "num_tokens": 59627934.0,
      "step": 27220
    },
    {
      "entropy": 5.6627147674560545,
      "epoch": 2.7019650655021836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004275238550639863,
      "loss": 5.066,
      "mean_token_accuracy": 0.18062058389186858,
      "num_tokens": 59638759.0,
      "step": 27225
    },
    {
      "entropy": 5.694469165802002,
      "epoch": 2.702461294164351,
      "grad_norm": 1.109375,
      "learning_rate": 0.000427497805033128,
      "loss": 5.075,
      "mean_token_accuracy": 0.17589386701583862,
      "num_tokens": 59650199.0,
      "step": 27230
    },
    {
      "entropy": 5.680690002441406,
      "epoch": 2.7029575228265186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004274717512206527,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17342855930328369,
      "num_tokens": 59661480.0,
      "step": 27235
    },
    {
      "entropy": 5.714049053192139,
      "epoch": 2.703453751488686,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004274456936272065,
      "loss": 5.1037,
      "mean_token_accuracy": 0.18322691470384597,
      "num_tokens": 59673048.0,
      "step": 27240
    },
    {
      "entropy": 5.734686231613159,
      "epoch": 2.7039499801508535,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042741963225343554,
      "loss": 5.1424,
      "mean_token_accuracy": 0.1750873163342476,
      "num_tokens": 59686218.0,
      "step": 27245
    },
    {
      "entropy": 5.666069459915161,
      "epoch": 2.704446208813021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004273935670999861,
      "loss": 5.0137,
      "mean_token_accuracy": 0.17784301191568375,
      "num_tokens": 59695870.0,
      "step": 27250
    },
    {
      "entropy": 5.622952127456665,
      "epoch": 2.7049424374751885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042736749816750476,
      "loss": 4.9382,
      "mean_token_accuracy": 0.18637916892766954,
      "num_tokens": 59706479.0,
      "step": 27255
    },
    {
      "entropy": 5.674301242828369,
      "epoch": 2.705438666137356,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004273414254566376,
      "loss": 5.1055,
      "mean_token_accuracy": 0.1822776049375534,
      "num_tokens": 59717411.0,
      "step": 27260
    },
    {
      "entropy": 5.716517114639283,
      "epoch": 2.7059348947995234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004273153489680315,
      "loss": 5.0929,
      "mean_token_accuracy": 0.1762610614299774,
      "num_tokens": 59728363.0,
      "step": 27265
    },
    {
      "entropy": 5.653161907196045,
      "epoch": 2.706431123461691,
      "grad_norm": 1.0,
      "learning_rate": 0.00042728926870233287,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18481247276067733,
      "num_tokens": 59738714.0,
      "step": 27270
    },
    {
      "entropy": 5.66207480430603,
      "epoch": 2.706927352123859,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042726318466018855,
      "loss": 5.0393,
      "mean_token_accuracy": 0.1895486071705818,
      "num_tokens": 59749292.0,
      "step": 27275
    },
    {
      "entropy": 5.687474250793457,
      "epoch": 2.7074235807860263,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004272370968422454,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18072600960731505,
      "num_tokens": 59760207.0,
      "step": 27280
    },
    {
      "entropy": 5.632325077056885,
      "epoch": 2.7079198094481938,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042721100524915017,
      "loss": 4.951,
      "mean_token_accuracy": 0.1917423278093338,
      "num_tokens": 59771780.0,
      "step": 27285
    },
    {
      "entropy": 5.711581563949585,
      "epoch": 2.7084160381103612,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042718490988155014,
      "loss": 5.0482,
      "mean_token_accuracy": 0.18491272777318954,
      "num_tokens": 59782298.0,
      "step": 27290
    },
    {
      "entropy": 5.678626346588135,
      "epoch": 2.7089122667725287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042715881074009216,
      "loss": 5.1107,
      "mean_token_accuracy": 0.1813090354204178,
      "num_tokens": 59794097.0,
      "step": 27295
    },
    {
      "entropy": 5.682178068161011,
      "epoch": 2.709408495434696,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042713270782542364,
      "loss": 4.988,
      "mean_token_accuracy": 0.19077412486076356,
      "num_tokens": 59804373.0,
      "step": 27300
    },
    {
      "entropy": 5.659429311752319,
      "epoch": 2.7099047240968637,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004271066011381917,
      "loss": 5.0464,
      "mean_token_accuracy": 0.18102400600910187,
      "num_tokens": 59817422.0,
      "step": 27305
    },
    {
      "entropy": 5.6492053985595705,
      "epoch": 2.7104009527590316,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004270804906790438,
      "loss": 5.0598,
      "mean_token_accuracy": 0.18094191402196885,
      "num_tokens": 59828531.0,
      "step": 27310
    },
    {
      "entropy": 5.712206268310547,
      "epoch": 2.710897181421199,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004270543764486275,
      "loss": 5.0734,
      "mean_token_accuracy": 0.18116782754659652,
      "num_tokens": 59839817.0,
      "step": 27315
    },
    {
      "entropy": 5.67339129447937,
      "epoch": 2.7113934100833665,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004270282584475903,
      "loss": 5.0448,
      "mean_token_accuracy": 0.1874105900526047,
      "num_tokens": 59850697.0,
      "step": 27320
    },
    {
      "entropy": 5.604316568374633,
      "epoch": 2.711889638745534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042700213667657983,
      "loss": 5.0139,
      "mean_token_accuracy": 0.18431638777256013,
      "num_tokens": 59861255.0,
      "step": 27325
    },
    {
      "entropy": 5.648484468460083,
      "epoch": 2.7123858674077015,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000426976011136244,
      "loss": 5.0075,
      "mean_token_accuracy": 0.18404365926980973,
      "num_tokens": 59872225.0,
      "step": 27330
    },
    {
      "entropy": 5.736332273483276,
      "epoch": 2.712882096069869,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042694988182723047,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18523466289043428,
      "num_tokens": 59882588.0,
      "step": 27335
    },
    {
      "entropy": 5.687605190277099,
      "epoch": 2.7133783247320364,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042692374875018735,
      "loss": 5.1606,
      "mean_token_accuracy": 0.17945859134197234,
      "num_tokens": 59895023.0,
      "step": 27340
    },
    {
      "entropy": 5.689088821411133,
      "epoch": 2.713874553394204,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042689761190576263,
      "loss": 4.9999,
      "mean_token_accuracy": 0.18411976099014282,
      "num_tokens": 59906497.0,
      "step": 27345
    },
    {
      "entropy": 5.683567953109741,
      "epoch": 2.7143707820563714,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042687147129460447,
      "loss": 4.9945,
      "mean_token_accuracy": 0.18264287561178208,
      "num_tokens": 59917261.0,
      "step": 27350
    },
    {
      "entropy": 5.624971199035644,
      "epoch": 2.714867010718539,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042684532691736104,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18388729691505432,
      "num_tokens": 59928398.0,
      "step": 27355
    },
    {
      "entropy": 5.622186851501465,
      "epoch": 2.715363239380707,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042681917877468076,
      "loss": 4.9528,
      "mean_token_accuracy": 0.18744801580905915,
      "num_tokens": 59939259.0,
      "step": 27360
    },
    {
      "entropy": 5.662555265426636,
      "epoch": 2.7158594680428743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042679302686721187,
      "loss": 4.9799,
      "mean_token_accuracy": 0.1831154927611351,
      "num_tokens": 59950495.0,
      "step": 27365
    },
    {
      "entropy": 5.671580791473389,
      "epoch": 2.7163556967050417,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004267668711956031,
      "loss": 5.0465,
      "mean_token_accuracy": 0.18440003097057342,
      "num_tokens": 59963038.0,
      "step": 27370
    },
    {
      "entropy": 5.660066604614258,
      "epoch": 2.716851925367209,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042674071176050296,
      "loss": 5.0686,
      "mean_token_accuracy": 0.18157610446214675,
      "num_tokens": 59973698.0,
      "step": 27375
    },
    {
      "entropy": 5.691748380661011,
      "epoch": 2.7173481540293767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042671454856256,
      "loss": 4.9966,
      "mean_token_accuracy": 0.18496868908405303,
      "num_tokens": 59984098.0,
      "step": 27380
    },
    {
      "entropy": 5.688077402114868,
      "epoch": 2.717844382691544,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004266883816024233,
      "loss": 4.949,
      "mean_token_accuracy": 0.18911459892988206,
      "num_tokens": 59995222.0,
      "step": 27385
    },
    {
      "entropy": 5.658065509796143,
      "epoch": 2.7183406113537116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004266622108807415,
      "loss": 5.0978,
      "mean_token_accuracy": 0.189106023311615,
      "num_tokens": 60006886.0,
      "step": 27390
    },
    {
      "entropy": 5.658921480178833,
      "epoch": 2.7188368400158796,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004266360363981637,
      "loss": 5.067,
      "mean_token_accuracy": 0.18053707033395766,
      "num_tokens": 60018408.0,
      "step": 27395
    },
    {
      "entropy": 5.714153623580932,
      "epoch": 2.719333068678047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042660985815533905,
      "loss": 5.0736,
      "mean_token_accuracy": 0.17644005119800568,
      "num_tokens": 60029927.0,
      "step": 27400
    },
    {
      "entropy": 5.7126476764678955,
      "epoch": 2.7198292973402145,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042658367615291644,
      "loss": 5.0216,
      "mean_token_accuracy": 0.17899354845285415,
      "num_tokens": 60040521.0,
      "step": 27405
    },
    {
      "entropy": 5.603881454467773,
      "epoch": 2.720325526002382,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004265574903915454,
      "loss": 4.9903,
      "mean_token_accuracy": 0.184011310338974,
      "num_tokens": 60051709.0,
      "step": 27410
    },
    {
      "entropy": 5.590267610549927,
      "epoch": 2.7208217546645495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004265313008718751,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18139002472162247,
      "num_tokens": 60062478.0,
      "step": 27415
    },
    {
      "entropy": 5.715270137786865,
      "epoch": 2.721317983326717,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042650510759455516,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1820229023694992,
      "num_tokens": 60072957.0,
      "step": 27420
    },
    {
      "entropy": 5.72929015159607,
      "epoch": 2.7218142119888844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042647891056023495,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18364216089248658,
      "num_tokens": 60084174.0,
      "step": 27425
    },
    {
      "entropy": 5.6623087406158445,
      "epoch": 2.722310440651052,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042645270976956426,
      "loss": 4.9702,
      "mean_token_accuracy": 0.19051365107297896,
      "num_tokens": 60095647.0,
      "step": 27430
    },
    {
      "entropy": 5.651272821426391,
      "epoch": 2.7228066693132194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042642650522319253,
      "loss": 5.0432,
      "mean_token_accuracy": 0.17688941061496735,
      "num_tokens": 60106355.0,
      "step": 27435
    },
    {
      "entropy": 5.599193906784057,
      "epoch": 2.723302897975387,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004264002969217699,
      "loss": 4.93,
      "mean_token_accuracy": 0.18901843279600145,
      "num_tokens": 60116616.0,
      "step": 27440
    },
    {
      "entropy": 5.710363245010376,
      "epoch": 2.7237991266375547,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004263740848659461,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1864851623773575,
      "num_tokens": 60127168.0,
      "step": 27445
    },
    {
      "entropy": 5.677166748046875,
      "epoch": 2.724295355299722,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042634786905637123,
      "loss": 5.0215,
      "mean_token_accuracy": 0.1879727065563202,
      "num_tokens": 60137953.0,
      "step": 27450
    },
    {
      "entropy": 5.688990354537964,
      "epoch": 2.7247915839618897,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004263216494936953,
      "loss": 5.0561,
      "mean_token_accuracy": 0.177470064163208,
      "num_tokens": 60149513.0,
      "step": 27455
    },
    {
      "entropy": 5.66733832359314,
      "epoch": 2.725287812624057,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042629542617856855,
      "loss": 4.9781,
      "mean_token_accuracy": 0.1858611971139908,
      "num_tokens": 60160977.0,
      "step": 27460
    },
    {
      "entropy": 5.653369474411011,
      "epoch": 2.7257840412862246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004262691991116412,
      "loss": 5.0078,
      "mean_token_accuracy": 0.1851443886756897,
      "num_tokens": 60171607.0,
      "step": 27465
    },
    {
      "entropy": 5.670451402664185,
      "epoch": 2.726280269948392,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004262429682935637,
      "loss": 5.0874,
      "mean_token_accuracy": 0.17951204627752304,
      "num_tokens": 60183386.0,
      "step": 27470
    },
    {
      "entropy": 5.70088267326355,
      "epoch": 2.7267764986105596,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004262167337249865,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18056659102439881,
      "num_tokens": 60194494.0,
      "step": 27475
    },
    {
      "entropy": 5.60939474105835,
      "epoch": 2.7272727272727275,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004261904954065601,
      "loss": 4.9391,
      "mean_token_accuracy": 0.19175520539283752,
      "num_tokens": 60206053.0,
      "step": 27480
    },
    {
      "entropy": 5.640282821655274,
      "epoch": 2.727768955934895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004261642533389352,
      "loss": 5.0462,
      "mean_token_accuracy": 0.18100040704011916,
      "num_tokens": 60218564.0,
      "step": 27485
    },
    {
      "entropy": 5.675661993026734,
      "epoch": 2.7282651845970625,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042613800752276264,
      "loss": 4.9882,
      "mean_token_accuracy": 0.1965646430850029,
      "num_tokens": 60228979.0,
      "step": 27490
    },
    {
      "entropy": 5.656196403503418,
      "epoch": 2.72876141325923,
      "grad_norm": 1.125,
      "learning_rate": 0.00042611175795869306,
      "loss": 5.0713,
      "mean_token_accuracy": 0.17374542653560637,
      "num_tokens": 60239143.0,
      "step": 27495
    },
    {
      "entropy": 5.7064450740814205,
      "epoch": 2.7292576419213974,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004260855046473775,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17528121173381805,
      "num_tokens": 60249815.0,
      "step": 27500
    },
    {
      "entropy": 5.661878395080566,
      "epoch": 2.729753870583565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042605924758946706,
      "loss": 4.9802,
      "mean_token_accuracy": 0.1891026273369789,
      "num_tokens": 60259909.0,
      "step": 27505
    },
    {
      "entropy": 5.745583724975586,
      "epoch": 2.7302500992457324,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042603298678561277,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1843745678663254,
      "num_tokens": 60271020.0,
      "step": 27510
    },
    {
      "entropy": 5.688380718231201,
      "epoch": 2.7307463279079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004260067222364659,
      "loss": 4.9675,
      "mean_token_accuracy": 0.1862027868628502,
      "num_tokens": 60281986.0,
      "step": 27515
    },
    {
      "entropy": 5.639588117599487,
      "epoch": 2.7312425565700673,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042598045394267757,
      "loss": 5.054,
      "mean_token_accuracy": 0.18055299073457717,
      "num_tokens": 60292435.0,
      "step": 27520
    },
    {
      "entropy": 5.692773246765137,
      "epoch": 2.731738785232235,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042595418190489944,
      "loss": 5.032,
      "mean_token_accuracy": 0.1839613512158394,
      "num_tokens": 60303783.0,
      "step": 27525
    },
    {
      "entropy": 5.659827375411988,
      "epoch": 2.7322350138944023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042592790612378277,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18445808440446854,
      "num_tokens": 60313525.0,
      "step": 27530
    },
    {
      "entropy": 5.6098198890686035,
      "epoch": 2.73273124255657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042590162659997933,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18745106756687163,
      "num_tokens": 60323788.0,
      "step": 27535
    },
    {
      "entropy": 5.6830543041229244,
      "epoch": 2.7332274712187377,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004258753433341407,
      "loss": 5.1642,
      "mean_token_accuracy": 0.18380267769098282,
      "num_tokens": 60334006.0,
      "step": 27540
    },
    {
      "entropy": 5.723185443878174,
      "epoch": 2.733723699880905,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042584905632691867,
      "loss": 5.0497,
      "mean_token_accuracy": 0.18398260921239853,
      "num_tokens": 60345339.0,
      "step": 27545
    },
    {
      "entropy": 5.691580867767334,
      "epoch": 2.7342199285430726,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004258227655789651,
      "loss": 5.0534,
      "mean_token_accuracy": 0.1846708670258522,
      "num_tokens": 60356642.0,
      "step": 27550
    },
    {
      "entropy": 5.618023872375488,
      "epoch": 2.73471615720524,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004257964710909318,
      "loss": 4.9804,
      "mean_token_accuracy": 0.19373053461313247,
      "num_tokens": 60367244.0,
      "step": 27555
    },
    {
      "entropy": 5.6149262428283695,
      "epoch": 2.7352123858674076,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042577017286347116,
      "loss": 4.9672,
      "mean_token_accuracy": 0.18537162095308304,
      "num_tokens": 60377992.0,
      "step": 27560
    },
    {
      "entropy": 5.598627805709839,
      "epoch": 2.7357086145295755,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042574387089723504,
      "loss": 4.9574,
      "mean_token_accuracy": 0.18166160881519317,
      "num_tokens": 60389007.0,
      "step": 27565
    },
    {
      "entropy": 5.689659690856933,
      "epoch": 2.736204843191743,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042571756519287567,
      "loss": 5.073,
      "mean_token_accuracy": 0.17996560633182526,
      "num_tokens": 60399274.0,
      "step": 27570
    },
    {
      "entropy": 5.715211725234985,
      "epoch": 2.7367010718539104,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042569125575104557,
      "loss": 5.0599,
      "mean_token_accuracy": 0.17801235020160674,
      "num_tokens": 60411158.0,
      "step": 27575
    },
    {
      "entropy": 5.596199798583984,
      "epoch": 2.737197300516078,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042566494257239695,
      "loss": 4.9828,
      "mean_token_accuracy": 0.18941421508789064,
      "num_tokens": 60422790.0,
      "step": 27580
    },
    {
      "entropy": 5.67625470161438,
      "epoch": 2.7376935291782454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004256386256575824,
      "loss": 5.1592,
      "mean_token_accuracy": 0.1707532674074173,
      "num_tokens": 60434353.0,
      "step": 27585
    },
    {
      "entropy": 5.6257336139678955,
      "epoch": 2.738189757840413,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004256123050072546,
      "loss": 4.9478,
      "mean_token_accuracy": 0.18624568432569505,
      "num_tokens": 60445497.0,
      "step": 27590
    },
    {
      "entropy": 5.666214895248413,
      "epoch": 2.7386859865025803,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042558598062206616,
      "loss": 4.9418,
      "mean_token_accuracy": 0.18859639912843704,
      "num_tokens": 60456358.0,
      "step": 27595
    },
    {
      "entropy": 5.694129276275635,
      "epoch": 2.739182215164748,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042555965250266984,
      "loss": 5.0623,
      "mean_token_accuracy": 0.1844484493136406,
      "num_tokens": 60467613.0,
      "step": 27600
    },
    {
      "entropy": 5.668718004226685,
      "epoch": 2.7396784438269153,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042553332064971866,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18021908402442932,
      "num_tokens": 60479366.0,
      "step": 27605
    },
    {
      "entropy": 5.739984655380249,
      "epoch": 2.7401746724890828,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042550698506386536,
      "loss": 5.0505,
      "mean_token_accuracy": 0.184269580245018,
      "num_tokens": 60489199.0,
      "step": 27610
    },
    {
      "entropy": 5.6630054950714115,
      "epoch": 2.7406709011512502,
      "grad_norm": 1.125,
      "learning_rate": 0.00042548064574576323,
      "loss": 5.1094,
      "mean_token_accuracy": 0.18128141462802888,
      "num_tokens": 60500051.0,
      "step": 27615
    },
    {
      "entropy": 5.691911220550537,
      "epoch": 2.741167129813418,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004254543026960654,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1710179179906845,
      "num_tokens": 60510699.0,
      "step": 27620
    },
    {
      "entropy": 5.69548864364624,
      "epoch": 2.7416633584755856,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004254279559154249,
      "loss": 4.9627,
      "mean_token_accuracy": 0.18421933203935623,
      "num_tokens": 60520224.0,
      "step": 27625
    },
    {
      "entropy": 5.700532674789429,
      "epoch": 2.742159587137753,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042540160540449536,
      "loss": 5.0494,
      "mean_token_accuracy": 0.18278458565473557,
      "num_tokens": 60531829.0,
      "step": 27630
    },
    {
      "entropy": 5.590357160568237,
      "epoch": 2.7426558157999206,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004253752511639301,
      "loss": 5.0008,
      "mean_token_accuracy": 0.17670934051275253,
      "num_tokens": 60543278.0,
      "step": 27635
    },
    {
      "entropy": 5.702305603027344,
      "epoch": 2.743152044462088,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004253488931943826,
      "loss": 5.0806,
      "mean_token_accuracy": 0.1797234445810318,
      "num_tokens": 60554158.0,
      "step": 27640
    },
    {
      "entropy": 5.652312469482422,
      "epoch": 2.7436482731242555,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042532253149650653,
      "loss": 5.0242,
      "mean_token_accuracy": 0.18576118648052214,
      "num_tokens": 60563981.0,
      "step": 27645
    },
    {
      "entropy": 5.644956827163696,
      "epoch": 2.7441445017864234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004252961660709556,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18352962732315065,
      "num_tokens": 60574648.0,
      "step": 27650
    },
    {
      "entropy": 5.6348865032196045,
      "epoch": 2.744640730448591,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042526979691838353,
      "loss": 5.0253,
      "mean_token_accuracy": 0.1828178122639656,
      "num_tokens": 60585715.0,
      "step": 27655
    },
    {
      "entropy": 5.701448249816894,
      "epoch": 2.7451369591107584,
      "grad_norm": 1.125,
      "learning_rate": 0.00042524342403944446,
      "loss": 5.1421,
      "mean_token_accuracy": 0.17786087691783906,
      "num_tokens": 60597523.0,
      "step": 27660
    },
    {
      "entropy": 5.741191816329956,
      "epoch": 2.745633187772926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042521704743479213,
      "loss": 5.0969,
      "mean_token_accuracy": 0.18079868704080582,
      "num_tokens": 60607887.0,
      "step": 27665
    },
    {
      "entropy": 5.71139121055603,
      "epoch": 2.7461294164350933,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042519066710508064,
      "loss": 5.1143,
      "mean_token_accuracy": 0.18030357509851455,
      "num_tokens": 60619007.0,
      "step": 27670
    },
    {
      "entropy": 5.6411285400390625,
      "epoch": 2.746625645097261,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042516428305096435,
      "loss": 4.9773,
      "mean_token_accuracy": 0.19495926648378373,
      "num_tokens": 60629351.0,
      "step": 27675
    },
    {
      "entropy": 5.677299165725708,
      "epoch": 2.7471218737594283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004251378952730973,
      "loss": 5.0465,
      "mean_token_accuracy": 0.18491855561733245,
      "num_tokens": 60640145.0,
      "step": 27680
    },
    {
      "entropy": 5.662543296813965,
      "epoch": 2.7476181024215958,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004251115037721341,
      "loss": 5.0244,
      "mean_token_accuracy": 0.18983085453510284,
      "num_tokens": 60649981.0,
      "step": 27685
    },
    {
      "entropy": 5.692355394363403,
      "epoch": 2.7481143310837632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042508510854872905,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18195392340421676,
      "num_tokens": 60660572.0,
      "step": 27690
    },
    {
      "entropy": 5.6685370922088625,
      "epoch": 2.7486105597459307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042505870960353665,
      "loss": 5.0154,
      "mean_token_accuracy": 0.18353605568408965,
      "num_tokens": 60673136.0,
      "step": 27695
    },
    {
      "entropy": 5.723109102249145,
      "epoch": 2.749106788408098,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004250323069372116,
      "loss": 5.0825,
      "mean_token_accuracy": 0.17790765911340714,
      "num_tokens": 60683220.0,
      "step": 27700
    },
    {
      "entropy": 5.7365149974823,
      "epoch": 2.749603017070266,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042500590055040866,
      "loss": 5.0949,
      "mean_token_accuracy": 0.17926021963357924,
      "num_tokens": 60695030.0,
      "step": 27705
    },
    {
      "entropy": 5.6273040771484375,
      "epoch": 2.7500992457324336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042497949044378267,
      "loss": 5.0307,
      "mean_token_accuracy": 0.17840767204761504,
      "num_tokens": 60705268.0,
      "step": 27710
    },
    {
      "entropy": 5.628432035446167,
      "epoch": 2.750595474394601,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042495307661798837,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18692008256912232,
      "num_tokens": 60716091.0,
      "step": 27715
    },
    {
      "entropy": 5.725919771194458,
      "epoch": 2.7510917030567685,
      "grad_norm": 1.125,
      "learning_rate": 0.00042492665907368094,
      "loss": 5.0905,
      "mean_token_accuracy": 0.18114401251077653,
      "num_tokens": 60727255.0,
      "step": 27720
    },
    {
      "entropy": 5.685586261749267,
      "epoch": 2.751587931718936,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042490023781151546,
      "loss": 5.0475,
      "mean_token_accuracy": 0.18139415681362153,
      "num_tokens": 60738014.0,
      "step": 27725
    },
    {
      "entropy": 5.704896450042725,
      "epoch": 2.7520841603811035,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042487381283214706,
      "loss": 5.1536,
      "mean_token_accuracy": 0.17577614188194274,
      "num_tokens": 60748935.0,
      "step": 27730
    },
    {
      "entropy": 5.6499535083770756,
      "epoch": 2.7525803890432714,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004248473841362311,
      "loss": 5.0046,
      "mean_token_accuracy": 0.1924273833632469,
      "num_tokens": 60760046.0,
      "step": 27735
    },
    {
      "entropy": 5.683707189559937,
      "epoch": 2.753076617705439,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042482095172442286,
      "loss": 5.0623,
      "mean_token_accuracy": 0.18121807873249055,
      "num_tokens": 60771198.0,
      "step": 27740
    },
    {
      "entropy": 5.679253053665161,
      "epoch": 2.7535728463676064,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004247945155973778,
      "loss": 5.0978,
      "mean_token_accuracy": 0.18010537326335907,
      "num_tokens": 60782018.0,
      "step": 27745
    },
    {
      "entropy": 5.754214286804199,
      "epoch": 2.754069075029774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004247680757557516,
      "loss": 5.1105,
      "mean_token_accuracy": 0.18158157765865326,
      "num_tokens": 60794108.0,
      "step": 27750
    },
    {
      "entropy": 5.706709718704223,
      "epoch": 2.7545653036919413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042474163220019987,
      "loss": 5.0063,
      "mean_token_accuracy": 0.18334481716156006,
      "num_tokens": 60803941.0,
      "step": 27755
    },
    {
      "entropy": 5.689007043838501,
      "epoch": 2.7550615323541088,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042471518493137836,
      "loss": 5.0392,
      "mean_token_accuracy": 0.1877899318933487,
      "num_tokens": 60814772.0,
      "step": 27760
    },
    {
      "entropy": 5.728980922698975,
      "epoch": 2.7555577610162763,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004246887339499427,
      "loss": 5.116,
      "mean_token_accuracy": 0.1743614301085472,
      "num_tokens": 60825289.0,
      "step": 27765
    },
    {
      "entropy": 5.674287462234497,
      "epoch": 2.7560539896784437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042466227925654917,
      "loss": 5.0573,
      "mean_token_accuracy": 0.1797366499900818,
      "num_tokens": 60836440.0,
      "step": 27770
    },
    {
      "entropy": 5.675889205932617,
      "epoch": 2.756550218340611,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004246358208518535,
      "loss": 5.0372,
      "mean_token_accuracy": 0.18264016807079314,
      "num_tokens": 60847002.0,
      "step": 27775
    },
    {
      "entropy": 5.6929505348205565,
      "epoch": 2.7570464470027787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042460935873651196,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1795224964618683,
      "num_tokens": 60858240.0,
      "step": 27780
    },
    {
      "entropy": 5.653959083557129,
      "epoch": 2.757542675664946,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042458289291118066,
      "loss": 5.0152,
      "mean_token_accuracy": 0.18393053859472275,
      "num_tokens": 60870018.0,
      "step": 27785
    },
    {
      "entropy": 5.668621301651001,
      "epoch": 2.758038904327114,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042455642337651595,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18775398880243302,
      "num_tokens": 60881143.0,
      "step": 27790
    },
    {
      "entropy": 5.675235891342163,
      "epoch": 2.7585351329892815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004245299501331743,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1785864531993866,
      "num_tokens": 60892347.0,
      "step": 27795
    },
    {
      "entropy": 5.712392902374267,
      "epoch": 2.759031361651449,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042450347318181194,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17859435081481934,
      "num_tokens": 60903420.0,
      "step": 27800
    },
    {
      "entropy": 5.656671714782715,
      "epoch": 2.7595275903136165,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042447699252308566,
      "loss": 4.9361,
      "mean_token_accuracy": 0.19202478229999542,
      "num_tokens": 60913213.0,
      "step": 27805
    },
    {
      "entropy": 5.72378511428833,
      "epoch": 2.760023818975784,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004244505081576521,
      "loss": 5.0488,
      "mean_token_accuracy": 0.18322014212608337,
      "num_tokens": 60924505.0,
      "step": 27810
    },
    {
      "entropy": 5.683966827392578,
      "epoch": 2.7605200476379514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042442402008616793,
      "loss": 5.0382,
      "mean_token_accuracy": 0.18681470304727554,
      "num_tokens": 60934583.0,
      "step": 27815
    },
    {
      "entropy": 5.68596887588501,
      "epoch": 2.7610162763001194,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042439752830929,
      "loss": 5.0409,
      "mean_token_accuracy": 0.18116606622934342,
      "num_tokens": 60944471.0,
      "step": 27820
    },
    {
      "entropy": 5.690449905395508,
      "epoch": 2.761512504962287,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004243710328276754,
      "loss": 5.0129,
      "mean_token_accuracy": 0.1838294669985771,
      "num_tokens": 60954298.0,
      "step": 27825
    },
    {
      "entropy": 5.6858008861541744,
      "epoch": 2.7620087336244543,
      "grad_norm": 1.03125,
      "learning_rate": 0.000424344533641981,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18820427656173705,
      "num_tokens": 60965885.0,
      "step": 27830
    },
    {
      "entropy": 5.654649639129639,
      "epoch": 2.762504962286622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000424318030752864,
      "loss": 5.0222,
      "mean_token_accuracy": 0.1802108347415924,
      "num_tokens": 60977177.0,
      "step": 27835
    },
    {
      "entropy": 5.693717765808105,
      "epoch": 2.7630011909487893,
      "grad_norm": 1.125,
      "learning_rate": 0.0004242915241609815,
      "loss": 5.0985,
      "mean_token_accuracy": 0.17906143367290497,
      "num_tokens": 60988530.0,
      "step": 27840
    },
    {
      "entropy": 5.661290645599365,
      "epoch": 2.7634974196109567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000424265013866991,
      "loss": 4.9497,
      "mean_token_accuracy": 0.189732326567173,
      "num_tokens": 60999298.0,
      "step": 27845
    },
    {
      "entropy": 5.677545881271362,
      "epoch": 2.763993648273124,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004242384998715497,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1913491189479828,
      "num_tokens": 61010641.0,
      "step": 27850
    },
    {
      "entropy": 5.710785245895385,
      "epoch": 2.7644898769352917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004242119821753153,
      "loss": 5.1237,
      "mean_token_accuracy": 0.18176388889551162,
      "num_tokens": 61021597.0,
      "step": 27855
    },
    {
      "entropy": 5.719920349121094,
      "epoch": 2.764986105597459,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004241854607789452,
      "loss": 5.0535,
      "mean_token_accuracy": 0.18060633838176726,
      "num_tokens": 61033507.0,
      "step": 27860
    },
    {
      "entropy": 5.648094224929809,
      "epoch": 2.7654823342596266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042415893568309714,
      "loss": 4.9372,
      "mean_token_accuracy": 0.1850799575448036,
      "num_tokens": 61044302.0,
      "step": 27865
    },
    {
      "entropy": 5.600972652435303,
      "epoch": 2.765978562921794,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042413240688842884,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1835790231823921,
      "num_tokens": 61056262.0,
      "step": 27870
    },
    {
      "entropy": 5.666871833801269,
      "epoch": 2.766474791583962,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004241058743955982,
      "loss": 5.0659,
      "mean_token_accuracy": 0.18694402724504472,
      "num_tokens": 61068689.0,
      "step": 27875
    },
    {
      "entropy": 5.7077583312988285,
      "epoch": 2.7669710202461295,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004240793382052633,
      "loss": 5.0311,
      "mean_token_accuracy": 0.18845242410898208,
      "num_tokens": 61080151.0,
      "step": 27880
    },
    {
      "entropy": 5.645232486724853,
      "epoch": 2.767467248908297,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000424052798318082,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18435874134302138,
      "num_tokens": 61090488.0,
      "step": 27885
    },
    {
      "entropy": 5.5971122741699215,
      "epoch": 2.7679634775704645,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004240262547347124,
      "loss": 4.9145,
      "mean_token_accuracy": 0.18984893560409546,
      "num_tokens": 61101250.0,
      "step": 27890
    },
    {
      "entropy": 5.648252534866333,
      "epoch": 2.768459706232632,
      "grad_norm": 1.125,
      "learning_rate": 0.00042399970745581283,
      "loss": 4.9753,
      "mean_token_accuracy": 0.18816154450178146,
      "num_tokens": 61112920.0,
      "step": 27895
    },
    {
      "entropy": 5.693484592437744,
      "epoch": 2.7689559348947994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042397315648204164,
      "loss": 4.9921,
      "mean_token_accuracy": 0.1941028892993927,
      "num_tokens": 61122716.0,
      "step": 27900
    },
    {
      "entropy": 5.6644508838653564,
      "epoch": 2.7694521635569673,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042394660181405716,
      "loss": 5.0369,
      "mean_token_accuracy": 0.1849090874195099,
      "num_tokens": 61133794.0,
      "step": 27905
    },
    {
      "entropy": 5.736408329010009,
      "epoch": 2.769948392219135,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004239200434525179,
      "loss": 5.0728,
      "mean_token_accuracy": 0.18088562935590743,
      "num_tokens": 61145690.0,
      "step": 27910
    },
    {
      "entropy": 5.649764156341552,
      "epoch": 2.7704446208813023,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004238934813980824,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18138804882764817,
      "num_tokens": 61157111.0,
      "step": 27915
    },
    {
      "entropy": 5.649707555770874,
      "epoch": 2.7709408495434698,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042386691565140947,
      "loss": 5.0284,
      "mean_token_accuracy": 0.18674833923578263,
      "num_tokens": 61168380.0,
      "step": 27920
    },
    {
      "entropy": 5.72059907913208,
      "epoch": 2.7714370782056372,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042384034621315773,
      "loss": 5.1273,
      "mean_token_accuracy": 0.17520366758108138,
      "num_tokens": 61179318.0,
      "step": 27925
    },
    {
      "entropy": 5.689018106460571,
      "epoch": 2.7719333068678047,
      "grad_norm": 1.125,
      "learning_rate": 0.0004238137730839862,
      "loss": 5.0852,
      "mean_token_accuracy": 0.17932333946228027,
      "num_tokens": 61190448.0,
      "step": 27930
    },
    {
      "entropy": 5.660954332351684,
      "epoch": 2.772429535529972,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004237871962645538,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18664309978485108,
      "num_tokens": 61200746.0,
      "step": 27935
    },
    {
      "entropy": 5.635933113098145,
      "epoch": 2.7729257641921397,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042376061575551946,
      "loss": 4.9829,
      "mean_token_accuracy": 0.18787101209163665,
      "num_tokens": 61213093.0,
      "step": 27940
    },
    {
      "entropy": 5.617995882034302,
      "epoch": 2.773421992854307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004237340315575425,
      "loss": 4.9602,
      "mean_token_accuracy": 0.18995022475719453,
      "num_tokens": 61223749.0,
      "step": 27945
    },
    {
      "entropy": 5.671149730682373,
      "epoch": 2.7739182215164746,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004237074436712819,
      "loss": 5.0667,
      "mean_token_accuracy": 0.17428015768527985,
      "num_tokens": 61234755.0,
      "step": 27950
    },
    {
      "entropy": 5.753393983840942,
      "epoch": 2.774414450178642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042368085209739715,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17827987670898438,
      "num_tokens": 61244669.0,
      "step": 27955
    },
    {
      "entropy": 5.6724865436553955,
      "epoch": 2.77491067884081,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004236542568365478,
      "loss": 4.9866,
      "mean_token_accuracy": 0.18747484385967256,
      "num_tokens": 61257240.0,
      "step": 27960
    },
    {
      "entropy": 5.636288356781006,
      "epoch": 2.7754069075029775,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042362765788939307,
      "loss": 4.9301,
      "mean_token_accuracy": 0.18918218612670898,
      "num_tokens": 61267389.0,
      "step": 27965
    },
    {
      "entropy": 5.577549695968628,
      "epoch": 2.775903136165145,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004236010552565927,
      "loss": 5.004,
      "mean_token_accuracy": 0.1917399913072586,
      "num_tokens": 61277381.0,
      "step": 27970
    },
    {
      "entropy": 5.631432151794433,
      "epoch": 2.7763993648273124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042357444893880645,
      "loss": 5.018,
      "mean_token_accuracy": 0.18363560140132903,
      "num_tokens": 61286931.0,
      "step": 27975
    },
    {
      "entropy": 5.692963457107544,
      "epoch": 2.77689559348948,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042354783893669385,
      "loss": 4.9918,
      "mean_token_accuracy": 0.18455958366394043,
      "num_tokens": 61297706.0,
      "step": 27980
    },
    {
      "entropy": 5.733993816375732,
      "epoch": 2.7773918221516474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004235212252509151,
      "loss": 5.0548,
      "mean_token_accuracy": 0.17550714313983917,
      "num_tokens": 61308516.0,
      "step": 27985
    },
    {
      "entropy": 5.706290006637573,
      "epoch": 2.777888050813815,
      "grad_norm": 1.0,
      "learning_rate": 0.00042349460788212995,
      "loss": 5.1109,
      "mean_token_accuracy": 0.17961603850126268,
      "num_tokens": 61319454.0,
      "step": 27990
    },
    {
      "entropy": 5.740304040908813,
      "epoch": 2.7783842794759828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042346798683099844,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1722816124558449,
      "num_tokens": 61330970.0,
      "step": 27995
    },
    {
      "entropy": 5.670155811309814,
      "epoch": 2.7788805081381502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004234413620981809,
      "loss": 4.9412,
      "mean_token_accuracy": 0.1919977992773056,
      "num_tokens": 61342466.0,
      "step": 28000
    },
    {
      "entropy": 5.665373659133911,
      "epoch": 2.7793767368003177,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042341473368433735,
      "loss": 4.9607,
      "mean_token_accuracy": 0.19382859617471696,
      "num_tokens": 61351988.0,
      "step": 28005
    },
    {
      "entropy": 5.6172754764556885,
      "epoch": 2.779872965462485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004233881015901282,
      "loss": 4.9931,
      "mean_token_accuracy": 0.18219754695892335,
      "num_tokens": 61364270.0,
      "step": 28010
    },
    {
      "entropy": 5.6524553298950195,
      "epoch": 2.7803691941246527,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004233614658162139,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18861483335494994,
      "num_tokens": 61374996.0,
      "step": 28015
    },
    {
      "entropy": 5.627207899093628,
      "epoch": 2.78086542278682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042333482636325504,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18344692438840865,
      "num_tokens": 61385394.0,
      "step": 28020
    },
    {
      "entropy": 5.65884690284729,
      "epoch": 2.7813616514489876,
      "grad_norm": 1.046875,
      "learning_rate": 0.000423308183231912,
      "loss": 5.0099,
      "mean_token_accuracy": 0.18317780643701553,
      "num_tokens": 61397110.0,
      "step": 28025
    },
    {
      "entropy": 5.6368104934692385,
      "epoch": 2.781857880111155,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042328153642284567,
      "loss": 5.0522,
      "mean_token_accuracy": 0.18481992036104203,
      "num_tokens": 61408462.0,
      "step": 28030
    },
    {
      "entropy": 5.676920557022095,
      "epoch": 2.7823541087733226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004232548859367167,
      "loss": 5.0301,
      "mean_token_accuracy": 0.1808696374297142,
      "num_tokens": 61419543.0,
      "step": 28035
    },
    {
      "entropy": 5.662796878814698,
      "epoch": 2.78285033743549,
      "grad_norm": 1.0625,
      "learning_rate": 0.000423228231774186,
      "loss": 5.0047,
      "mean_token_accuracy": 0.18083580285310746,
      "num_tokens": 61430476.0,
      "step": 28040
    },
    {
      "entropy": 5.652567148208618,
      "epoch": 2.783346566097658,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004232015739359146,
      "loss": 5.021,
      "mean_token_accuracy": 0.18551425784826278,
      "num_tokens": 61440384.0,
      "step": 28045
    },
    {
      "entropy": 5.6659903049469,
      "epoch": 2.7838427947598254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042317491242256346,
      "loss": 5.0731,
      "mean_token_accuracy": 0.18212960064411163,
      "num_tokens": 61450957.0,
      "step": 28050
    },
    {
      "entropy": 5.641305303573608,
      "epoch": 2.784339023421993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004231482472347938,
      "loss": 5.0476,
      "mean_token_accuracy": 0.19033270031213761,
      "num_tokens": 61462033.0,
      "step": 28055
    },
    {
      "entropy": 5.738147830963134,
      "epoch": 2.7848352520841604,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004231215783732668,
      "loss": 5.0907,
      "mean_token_accuracy": 0.17696181684732437,
      "num_tokens": 61472153.0,
      "step": 28060
    },
    {
      "entropy": 5.746042966842651,
      "epoch": 2.785331480746328,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00042309490583864395,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1830942615866661,
      "num_tokens": 61483836.0,
      "step": 28065
    },
    {
      "entropy": 5.677969217300415,
      "epoch": 2.7858277094084953,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004230682296315865,
      "loss": 5.0178,
      "mean_token_accuracy": 0.18994626104831697,
      "num_tokens": 61494819.0,
      "step": 28070
    },
    {
      "entropy": 5.695749711990357,
      "epoch": 2.786323938070663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000423041549752756,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18049008399248123,
      "num_tokens": 61506385.0,
      "step": 28075
    },
    {
      "entropy": 5.696149921417236,
      "epoch": 2.7868201667328307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004230148662028141,
      "loss": 5.0197,
      "mean_token_accuracy": 0.18077728599309922,
      "num_tokens": 61517679.0,
      "step": 28080
    },
    {
      "entropy": 5.666807365417481,
      "epoch": 2.787316395394998,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042298817898242237,
      "loss": 5.0181,
      "mean_token_accuracy": 0.19297779202461243,
      "num_tokens": 61529548.0,
      "step": 28085
    },
    {
      "entropy": 5.688243055343628,
      "epoch": 2.7878126240571657,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042296148809224275,
      "loss": 5.122,
      "mean_token_accuracy": 0.17636170238256454,
      "num_tokens": 61540341.0,
      "step": 28090
    },
    {
      "entropy": 5.728671836853027,
      "epoch": 2.788308852719333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004229347935329371,
      "loss": 5.0063,
      "mean_token_accuracy": 0.1879989743232727,
      "num_tokens": 61550692.0,
      "step": 28095
    },
    {
      "entropy": 5.671541690826416,
      "epoch": 2.7888050813815006,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042290809530516724,
      "loss": 4.9679,
      "mean_token_accuracy": 0.18136851340532303,
      "num_tokens": 61562126.0,
      "step": 28100
    },
    {
      "entropy": 5.62607479095459,
      "epoch": 2.789301310043668,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004228813934095954,
      "loss": 5.0038,
      "mean_token_accuracy": 0.18457031548023223,
      "num_tokens": 61573109.0,
      "step": 28105
    },
    {
      "entropy": 5.673747158050537,
      "epoch": 2.7897975387058356,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042285468784688363,
      "loss": 5.0687,
      "mean_token_accuracy": 0.18172443956136702,
      "num_tokens": 61584389.0,
      "step": 28110
    },
    {
      "entropy": 5.682768058776856,
      "epoch": 2.790293767368003,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042282797861769423,
      "loss": 5.0176,
      "mean_token_accuracy": 0.1843493774533272,
      "num_tokens": 61595173.0,
      "step": 28115
    },
    {
      "entropy": 5.687092304229736,
      "epoch": 2.7907899960301705,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004228012657226895,
      "loss": 5.0627,
      "mean_token_accuracy": 0.1806496262550354,
      "num_tokens": 61605077.0,
      "step": 28120
    },
    {
      "entropy": 5.735343742370605,
      "epoch": 2.791286224692338,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042277454916253186,
      "loss": 5.0972,
      "mean_token_accuracy": 0.17761319726705552,
      "num_tokens": 61617265.0,
      "step": 28125
    },
    {
      "entropy": 5.681867885589599,
      "epoch": 2.791782453354506,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004227478289378838,
      "loss": 5.0509,
      "mean_token_accuracy": 0.17891421765089036,
      "num_tokens": 61629441.0,
      "step": 28130
    },
    {
      "entropy": 5.642759895324707,
      "epoch": 2.7922786820166734,
      "grad_norm": 1.25,
      "learning_rate": 0.00042272110504940795,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1780583843588829,
      "num_tokens": 61641032.0,
      "step": 28135
    },
    {
      "entropy": 5.697315645217896,
      "epoch": 2.792774910678841,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042269437749776703,
      "loss": 5.0428,
      "mean_token_accuracy": 0.17877932637929916,
      "num_tokens": 61652625.0,
      "step": 28140
    },
    {
      "entropy": 5.674192762374878,
      "epoch": 2.7932711393410083,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004226676462836238,
      "loss": 4.9468,
      "mean_token_accuracy": 0.19116777032613755,
      "num_tokens": 61664386.0,
      "step": 28145
    },
    {
      "entropy": 5.677223253250122,
      "epoch": 2.793767368003176,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004226409114076411,
      "loss": 5.002,
      "mean_token_accuracy": 0.18810195624828338,
      "num_tokens": 61675040.0,
      "step": 28150
    },
    {
      "entropy": 5.591964292526245,
      "epoch": 2.7942635966653433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042261417287048196,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18577542155981064,
      "num_tokens": 61684702.0,
      "step": 28155
    },
    {
      "entropy": 5.60604567527771,
      "epoch": 2.7947598253275108,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004225874306728094,
      "loss": 5.0568,
      "mean_token_accuracy": 0.18942238092422486,
      "num_tokens": 61695888.0,
      "step": 28160
    },
    {
      "entropy": 5.69886212348938,
      "epoch": 2.7952560539896787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042256068481528666,
      "loss": 5.0576,
      "mean_token_accuracy": 0.18161840587854386,
      "num_tokens": 61705655.0,
      "step": 28165
    },
    {
      "entropy": 5.713629865646363,
      "epoch": 2.795752282651846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004225339352985768,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18283617198467256,
      "num_tokens": 61715365.0,
      "step": 28170
    },
    {
      "entropy": 5.6845383644104,
      "epoch": 2.7962485113140136,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042250718212334334,
      "loss": 5.1017,
      "mean_token_accuracy": 0.1800490215420723,
      "num_tokens": 61727104.0,
      "step": 28175
    },
    {
      "entropy": 5.667980432510376,
      "epoch": 2.796744739976181,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042248042529024956,
      "loss": 5.0302,
      "mean_token_accuracy": 0.17897475957870485,
      "num_tokens": 61737000.0,
      "step": 28180
    },
    {
      "entropy": 5.606955051422119,
      "epoch": 2.7972409686383486,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042245366479995894,
      "loss": 4.994,
      "mean_token_accuracy": 0.18621134907007217,
      "num_tokens": 61748801.0,
      "step": 28185
    },
    {
      "entropy": 5.64434175491333,
      "epoch": 2.797737197300516,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042242690065313535,
      "loss": 5.0239,
      "mean_token_accuracy": 0.1843794047832489,
      "num_tokens": 61760202.0,
      "step": 28190
    },
    {
      "entropy": 5.697030162811279,
      "epoch": 2.7982334259626835,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042240013285044215,
      "loss": 5.0878,
      "mean_token_accuracy": 0.18190304040908814,
      "num_tokens": 61770886.0,
      "step": 28195
    },
    {
      "entropy": 5.7401893615722654,
      "epoch": 2.798729654624851,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042237336139254336,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1807926893234253,
      "num_tokens": 61781464.0,
      "step": 28200
    },
    {
      "entropy": 5.616227531433106,
      "epoch": 2.7992258832870185,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042234658628010274,
      "loss": 4.9628,
      "mean_token_accuracy": 0.18387981802225112,
      "num_tokens": 61791183.0,
      "step": 28205
    },
    {
      "entropy": 5.639111852645874,
      "epoch": 2.799722111949186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042231980751378435,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18143371492624283,
      "num_tokens": 61801113.0,
      "step": 28210
    },
    {
      "entropy": 5.696996355056763,
      "epoch": 2.8002183406113534,
      "grad_norm": 1.125,
      "learning_rate": 0.00042229302509425205,
      "loss": 5.0973,
      "mean_token_accuracy": 0.1762745574116707,
      "num_tokens": 61811806.0,
      "step": 28215
    },
    {
      "entropy": 5.758212041854859,
      "epoch": 2.8007145692735214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004222662390221702,
      "loss": 5.0746,
      "mean_token_accuracy": 0.1820446327328682,
      "num_tokens": 61822279.0,
      "step": 28220
    },
    {
      "entropy": 5.647297811508179,
      "epoch": 2.801210797935689,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004222394492982029,
      "loss": 4.9682,
      "mean_token_accuracy": 0.18418197631835936,
      "num_tokens": 61832303.0,
      "step": 28225
    },
    {
      "entropy": 5.745135498046875,
      "epoch": 2.8017070265978563,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222126559230146,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1801106944680214,
      "num_tokens": 61844187.0,
      "step": 28230
    },
    {
      "entropy": 5.592764186859131,
      "epoch": 2.802203255260024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004221858588972696,
      "loss": 4.963,
      "mean_token_accuracy": 0.1872100353240967,
      "num_tokens": 61855375.0,
      "step": 28235
    },
    {
      "entropy": 5.651000499725342,
      "epoch": 2.8026994839221913,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042215905822163245,
      "loss": 5.0207,
      "mean_token_accuracy": 0.1774177610874176,
      "num_tokens": 61865810.0,
      "step": 28240
    },
    {
      "entropy": 5.637115097045898,
      "epoch": 2.8031957125843587,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004221322538967678,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18531113862991333,
      "num_tokens": 61876641.0,
      "step": 28245
    },
    {
      "entropy": 5.670738744735718,
      "epoch": 2.8036919412465267,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004221054459233403,
      "loss": 5.0255,
      "mean_token_accuracy": 0.19179772585630417,
      "num_tokens": 61887854.0,
      "step": 28250
    },
    {
      "entropy": 5.73931303024292,
      "epoch": 2.804188169908694,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042207863430201466,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1794213905930519,
      "num_tokens": 61898879.0,
      "step": 28255
    },
    {
      "entropy": 5.697976541519165,
      "epoch": 2.8046843985708616,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042205181903345586,
      "loss": 5.0426,
      "mean_token_accuracy": 0.17621983885765075,
      "num_tokens": 61909425.0,
      "step": 28260
    },
    {
      "entropy": 5.683617687225341,
      "epoch": 2.805180627233029,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004220250001183288,
      "loss": 5.0298,
      "mean_token_accuracy": 0.18984633237123488,
      "num_tokens": 61922097.0,
      "step": 28265
    },
    {
      "entropy": 5.688587331771851,
      "epoch": 2.8056768558951966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004219981775572986,
      "loss": 4.966,
      "mean_token_accuracy": 0.19014500826597214,
      "num_tokens": 61932457.0,
      "step": 28270
    },
    {
      "entropy": 5.654702472686767,
      "epoch": 2.806173084557364,
      "grad_norm": 1.125,
      "learning_rate": 0.0004219713513510303,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1881965696811676,
      "num_tokens": 61942354.0,
      "step": 28275
    },
    {
      "entropy": 5.6790442943573,
      "epoch": 2.8066693132195315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042194452150018917,
      "loss": 5.0379,
      "mean_token_accuracy": 0.18333384096622468,
      "num_tokens": 61952026.0,
      "step": 28280
    },
    {
      "entropy": 5.612762403488159,
      "epoch": 2.807165541881699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042191768800544055,
      "loss": 4.9866,
      "mean_token_accuracy": 0.18618517816066743,
      "num_tokens": 61961976.0,
      "step": 28285
    },
    {
      "entropy": 5.623417520523072,
      "epoch": 2.8076617705438665,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042189085086744986,
      "loss": 5.0268,
      "mean_token_accuracy": 0.18981678634881974,
      "num_tokens": 61974529.0,
      "step": 28290
    },
    {
      "entropy": 5.679486083984375,
      "epoch": 2.808157999206034,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004218640100868826,
      "loss": 5.0878,
      "mean_token_accuracy": 0.18107393383979797,
      "num_tokens": 61985912.0,
      "step": 28295
    },
    {
      "entropy": 5.704853534698486,
      "epoch": 2.8086542278682014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042183716566440433,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17957653254270553,
      "num_tokens": 61997534.0,
      "step": 28300
    },
    {
      "entropy": 5.6744460105896,
      "epoch": 2.8091504565303693,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004218103176006808,
      "loss": 4.9445,
      "mean_token_accuracy": 0.1822303518652916,
      "num_tokens": 62008292.0,
      "step": 28305
    },
    {
      "entropy": 5.619553279876709,
      "epoch": 2.809646685192537,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004217834658963777,
      "loss": 5.0052,
      "mean_token_accuracy": 0.18934312611818313,
      "num_tokens": 62019005.0,
      "step": 28310
    },
    {
      "entropy": 5.66877007484436,
      "epoch": 2.8101429138547043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042175661055216093,
      "loss": 5.0806,
      "mean_token_accuracy": 0.17808687686920166,
      "num_tokens": 62030127.0,
      "step": 28315
    },
    {
      "entropy": 5.669171237945557,
      "epoch": 2.8106391425168717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004217297515686965,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17438411861658096,
      "num_tokens": 62041080.0,
      "step": 28320
    },
    {
      "entropy": 5.633999443054199,
      "epoch": 2.811135371179039,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004217028889466503,
      "loss": 4.9932,
      "mean_token_accuracy": 0.18761467933654785,
      "num_tokens": 62051420.0,
      "step": 28325
    },
    {
      "entropy": 5.57522029876709,
      "epoch": 2.8116315998412067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004216760226866887,
      "loss": 4.9269,
      "mean_token_accuracy": 0.18232733011245728,
      "num_tokens": 62063303.0,
      "step": 28330
    },
    {
      "entropy": 5.7022271156311035,
      "epoch": 2.8121278285033746,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042164915278947775,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18366734236478804,
      "num_tokens": 62074654.0,
      "step": 28335
    },
    {
      "entropy": 5.6571790218353275,
      "epoch": 2.812624057165542,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004216222792556838,
      "loss": 4.9232,
      "mean_token_accuracy": 0.19052259176969527,
      "num_tokens": 62085872.0,
      "step": 28340
    },
    {
      "entropy": 5.630660581588745,
      "epoch": 2.8131202858277096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004215954020859732,
      "loss": 4.9826,
      "mean_token_accuracy": 0.189591483771801,
      "num_tokens": 62095185.0,
      "step": 28345
    },
    {
      "entropy": 5.68410096168518,
      "epoch": 2.813616514489877,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042156852128101264,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18767166882753372,
      "num_tokens": 62105643.0,
      "step": 28350
    },
    {
      "entropy": 5.60864953994751,
      "epoch": 2.8141127431520445,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042154163684146855,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1906431183218956,
      "num_tokens": 62115290.0,
      "step": 28355
    },
    {
      "entropy": 5.633398723602295,
      "epoch": 2.814608971814212,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004215147487680076,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18802594840526582,
      "num_tokens": 62125493.0,
      "step": 28360
    },
    {
      "entropy": 5.665056371688843,
      "epoch": 2.8151052004763795,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042148785706129657,
      "loss": 5.0353,
      "mean_token_accuracy": 0.1783938392996788,
      "num_tokens": 62135813.0,
      "step": 28365
    },
    {
      "entropy": 5.614195442199707,
      "epoch": 2.815601429138547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004214609617220024,
      "loss": 4.9826,
      "mean_token_accuracy": 0.19315813034772872,
      "num_tokens": 62146423.0,
      "step": 28370
    },
    {
      "entropy": 5.666711282730103,
      "epoch": 2.8160976578007144,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042143406275079186,
      "loss": 5.0842,
      "mean_token_accuracy": 0.18906779289245607,
      "num_tokens": 62157601.0,
      "step": 28375
    },
    {
      "entropy": 5.684387159347534,
      "epoch": 2.816593886462882,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004214071601483322,
      "loss": 5.0851,
      "mean_token_accuracy": 0.18194715529680253,
      "num_tokens": 62167918.0,
      "step": 28380
    },
    {
      "entropy": 5.71801347732544,
      "epoch": 2.8170901151250494,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004213802539152905,
      "loss": 5.0083,
      "mean_token_accuracy": 0.19121670573949814,
      "num_tokens": 62179176.0,
      "step": 28385
    },
    {
      "entropy": 5.710136985778808,
      "epoch": 2.8175863437872173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042135334405233384,
      "loss": 5.0292,
      "mean_token_accuracy": 0.19429441243410112,
      "num_tokens": 62190238.0,
      "step": 28390
    },
    {
      "entropy": 5.653854846954346,
      "epoch": 2.8180825724493848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004213264305601296,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18345659077167512,
      "num_tokens": 62202041.0,
      "step": 28395
    },
    {
      "entropy": 5.628012990951538,
      "epoch": 2.8185788011115522,
      "grad_norm": 1.125,
      "learning_rate": 0.00042129951343934527,
      "loss": 5.04,
      "mean_token_accuracy": 0.18051311522722244,
      "num_tokens": 62213309.0,
      "step": 28400
    },
    {
      "entropy": 5.718500995635987,
      "epoch": 2.8190750297737197,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004212725926906482,
      "loss": 5.0987,
      "mean_token_accuracy": 0.1836406722664833,
      "num_tokens": 62223936.0,
      "step": 28405
    },
    {
      "entropy": 5.665398025512696,
      "epoch": 2.819571258435887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042124566831470604,
      "loss": 4.9814,
      "mean_token_accuracy": 0.19401148706674576,
      "num_tokens": 62233800.0,
      "step": 28410
    },
    {
      "entropy": 5.723420858383179,
      "epoch": 2.8200674870980547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004212187403121864,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18786625266075135,
      "num_tokens": 62245888.0,
      "step": 28415
    },
    {
      "entropy": 5.623145532608032,
      "epoch": 2.8205637157602226,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042119180868375707,
      "loss": 4.9752,
      "mean_token_accuracy": 0.1906639114022255,
      "num_tokens": 62255796.0,
      "step": 28420
    },
    {
      "entropy": 5.687994575500488,
      "epoch": 2.82105994442239,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004211648734300859,
      "loss": 5.1109,
      "mean_token_accuracy": 0.1851850390434265,
      "num_tokens": 62266463.0,
      "step": 28425
    },
    {
      "entropy": 5.6016090393066404,
      "epoch": 2.8215561730845575,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042113793455184084,
      "loss": 4.9779,
      "mean_token_accuracy": 0.18710190206766128,
      "num_tokens": 62278077.0,
      "step": 28430
    },
    {
      "entropy": 5.703857707977295,
      "epoch": 2.822052401746725,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004211109920496899,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1886189326643944,
      "num_tokens": 62289489.0,
      "step": 28435
    },
    {
      "entropy": 5.635318231582642,
      "epoch": 2.8225486304088925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042108404592430126,
      "loss": 4.9373,
      "mean_token_accuracy": 0.1961933732032776,
      "num_tokens": 62301402.0,
      "step": 28440
    },
    {
      "entropy": 5.615809774398803,
      "epoch": 2.82304485907106,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000421057096176343,
      "loss": 4.9112,
      "mean_token_accuracy": 0.1869183287024498,
      "num_tokens": 62311253.0,
      "step": 28445
    },
    {
      "entropy": 5.612711000442505,
      "epoch": 2.8235410877332274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004210301428064835,
      "loss": 4.9903,
      "mean_token_accuracy": 0.1841856449842453,
      "num_tokens": 62321312.0,
      "step": 28450
    },
    {
      "entropy": 5.650890827178955,
      "epoch": 2.824037316395395,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004210031858153911,
      "loss": 5.0422,
      "mean_token_accuracy": 0.18081861436367036,
      "num_tokens": 62332318.0,
      "step": 28455
    },
    {
      "entropy": 5.722333478927612,
      "epoch": 2.8245335450575624,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042097622520373434,
      "loss": 5.0307,
      "mean_token_accuracy": 0.1882147029042244,
      "num_tokens": 62341945.0,
      "step": 28460
    },
    {
      "entropy": 5.663328647613525,
      "epoch": 2.82502977371973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004209492609721817,
      "loss": 5.0205,
      "mean_token_accuracy": 0.18918772786855698,
      "num_tokens": 62353509.0,
      "step": 28465
    },
    {
      "entropy": 5.649641323089599,
      "epoch": 2.8255260023818973,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004209222931214019,
      "loss": 5.0183,
      "mean_token_accuracy": 0.1876890853047371,
      "num_tokens": 62364044.0,
      "step": 28470
    },
    {
      "entropy": 5.605014705657959,
      "epoch": 2.8260222310440652,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042089532165206367,
      "loss": 4.9118,
      "mean_token_accuracy": 0.18637754917144775,
      "num_tokens": 62375113.0,
      "step": 28475
    },
    {
      "entropy": 5.638198328018189,
      "epoch": 2.8265184597062327,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004208683465648358,
      "loss": 4.9915,
      "mean_token_accuracy": 0.1877106249332428,
      "num_tokens": 62387077.0,
      "step": 28480
    },
    {
      "entropy": 5.686506223678589,
      "epoch": 2.8270146883684,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042084136786038737,
      "loss": 5.0262,
      "mean_token_accuracy": 0.18572545498609544,
      "num_tokens": 62397320.0,
      "step": 28485
    },
    {
      "entropy": 5.646945858001709,
      "epoch": 2.8275109170305677,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004208143855393872,
      "loss": 4.9729,
      "mean_token_accuracy": 0.19180663526058198,
      "num_tokens": 62407675.0,
      "step": 28490
    },
    {
      "entropy": 5.628590059280396,
      "epoch": 2.828007145692735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042078739960250446,
      "loss": 4.9739,
      "mean_token_accuracy": 0.19331900626420975,
      "num_tokens": 62417969.0,
      "step": 28495
    },
    {
      "entropy": 5.675786828994751,
      "epoch": 2.8285033743549026,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042076041005040835,
      "loss": 4.9714,
      "mean_token_accuracy": 0.18950010538101197,
      "num_tokens": 62428761.0,
      "step": 28500
    },
    {
      "entropy": 5.642805480957032,
      "epoch": 2.8289996030170705,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004207334168837682,
      "loss": 4.9645,
      "mean_token_accuracy": 0.17969236224889756,
      "num_tokens": 62438425.0,
      "step": 28505
    },
    {
      "entropy": 5.612138557434082,
      "epoch": 2.829495831679238,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004207064201032533,
      "loss": 4.9385,
      "mean_token_accuracy": 0.18957892507314683,
      "num_tokens": 62448819.0,
      "step": 28510
    },
    {
      "entropy": 5.692314195632934,
      "epoch": 2.8299920603414055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042067941970953316,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1819077640771866,
      "num_tokens": 62458110.0,
      "step": 28515
    },
    {
      "entropy": 5.645751094818115,
      "epoch": 2.830488289003573,
      "grad_norm": 1.125,
      "learning_rate": 0.00042065241570327734,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18571601659059525,
      "num_tokens": 62468158.0,
      "step": 28520
    },
    {
      "entropy": 5.666433525085449,
      "epoch": 2.8309845176657404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004206254080851555,
      "loss": 5.1141,
      "mean_token_accuracy": 0.18016449809074403,
      "num_tokens": 62479926.0,
      "step": 28525
    },
    {
      "entropy": 5.714016008377075,
      "epoch": 2.831480746327908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042059839685583726,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1826118364930153,
      "num_tokens": 62490012.0,
      "step": 28530
    },
    {
      "entropy": 5.654635906219482,
      "epoch": 2.8319769749900754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042057138201599247,
      "loss": 5.0049,
      "mean_token_accuracy": 0.1892981767654419,
      "num_tokens": 62501268.0,
      "step": 28535
    },
    {
      "entropy": 5.646989917755127,
      "epoch": 2.832473203652243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042054436356629113,
      "loss": 5.0154,
      "mean_token_accuracy": 0.1844584360718727,
      "num_tokens": 62513117.0,
      "step": 28540
    },
    {
      "entropy": 5.730324935913086,
      "epoch": 2.8329694323144103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004205173415074032,
      "loss": 5.1347,
      "mean_token_accuracy": 0.17821938395500184,
      "num_tokens": 62524673.0,
      "step": 28545
    },
    {
      "entropy": 5.646428394317627,
      "epoch": 2.833465660976578,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004204903158399989,
      "loss": 4.971,
      "mean_token_accuracy": 0.1874020963907242,
      "num_tokens": 62536336.0,
      "step": 28550
    },
    {
      "entropy": 5.7503026008605955,
      "epoch": 2.8339618896387453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004204632865647481,
      "loss": 5.1508,
      "mean_token_accuracy": 0.17748499661684036,
      "num_tokens": 62546976.0,
      "step": 28555
    },
    {
      "entropy": 5.657270717620849,
      "epoch": 2.834458118300913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204362536823213,
      "loss": 4.9315,
      "mean_token_accuracy": 0.19330254048109055,
      "num_tokens": 62558081.0,
      "step": 28560
    },
    {
      "entropy": 5.671589326858521,
      "epoch": 2.8349543469630807,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042040921719338877,
      "loss": 4.9566,
      "mean_token_accuracy": 0.18693988919258117,
      "num_tokens": 62568909.0,
      "step": 28565
    },
    {
      "entropy": 5.644330978393555,
      "epoch": 2.835450575625248,
      "grad_norm": 1.078125,
      "learning_rate": 0.000420382177098621,
      "loss": 4.9644,
      "mean_token_accuracy": 0.19049750566482543,
      "num_tokens": 62579765.0,
      "step": 28570
    },
    {
      "entropy": 5.660172462463379,
      "epoch": 2.8359468042874156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042035513339868855,
      "loss": 5.0155,
      "mean_token_accuracy": 0.1813484251499176,
      "num_tokens": 62591150.0,
      "step": 28575
    },
    {
      "entropy": 5.843495941162109,
      "epoch": 2.836443032949583,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004203280860942619,
      "loss": 5.1366,
      "mean_token_accuracy": 0.1762630820274353,
      "num_tokens": 62601582.0,
      "step": 28580
    },
    {
      "entropy": 5.70407772064209,
      "epoch": 2.8369392616117506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004203010351860119,
      "loss": 4.9305,
      "mean_token_accuracy": 0.18966541886329652,
      "num_tokens": 62612381.0,
      "step": 28585
    },
    {
      "entropy": 5.686920309066773,
      "epoch": 2.837435490273918,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004202739806746094,
      "loss": 5.0621,
      "mean_token_accuracy": 0.1793050542473793,
      "num_tokens": 62623302.0,
      "step": 28590
    },
    {
      "entropy": 5.677908086776734,
      "epoch": 2.837931718936086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004202469225607252,
      "loss": 5.0243,
      "mean_token_accuracy": 0.17442430257797242,
      "num_tokens": 62633850.0,
      "step": 28595
    },
    {
      "entropy": 5.739142847061157,
      "epoch": 2.8384279475982535,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004202198608450302,
      "loss": 5.0478,
      "mean_token_accuracy": 0.17944031953811646,
      "num_tokens": 62644869.0,
      "step": 28600
    },
    {
      "entropy": 5.6770987033844,
      "epoch": 2.838924176260421,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004201927955281957,
      "loss": 5.0363,
      "mean_token_accuracy": 0.18179007470607758,
      "num_tokens": 62655553.0,
      "step": 28605
    },
    {
      "entropy": 5.677303647994995,
      "epoch": 2.8394204049225884,
      "grad_norm": 1.125,
      "learning_rate": 0.0004201657266108926,
      "loss": 4.9931,
      "mean_token_accuracy": 0.18801648318767547,
      "num_tokens": 62667043.0,
      "step": 28610
    },
    {
      "entropy": 5.6896837711334225,
      "epoch": 2.839916633584756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042013865409379244,
      "loss": 5.0589,
      "mean_token_accuracy": 0.18356267511844634,
      "num_tokens": 62678081.0,
      "step": 28615
    },
    {
      "entropy": 5.6973865032196045,
      "epoch": 2.8404128622469234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004201115779775663,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18464255779981614,
      "num_tokens": 62689101.0,
      "step": 28620
    },
    {
      "entropy": 5.654908514022827,
      "epoch": 2.840909090909091,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004200844982628857,
      "loss": 4.9736,
      "mean_token_accuracy": 0.19054378867149352,
      "num_tokens": 62698584.0,
      "step": 28625
    },
    {
      "entropy": 5.639261531829834,
      "epoch": 2.8414053195712583,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004200574149504222,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18390269130468367,
      "num_tokens": 62709936.0,
      "step": 28630
    },
    {
      "entropy": 5.669144868850708,
      "epoch": 2.841901548233426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004200303280408474,
      "loss": 5.0065,
      "mean_token_accuracy": 0.18763557970523834,
      "num_tokens": 62720926.0,
      "step": 28635
    },
    {
      "entropy": 5.6814470291137695,
      "epoch": 2.8423977768955933,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042000323753483296,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18607867807149886,
      "num_tokens": 62732221.0,
      "step": 28640
    },
    {
      "entropy": 5.669524669647217,
      "epoch": 2.842894005557761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041997614343305067,
      "loss": 5.0229,
      "mean_token_accuracy": 0.18321057111024858,
      "num_tokens": 62743579.0,
      "step": 28645
    },
    {
      "entropy": 5.700817728042603,
      "epoch": 2.8433902342199286,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041994904573617247,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18334050625562667,
      "num_tokens": 62754246.0,
      "step": 28650
    },
    {
      "entropy": 5.704958724975586,
      "epoch": 2.843886462882096,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041992194444487026,
      "loss": 5.1364,
      "mean_token_accuracy": 0.18020118325948714,
      "num_tokens": 62765813.0,
      "step": 28655
    },
    {
      "entropy": 5.710921716690064,
      "epoch": 2.8443826915442636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041989483955981607,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17978915125131606,
      "num_tokens": 62777208.0,
      "step": 28660
    },
    {
      "entropy": 5.688695859909058,
      "epoch": 2.844878920206431,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041986773108168216,
      "loss": 4.9976,
      "mean_token_accuracy": 0.17804140150547026,
      "num_tokens": 62787549.0,
      "step": 28665
    },
    {
      "entropy": 5.67550311088562,
      "epoch": 2.8453751488685985,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041984061901114057,
      "loss": 4.9621,
      "mean_token_accuracy": 0.19399717599153518,
      "num_tokens": 62798609.0,
      "step": 28670
    },
    {
      "entropy": 5.700813150405883,
      "epoch": 2.845871377530766,
      "grad_norm": 1.125,
      "learning_rate": 0.0004198135033488638,
      "loss": 4.9687,
      "mean_token_accuracy": 0.18532646000385283,
      "num_tokens": 62809705.0,
      "step": 28675
    },
    {
      "entropy": 5.738010311126709,
      "epoch": 2.846367606192934,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041978638409552417,
      "loss": 5.042,
      "mean_token_accuracy": 0.1845290720462799,
      "num_tokens": 62820740.0,
      "step": 28680
    },
    {
      "entropy": 5.695954132080078,
      "epoch": 2.8468638348551014,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004197592612517942,
      "loss": 5.1063,
      "mean_token_accuracy": 0.18355564326047896,
      "num_tokens": 62832294.0,
      "step": 28685
    },
    {
      "entropy": 5.583215999603271,
      "epoch": 2.847360063517269,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004197321348183465,
      "loss": 4.9298,
      "mean_token_accuracy": 0.18695326149463654,
      "num_tokens": 62843174.0,
      "step": 28690
    },
    {
      "entropy": 5.6823036670684814,
      "epoch": 2.8478562921794364,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041970500479585364,
      "loss": 5.0421,
      "mean_token_accuracy": 0.1833550900220871,
      "num_tokens": 62852983.0,
      "step": 28695
    },
    {
      "entropy": 5.630422639846802,
      "epoch": 2.848352520841604,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041967787118498854,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18881481289863586,
      "num_tokens": 62863587.0,
      "step": 28700
    },
    {
      "entropy": 5.701425266265869,
      "epoch": 2.8488487495037713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004196507339864241,
      "loss": 5.0828,
      "mean_token_accuracy": 0.180396568775177,
      "num_tokens": 62874938.0,
      "step": 28705
    },
    {
      "entropy": 5.656632041931152,
      "epoch": 2.849344978165939,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000419623593200833,
      "loss": 4.96,
      "mean_token_accuracy": 0.1860442653298378,
      "num_tokens": 62885987.0,
      "step": 28710
    },
    {
      "entropy": 5.611132049560547,
      "epoch": 2.8498412068281063,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041959644882888844,
      "loss": 4.9349,
      "mean_token_accuracy": 0.1855830043554306,
      "num_tokens": 62896827.0,
      "step": 28715
    },
    {
      "entropy": 5.705468320846558,
      "epoch": 2.8503374354902737,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041956930087126357,
      "loss": 5.0231,
      "mean_token_accuracy": 0.18162171840667723,
      "num_tokens": 62906596.0,
      "step": 28720
    },
    {
      "entropy": 5.620092678070068,
      "epoch": 2.850833664152441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041954214932863156,
      "loss": 4.9943,
      "mean_token_accuracy": 0.184586501121521,
      "num_tokens": 62918773.0,
      "step": 28725
    },
    {
      "entropy": 5.692898559570312,
      "epoch": 2.851329892814609,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004195149942016657,
      "loss": 4.9902,
      "mean_token_accuracy": 0.18341905921697615,
      "num_tokens": 62929395.0,
      "step": 28730
    },
    {
      "entropy": 5.671969985961914,
      "epoch": 2.8518261214767766,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004194878354910394,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18853245675563812,
      "num_tokens": 62939905.0,
      "step": 28735
    },
    {
      "entropy": 5.696950244903564,
      "epoch": 2.852322350138944,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041946067319742605,
      "loss": 5.0019,
      "mean_token_accuracy": 0.19305487275123595,
      "num_tokens": 62950684.0,
      "step": 28740
    },
    {
      "entropy": 5.634122371673584,
      "epoch": 2.8528185788011116,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041943350732149935,
      "loss": 4.9644,
      "mean_token_accuracy": 0.1873519390821457,
      "num_tokens": 62960754.0,
      "step": 28745
    },
    {
      "entropy": 5.635667181015014,
      "epoch": 2.853314807463279,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041940633786393295,
      "loss": 4.9561,
      "mean_token_accuracy": 0.1840004727244377,
      "num_tokens": 62970659.0,
      "step": 28750
    },
    {
      "entropy": 5.6372884750366214,
      "epoch": 2.8538110361254465,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041937916482540045,
      "loss": 4.9424,
      "mean_token_accuracy": 0.18876208662986754,
      "num_tokens": 62981135.0,
      "step": 28755
    },
    {
      "entropy": 5.686242198944091,
      "epoch": 2.854307264787614,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041935198820657575,
      "loss": 4.993,
      "mean_token_accuracy": 0.18611033856868744,
      "num_tokens": 62991979.0,
      "step": 28760
    },
    {
      "entropy": 5.706282138824463,
      "epoch": 2.854803493449782,
      "grad_norm": 1.125,
      "learning_rate": 0.0004193248080081329,
      "loss": 5.0488,
      "mean_token_accuracy": 0.18621266782283782,
      "num_tokens": 63003180.0,
      "step": 28765
    },
    {
      "entropy": 5.683365058898926,
      "epoch": 2.8552997221119494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004192976242307457,
      "loss": 5.0439,
      "mean_token_accuracy": 0.1873775839805603,
      "num_tokens": 63014529.0,
      "step": 28770
    },
    {
      "entropy": 5.649776840209961,
      "epoch": 2.855795950774117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004192704368750884,
      "loss": 5.0254,
      "mean_token_accuracy": 0.18233649730682372,
      "num_tokens": 63025293.0,
      "step": 28775
    },
    {
      "entropy": 5.69970669746399,
      "epoch": 2.8562921794362843,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004192432459418351,
      "loss": 5.0365,
      "mean_token_accuracy": 0.18548420667648316,
      "num_tokens": 63036210.0,
      "step": 28780
    },
    {
      "entropy": 5.728966808319091,
      "epoch": 2.856788408098452,
      "grad_norm": 1.0,
      "learning_rate": 0.0004192160514316602,
      "loss": 5.0384,
      "mean_token_accuracy": 0.18292864710092543,
      "num_tokens": 63047212.0,
      "step": 28785
    },
    {
      "entropy": 5.697925090789795,
      "epoch": 2.8572846367606193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004191888533452379,
      "loss": 5.0079,
      "mean_token_accuracy": 0.1849217399954796,
      "num_tokens": 63058398.0,
      "step": 28790
    },
    {
      "entropy": 5.7046221733093265,
      "epoch": 2.8577808654227868,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004191616516832428,
      "loss": 5.0437,
      "mean_token_accuracy": 0.1799236997961998,
      "num_tokens": 63070099.0,
      "step": 28795
    },
    {
      "entropy": 5.655193948745728,
      "epoch": 2.8582770940849542,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004191344464463494,
      "loss": 4.9695,
      "mean_token_accuracy": 0.19006676524877547,
      "num_tokens": 63081743.0,
      "step": 28800
    },
    {
      "entropy": 5.705867052078247,
      "epoch": 2.8587733227471217,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041910723763523224,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18126917332410813,
      "num_tokens": 63092548.0,
      "step": 28805
    },
    {
      "entropy": 5.699216079711914,
      "epoch": 2.859269551409289,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004190800252505661,
      "loss": 5.0488,
      "mean_token_accuracy": 0.18702365159988404,
      "num_tokens": 63102765.0,
      "step": 28810
    },
    {
      "entropy": 5.615594863891602,
      "epoch": 2.8597657800714567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004190528092930259,
      "loss": 4.937,
      "mean_token_accuracy": 0.1919094815850258,
      "num_tokens": 63113847.0,
      "step": 28815
    },
    {
      "entropy": 5.6603984355926515,
      "epoch": 2.8602620087336246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004190255897632863,
      "loss": 5.0183,
      "mean_token_accuracy": 0.18103066980838775,
      "num_tokens": 63125288.0,
      "step": 28820
    },
    {
      "entropy": 5.71246829032898,
      "epoch": 2.860758237395792,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041899836666202254,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18262096047401427,
      "num_tokens": 63136590.0,
      "step": 28825
    },
    {
      "entropy": 5.714640331268311,
      "epoch": 2.8612544660579595,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004189711399899095,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17822401374578475,
      "num_tokens": 63146711.0,
      "step": 28830
    },
    {
      "entropy": 5.7055888175964355,
      "epoch": 2.861750694720127,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004189439097476224,
      "loss": 5.0578,
      "mean_token_accuracy": 0.18103123754262923,
      "num_tokens": 63157805.0,
      "step": 28835
    },
    {
      "entropy": 5.684500980377197,
      "epoch": 2.8622469233822945,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004189166759358366,
      "loss": 4.9943,
      "mean_token_accuracy": 0.19726363122463225,
      "num_tokens": 63167891.0,
      "step": 28840
    },
    {
      "entropy": 5.662652635574341,
      "epoch": 2.862743152044462,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041888943855522727,
      "loss": 4.9896,
      "mean_token_accuracy": 0.18864350914955139,
      "num_tokens": 63179438.0,
      "step": 28845
    },
    {
      "entropy": 5.662731122970581,
      "epoch": 2.86323938070663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041886219760646997,
      "loss": 5.0524,
      "mean_token_accuracy": 0.180850987136364,
      "num_tokens": 63190234.0,
      "step": 28850
    },
    {
      "entropy": 5.666010046005249,
      "epoch": 2.8637356093687973,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004188349530902402,
      "loss": 4.9858,
      "mean_token_accuracy": 0.18109796047210694,
      "num_tokens": 63201880.0,
      "step": 28855
    },
    {
      "entropy": 5.6292376041412355,
      "epoch": 2.864231838030965,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041880770500721346,
      "loss": 4.8846,
      "mean_token_accuracy": 0.19670591205358506,
      "num_tokens": 63211906.0,
      "step": 28860
    },
    {
      "entropy": 5.689622068405152,
      "epoch": 2.8647280666931323,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004187804533580655,
      "loss": 5.0789,
      "mean_token_accuracy": 0.18573503345251083,
      "num_tokens": 63223365.0,
      "step": 28865
    },
    {
      "entropy": 5.723799228668213,
      "epoch": 2.8652242953552998,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004187531981434721,
      "loss": 5.1066,
      "mean_token_accuracy": 0.17631808221340178,
      "num_tokens": 63233849.0,
      "step": 28870
    },
    {
      "entropy": 5.701140213012695,
      "epoch": 2.8657205240174672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041872593936410923,
      "loss": 5.0749,
      "mean_token_accuracy": 0.181366029381752,
      "num_tokens": 63245742.0,
      "step": 28875
    },
    {
      "entropy": 5.67630558013916,
      "epoch": 2.8662167526796347,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041869867702065266,
      "loss": 5.0417,
      "mean_token_accuracy": 0.18718888908624648,
      "num_tokens": 63257284.0,
      "step": 28880
    },
    {
      "entropy": 5.680598974227905,
      "epoch": 2.866712981341802,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004186714111137787,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17784785628318786,
      "num_tokens": 63268893.0,
      "step": 28885
    },
    {
      "entropy": 5.74768762588501,
      "epoch": 2.8672092100039697,
      "grad_norm": 1.0,
      "learning_rate": 0.0004186441416441632,
      "loss": 5.0787,
      "mean_token_accuracy": 0.18173172622919082,
      "num_tokens": 63281763.0,
      "step": 28890
    },
    {
      "entropy": 5.703610181808472,
      "epoch": 2.867705438666137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004186168686124825,
      "loss": 5.0377,
      "mean_token_accuracy": 0.18338662534952163,
      "num_tokens": 63292304.0,
      "step": 28895
    },
    {
      "entropy": 5.7471226215362545,
      "epoch": 2.8682016673283046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041858959201941307,
      "loss": 5.1227,
      "mean_token_accuracy": 0.1812385931611061,
      "num_tokens": 63303539.0,
      "step": 28900
    },
    {
      "entropy": 5.741182613372803,
      "epoch": 2.8686978959904725,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041856231186563107,
      "loss": 5.0775,
      "mean_token_accuracy": 0.180435448884964,
      "num_tokens": 63313678.0,
      "step": 28905
    },
    {
      "entropy": 5.68811182975769,
      "epoch": 2.86919412465264,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041853502815181306,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18144164085388184,
      "num_tokens": 63324988.0,
      "step": 28910
    },
    {
      "entropy": 5.763414287567139,
      "epoch": 2.8696903533148075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004185077408786357,
      "loss": 5.124,
      "mean_token_accuracy": 0.1798551008105278,
      "num_tokens": 63335325.0,
      "step": 28915
    },
    {
      "entropy": 5.687190914154053,
      "epoch": 2.870186581976975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004184804500467756,
      "loss": 5.0552,
      "mean_token_accuracy": 0.17829488962888718,
      "num_tokens": 63346838.0,
      "step": 28920
    },
    {
      "entropy": 5.696912956237793,
      "epoch": 2.8706828106391424,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041845315565690945,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1837082475423813,
      "num_tokens": 63356752.0,
      "step": 28925
    },
    {
      "entropy": 5.7445573806762695,
      "epoch": 2.87117903930131,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041842585770971417,
      "loss": 5.0732,
      "mean_token_accuracy": 0.18234584480524063,
      "num_tokens": 63367464.0,
      "step": 28930
    },
    {
      "entropy": 5.641614532470703,
      "epoch": 2.871675267963478,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004183985562058667,
      "loss": 5.0242,
      "mean_token_accuracy": 0.18561359792947768,
      "num_tokens": 63377986.0,
      "step": 28935
    },
    {
      "entropy": 5.704216051101684,
      "epoch": 2.8721714966256453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000418371251146044,
      "loss": 5.0481,
      "mean_token_accuracy": 0.18212876170873643,
      "num_tokens": 63391026.0,
      "step": 28940
    },
    {
      "entropy": 5.694080448150634,
      "epoch": 2.872667725287813,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004183439425309232,
      "loss": 5.0703,
      "mean_token_accuracy": 0.18046136647462846,
      "num_tokens": 63401946.0,
      "step": 28945
    },
    {
      "entropy": 5.640472650527954,
      "epoch": 2.8731639539499803,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004183166303611816,
      "loss": 4.9772,
      "mean_token_accuracy": 0.18306902199983596,
      "num_tokens": 63412800.0,
      "step": 28950
    },
    {
      "entropy": 5.643978595733643,
      "epoch": 2.8736601826121477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004182893146374963,
      "loss": 4.9422,
      "mean_token_accuracy": 0.18879874497652055,
      "num_tokens": 63423153.0,
      "step": 28955
    },
    {
      "entropy": 5.624293613433838,
      "epoch": 2.874156411274315,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041826199536054476,
      "loss": 4.9169,
      "mean_token_accuracy": 0.19262148737907409,
      "num_tokens": 63433115.0,
      "step": 28960
    },
    {
      "entropy": 5.688516664505005,
      "epoch": 2.8746526399364827,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004182346725310045,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17991043031215667,
      "num_tokens": 63443215.0,
      "step": 28965
    },
    {
      "entropy": 5.687531042098999,
      "epoch": 2.87514886859865,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041820734614955295,
      "loss": 4.9751,
      "mean_token_accuracy": 0.18309463262557985,
      "num_tokens": 63453619.0,
      "step": 28970
    },
    {
      "entropy": 5.6914567947387695,
      "epoch": 2.8756450972608176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004181800162168678,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1848748207092285,
      "num_tokens": 63464284.0,
      "step": 28975
    },
    {
      "entropy": 5.645015811920166,
      "epoch": 2.876141325922985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004181526827336268,
      "loss": 4.9237,
      "mean_token_accuracy": 0.19221092909574508,
      "num_tokens": 63474318.0,
      "step": 28980
    },
    {
      "entropy": 5.64661660194397,
      "epoch": 2.8766375545851526,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004181253457005077,
      "loss": 4.9838,
      "mean_token_accuracy": 0.19898854494094848,
      "num_tokens": 63484179.0,
      "step": 28985
    },
    {
      "entropy": 5.742045307159424,
      "epoch": 2.8771337832473205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041809800511818853,
      "loss": 5.0487,
      "mean_token_accuracy": 0.18486587107181549,
      "num_tokens": 63494876.0,
      "step": 28990
    },
    {
      "entropy": 5.79269061088562,
      "epoch": 2.877630011909488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041807066098734704,
      "loss": 5.159,
      "mean_token_accuracy": 0.18276361078023912,
      "num_tokens": 63505563.0,
      "step": 28995
    },
    {
      "entropy": 5.705702257156372,
      "epoch": 2.8781262405716554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004180433133086615,
      "loss": 5.1241,
      "mean_token_accuracy": 0.17459706217050552,
      "num_tokens": 63517493.0,
      "step": 29000
    },
    {
      "entropy": 5.74765453338623,
      "epoch": 2.878622469233823,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004180159620828101,
      "loss": 5.0805,
      "mean_token_accuracy": 0.1900472015142441,
      "num_tokens": 63528187.0,
      "step": 29005
    },
    {
      "entropy": 5.744244241714478,
      "epoch": 2.8791186978959904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004179886073104709,
      "loss": 5.1477,
      "mean_token_accuracy": 0.18182885348796846,
      "num_tokens": 63539297.0,
      "step": 29010
    },
    {
      "entropy": 5.680358457565307,
      "epoch": 2.879614926558158,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004179612489923224,
      "loss": 5.066,
      "mean_token_accuracy": 0.18196721524000167,
      "num_tokens": 63551162.0,
      "step": 29015
    },
    {
      "entropy": 5.676433515548706,
      "epoch": 2.880111155220326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417933887129043,
      "loss": 5.0411,
      "mean_token_accuracy": 0.17896981537342072,
      "num_tokens": 63561472.0,
      "step": 29020
    },
    {
      "entropy": 5.735057926177978,
      "epoch": 2.8806073838824933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004179065217213111,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17919723838567733,
      "num_tokens": 63571682.0,
      "step": 29025
    },
    {
      "entropy": 5.706338405609131,
      "epoch": 2.8811036125446607,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041787915276980556,
      "loss": 5.0455,
      "mean_token_accuracy": 0.1825241580605507,
      "num_tokens": 63583288.0,
      "step": 29030
    },
    {
      "entropy": 5.689586210250854,
      "epoch": 2.881599841206828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041785178027520486,
      "loss": 5.0422,
      "mean_token_accuracy": 0.18079305291175843,
      "num_tokens": 63594774.0,
      "step": 29035
    },
    {
      "entropy": 5.696850728988648,
      "epoch": 2.8820960698689957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041782440423818775,
      "loss": 5.06,
      "mean_token_accuracy": 0.18304894119501114,
      "num_tokens": 63605156.0,
      "step": 29040
    },
    {
      "entropy": 5.693200635910034,
      "epoch": 2.882592298531163,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004177970246594332,
      "loss": 4.9594,
      "mean_token_accuracy": 0.18463135808706282,
      "num_tokens": 63615224.0,
      "step": 29045
    },
    {
      "entropy": 5.68426365852356,
      "epoch": 2.8830885271933306,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041776964153962017,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18319242000579833,
      "num_tokens": 63625566.0,
      "step": 29050
    },
    {
      "entropy": 5.749634885787964,
      "epoch": 2.883584755855498,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041774225487942765,
      "loss": 5.1239,
      "mean_token_accuracy": 0.18009073436260223,
      "num_tokens": 63635697.0,
      "step": 29055
    },
    {
      "entropy": 5.677986574172974,
      "epoch": 2.8840809845176656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004177148646795348,
      "loss": 5.0186,
      "mean_token_accuracy": 0.18748359829187394,
      "num_tokens": 63646377.0,
      "step": 29060
    },
    {
      "entropy": 5.710941934585572,
      "epoch": 2.884577213179833,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041768747094062084,
      "loss": 4.9659,
      "mean_token_accuracy": 0.18338521420955659,
      "num_tokens": 63657356.0,
      "step": 29065
    },
    {
      "entropy": 5.782914829254151,
      "epoch": 2.8850734418420005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041766007366336494,
      "loss": 5.0836,
      "mean_token_accuracy": 0.1889672502875328,
      "num_tokens": 63667267.0,
      "step": 29070
    },
    {
      "entropy": 5.638280916213989,
      "epoch": 2.8855696705041685,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041763267284844675,
      "loss": 5.0203,
      "mean_token_accuracy": 0.1843297392129898,
      "num_tokens": 63678242.0,
      "step": 29075
    },
    {
      "entropy": 5.7053803443908695,
      "epoch": 2.886065899166336,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004176052684965456,
      "loss": 5.02,
      "mean_token_accuracy": 0.18354716151952744,
      "num_tokens": 63689757.0,
      "step": 29080
    },
    {
      "entropy": 5.649961805343628,
      "epoch": 2.8865621278285034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004175778606083411,
      "loss": 4.9596,
      "mean_token_accuracy": 0.18754420578479766,
      "num_tokens": 63701196.0,
      "step": 29085
    },
    {
      "entropy": 5.6583668231964115,
      "epoch": 2.887058356490671,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004175504491845128,
      "loss": 5.0756,
      "mean_token_accuracy": 0.1847450852394104,
      "num_tokens": 63712802.0,
      "step": 29090
    },
    {
      "entropy": 5.697234630584717,
      "epoch": 2.8875545851528384,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041752303422574067,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18503907918930054,
      "num_tokens": 63723543.0,
      "step": 29095
    },
    {
      "entropy": 5.686252117156982,
      "epoch": 2.888050813815006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004174956157327043,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18003454357385634,
      "num_tokens": 63734954.0,
      "step": 29100
    },
    {
      "entropy": 5.743284368515015,
      "epoch": 2.8885470424771738,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004174681937060837,
      "loss": 5.0674,
      "mean_token_accuracy": 0.18115601390600206,
      "num_tokens": 63746991.0,
      "step": 29105
    },
    {
      "entropy": 5.655759668350219,
      "epoch": 2.8890432711393412,
      "grad_norm": 1.125,
      "learning_rate": 0.00041744076814655893,
      "loss": 4.9108,
      "mean_token_accuracy": 0.19251292198896408,
      "num_tokens": 63756047.0,
      "step": 29110
    },
    {
      "entropy": 5.690530776977539,
      "epoch": 2.8895394998015087,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041741333905481,
      "loss": 5.0514,
      "mean_token_accuracy": 0.18120466023683549,
      "num_tokens": 63768006.0,
      "step": 29115
    },
    {
      "entropy": 5.694486141204834,
      "epoch": 2.890035728463676,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004173859064315172,
      "loss": 5.0213,
      "mean_token_accuracy": 0.17908615469932557,
      "num_tokens": 63778765.0,
      "step": 29120
    },
    {
      "entropy": 5.69671778678894,
      "epoch": 2.8905319571258437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041735847027736067,
      "loss": 5.0136,
      "mean_token_accuracy": 0.19097760319709778,
      "num_tokens": 63791305.0,
      "step": 29125
    },
    {
      "entropy": 5.6473822593688965,
      "epoch": 2.891028185788011,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004173310305930209,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18920005410909652,
      "num_tokens": 63803378.0,
      "step": 29130
    },
    {
      "entropy": 5.666828775405884,
      "epoch": 2.8915244144501786,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041730358737917823,
      "loss": 4.9867,
      "mean_token_accuracy": 0.19120000153779984,
      "num_tokens": 63815621.0,
      "step": 29135
    },
    {
      "entropy": 5.687378358840943,
      "epoch": 2.892020643112346,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041727614063651313,
      "loss": 5.0503,
      "mean_token_accuracy": 0.17901761680841446,
      "num_tokens": 63828009.0,
      "step": 29140
    },
    {
      "entropy": 5.643346166610717,
      "epoch": 2.8925168717745136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004172486903657065,
      "loss": 5.0047,
      "mean_token_accuracy": 0.18655674606561662,
      "num_tokens": 63840506.0,
      "step": 29145
    },
    {
      "entropy": 5.6852678775787355,
      "epoch": 2.893013100436681,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004172212365674388,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18775244504213334,
      "num_tokens": 63851894.0,
      "step": 29150
    },
    {
      "entropy": 5.74419903755188,
      "epoch": 2.8935093290988485,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041719377924239083,
      "loss": 5.1232,
      "mean_token_accuracy": 0.17339274436235427,
      "num_tokens": 63863815.0,
      "step": 29155
    },
    {
      "entropy": 5.718755769729614,
      "epoch": 2.8940055577610164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041716631839124356,
      "loss": 5.0563,
      "mean_token_accuracy": 0.18173293322324752,
      "num_tokens": 63875115.0,
      "step": 29160
    },
    {
      "entropy": 5.725619697570801,
      "epoch": 2.894501786423184,
      "grad_norm": 1.203125,
      "learning_rate": 0.000417138854014678,
      "loss": 5.0601,
      "mean_token_accuracy": 0.18389911502599715,
      "num_tokens": 63885652.0,
      "step": 29165
    },
    {
      "entropy": 5.651135587692261,
      "epoch": 2.8949980150853514,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004171113861133751,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18616945147514344,
      "num_tokens": 63896861.0,
      "step": 29170
    },
    {
      "entropy": 5.666819620132446,
      "epoch": 2.895494243747519,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004170839146880161,
      "loss": 5.0775,
      "mean_token_accuracy": 0.1804726541042328,
      "num_tokens": 63908736.0,
      "step": 29175
    },
    {
      "entropy": 5.6728309154510494,
      "epoch": 2.8959904724096863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004170564397392821,
      "loss": 5.04,
      "mean_token_accuracy": 0.18473106175661086,
      "num_tokens": 63920841.0,
      "step": 29180
    },
    {
      "entropy": 5.727948760986328,
      "epoch": 2.896486701071854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004170289612678545,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1811899274587631,
      "num_tokens": 63931421.0,
      "step": 29185
    },
    {
      "entropy": 5.684969425201416,
      "epoch": 2.8969829297340213,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041700147927441477,
      "loss": 5.0285,
      "mean_token_accuracy": 0.18739347904920578,
      "num_tokens": 63941548.0,
      "step": 29190
    },
    {
      "entropy": 5.699454641342163,
      "epoch": 2.897479158396189,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041697399375964435,
      "loss": 5.0521,
      "mean_token_accuracy": 0.18205295205116273,
      "num_tokens": 63953492.0,
      "step": 29195
    },
    {
      "entropy": 5.676342296600342,
      "epoch": 2.8979753870583567,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041694650472422467,
      "loss": 5.0369,
      "mean_token_accuracy": 0.1841968074440956,
      "num_tokens": 63965601.0,
      "step": 29200
    },
    {
      "entropy": 5.722444915771485,
      "epoch": 2.898471615720524,
      "grad_norm": 1.125,
      "learning_rate": 0.00041691901216883763,
      "loss": 5.0598,
      "mean_token_accuracy": 0.1841201588511467,
      "num_tokens": 63975857.0,
      "step": 29205
    },
    {
      "entropy": 5.736050748825074,
      "epoch": 2.8989678443826916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000416891516094165,
      "loss": 5.1017,
      "mean_token_accuracy": 0.17704980522394181,
      "num_tokens": 63987504.0,
      "step": 29210
    },
    {
      "entropy": 5.651437044143677,
      "epoch": 2.899464073044859,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004168640165008884,
      "loss": 4.9853,
      "mean_token_accuracy": 0.18639895617961882,
      "num_tokens": 63998362.0,
      "step": 29215
    },
    {
      "entropy": 5.687715625762939,
      "epoch": 2.8999603017070266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041683651338968994,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18364009410142898,
      "num_tokens": 64008523.0,
      "step": 29220
    },
    {
      "entropy": 5.631632137298584,
      "epoch": 2.900456530369194,
      "grad_norm": 1.125,
      "learning_rate": 0.00041680900676125147,
      "loss": 4.9789,
      "mean_token_accuracy": 0.18609965592622757,
      "num_tokens": 64019310.0,
      "step": 29225
    },
    {
      "entropy": 5.768888235092163,
      "epoch": 2.9009527590313615,
      "grad_norm": 1.125,
      "learning_rate": 0.00041678149661625525,
      "loss": 5.1158,
      "mean_token_accuracy": 0.18126425594091417,
      "num_tokens": 64030625.0,
      "step": 29230
    },
    {
      "entropy": 5.610435056686401,
      "epoch": 2.901448987693529,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004167539829553835,
      "loss": 4.9019,
      "mean_token_accuracy": 0.19302420765161515,
      "num_tokens": 64041947.0,
      "step": 29235
    },
    {
      "entropy": 5.621342611312866,
      "epoch": 2.9019452163556965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041672646577931835,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18264070600271226,
      "num_tokens": 64052709.0,
      "step": 29240
    },
    {
      "entropy": 5.595081949234009,
      "epoch": 2.9024414450178644,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041669894508874225,
      "loss": 4.9628,
      "mean_token_accuracy": 0.19245805889368056,
      "num_tokens": 64064284.0,
      "step": 29245
    },
    {
      "entropy": 5.666966390609741,
      "epoch": 2.902937673680032,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041667142088433766,
      "loss": 5.036,
      "mean_token_accuracy": 0.18994693756103515,
      "num_tokens": 64074086.0,
      "step": 29250
    },
    {
      "entropy": 5.605280017852783,
      "epoch": 2.9034339023421993,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041664389316678713,
      "loss": 4.9675,
      "mean_token_accuracy": 0.18711818307638167,
      "num_tokens": 64085958.0,
      "step": 29255
    },
    {
      "entropy": 5.633073949813843,
      "epoch": 2.903930131004367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004166163619367733,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18131861388683318,
      "num_tokens": 64098027.0,
      "step": 29260
    },
    {
      "entropy": 5.797200870513916,
      "epoch": 2.9044263596665343,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041658882719497876,
      "loss": 5.1251,
      "mean_token_accuracy": 0.18422953337430953,
      "num_tokens": 64109452.0,
      "step": 29265
    },
    {
      "entropy": 5.76592607498169,
      "epoch": 2.9049225883287018,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004165612889420864,
      "loss": 5.0255,
      "mean_token_accuracy": 0.19148880541324614,
      "num_tokens": 64120803.0,
      "step": 29270
    },
    {
      "entropy": 5.704599285125733,
      "epoch": 2.9054188169908692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004165337471787792,
      "loss": 5.0474,
      "mean_token_accuracy": 0.1857011467218399,
      "num_tokens": 64132594.0,
      "step": 29275
    },
    {
      "entropy": 5.665228080749512,
      "epoch": 2.905915045653037,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041650620190574004,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18022168576717376,
      "num_tokens": 64143039.0,
      "step": 29280
    },
    {
      "entropy": 5.672952938079834,
      "epoch": 2.9064112743152046,
      "grad_norm": 1.25,
      "learning_rate": 0.00041647865312365186,
      "loss": 5.0135,
      "mean_token_accuracy": 0.17819776087999345,
      "num_tokens": 64153171.0,
      "step": 29285
    },
    {
      "entropy": 5.525965070724487,
      "epoch": 2.906907502977372,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041645110083319805,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2069973260164261,
      "num_tokens": 64163456.0,
      "step": 29290
    },
    {
      "entropy": 5.697732448577881,
      "epoch": 2.9074037316395396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041642354503506175,
      "loss": 5.0902,
      "mean_token_accuracy": 0.18337951004505157,
      "num_tokens": 64175728.0,
      "step": 29295
    },
    {
      "entropy": 5.738635015487671,
      "epoch": 2.907899960301707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041639598572992625,
      "loss": 5.0722,
      "mean_token_accuracy": 0.18063487559556962,
      "num_tokens": 64186600.0,
      "step": 29300
    },
    {
      "entropy": 5.663158988952636,
      "epoch": 2.9083961889638745,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041636842291847495,
      "loss": 5.0309,
      "mean_token_accuracy": 0.1851910173892975,
      "num_tokens": 64198112.0,
      "step": 29305
    },
    {
      "entropy": 5.697609710693359,
      "epoch": 2.908892417626042,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041634085660139145,
      "loss": 4.9849,
      "mean_token_accuracy": 0.19187327176332475,
      "num_tokens": 64209561.0,
      "step": 29310
    },
    {
      "entropy": 5.664290046691894,
      "epoch": 2.9093886462882095,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004163132867793592,
      "loss": 5.0539,
      "mean_token_accuracy": 0.18524486422538758,
      "num_tokens": 64221100.0,
      "step": 29315
    },
    {
      "entropy": 5.649159860610962,
      "epoch": 2.909884874950377,
      "grad_norm": 1.078125,
      "learning_rate": 0.000416285713453062,
      "loss": 4.9882,
      "mean_token_accuracy": 0.1867428407073021,
      "num_tokens": 64232262.0,
      "step": 29320
    },
    {
      "entropy": 5.7117901802062985,
      "epoch": 2.9103811036125444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004162581366231835,
      "loss": 5.0079,
      "mean_token_accuracy": 0.18557331711053848,
      "num_tokens": 64242722.0,
      "step": 29325
    },
    {
      "entropy": 5.727777671813965,
      "epoch": 2.9108773322747123,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041623055629040764,
      "loss": 5.0729,
      "mean_token_accuracy": 0.1792661115527153,
      "num_tokens": 64253286.0,
      "step": 29330
    },
    {
      "entropy": 5.723960256576538,
      "epoch": 2.91137356093688,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004162029724554183,
      "loss": 5.0901,
      "mean_token_accuracy": 0.17676005363464356,
      "num_tokens": 64264944.0,
      "step": 29335
    },
    {
      "entropy": 5.6507007598876955,
      "epoch": 2.9118697895990473,
      "grad_norm": 1.125,
      "learning_rate": 0.00041617538511889946,
      "loss": 5.0491,
      "mean_token_accuracy": 0.18365956842899323,
      "num_tokens": 64276681.0,
      "step": 29340
    },
    {
      "entropy": 5.679688119888306,
      "epoch": 2.9123660182612148,
      "grad_norm": 1.125,
      "learning_rate": 0.0004161477942815353,
      "loss": 4.9742,
      "mean_token_accuracy": 0.1892792820930481,
      "num_tokens": 64286586.0,
      "step": 29345
    },
    {
      "entropy": 5.688318967819214,
      "epoch": 2.9128622469233822,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041612019994401,
      "loss": 5.0524,
      "mean_token_accuracy": 0.17622076123952865,
      "num_tokens": 64297231.0,
      "step": 29350
    },
    {
      "entropy": 5.668228387832642,
      "epoch": 2.9133584755855497,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004160926021070078,
      "loss": 5.0431,
      "mean_token_accuracy": 0.18178654909133912,
      "num_tokens": 64308405.0,
      "step": 29355
    },
    {
      "entropy": 5.733885288238525,
      "epoch": 2.913854704247717,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004160650007712132,
      "loss": 5.013,
      "mean_token_accuracy": 0.18067943751811982,
      "num_tokens": 64318287.0,
      "step": 29360
    },
    {
      "entropy": 5.7145188331604,
      "epoch": 2.914350932909885,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041603739593731046,
      "loss": 5.063,
      "mean_token_accuracy": 0.1839607059955597,
      "num_tokens": 64330627.0,
      "step": 29365
    },
    {
      "entropy": 5.687706041336059,
      "epoch": 2.9148471615720526,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004160097876059842,
      "loss": 4.9624,
      "mean_token_accuracy": 0.18651782870292663,
      "num_tokens": 64340967.0,
      "step": 29370
    },
    {
      "entropy": 5.685855102539063,
      "epoch": 2.91534339023422,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004159821757779191,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18868025690317153,
      "num_tokens": 64351060.0,
      "step": 29375
    },
    {
      "entropy": 5.691730785369873,
      "epoch": 2.9158396188963875,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041595456045379975,
      "loss": 4.952,
      "mean_token_accuracy": 0.19553986340761184,
      "num_tokens": 64362177.0,
      "step": 29380
    },
    {
      "entropy": 5.77213077545166,
      "epoch": 2.916335847558555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004159269416343111,
      "loss": 5.0791,
      "mean_token_accuracy": 0.17984328418970108,
      "num_tokens": 64374188.0,
      "step": 29385
    },
    {
      "entropy": 5.742061519622803,
      "epoch": 2.9168320762207225,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000415899319320138,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1805914744734764,
      "num_tokens": 64384727.0,
      "step": 29390
    },
    {
      "entropy": 5.733381700515747,
      "epoch": 2.91732830488289,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041587169351196536,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18189629912376404,
      "num_tokens": 64395305.0,
      "step": 29395
    },
    {
      "entropy": 5.701836347579956,
      "epoch": 2.9178245335450574,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041584406421047825,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18514123558998108,
      "num_tokens": 64406746.0,
      "step": 29400
    },
    {
      "entropy": 5.608036518096924,
      "epoch": 2.918320762207225,
      "grad_norm": 1.125,
      "learning_rate": 0.00041581643141636184,
      "loss": 4.9296,
      "mean_token_accuracy": 0.18642636388540268,
      "num_tokens": 64417044.0,
      "step": 29405
    },
    {
      "entropy": 5.63268404006958,
      "epoch": 2.9188169908693924,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004157887951303014,
      "loss": 5.0013,
      "mean_token_accuracy": 0.1823923572897911,
      "num_tokens": 64427875.0,
      "step": 29410
    },
    {
      "entropy": 5.663475418090821,
      "epoch": 2.91931321953156,
      "grad_norm": 1.125,
      "learning_rate": 0.0004157611553529822,
      "loss": 5.0184,
      "mean_token_accuracy": 0.1892357811331749,
      "num_tokens": 64440086.0,
      "step": 29415
    },
    {
      "entropy": 5.699763011932373,
      "epoch": 2.919809448193728,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041573351208508966,
      "loss": 5.0353,
      "mean_token_accuracy": 0.18573205322027206,
      "num_tokens": 64450988.0,
      "step": 29420
    },
    {
      "entropy": 5.670711612701416,
      "epoch": 2.9203056768558953,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041570586532730926,
      "loss": 5.0185,
      "mean_token_accuracy": 0.17858829200267792,
      "num_tokens": 64461666.0,
      "step": 29425
    },
    {
      "entropy": 5.677643299102783,
      "epoch": 2.9208019055180627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004156782150803266,
      "loss": 5.0461,
      "mean_token_accuracy": 0.18481798619031906,
      "num_tokens": 64473111.0,
      "step": 29430
    },
    {
      "entropy": 5.672317028045654,
      "epoch": 2.92129813418023,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041565056134482733,
      "loss": 5.0306,
      "mean_token_accuracy": 0.1795763999223709,
      "num_tokens": 64484047.0,
      "step": 29435
    },
    {
      "entropy": 5.723748064041137,
      "epoch": 2.9217943628423977,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004156229041214972,
      "loss": 5.0811,
      "mean_token_accuracy": 0.18116793185472488,
      "num_tokens": 64496147.0,
      "step": 29440
    },
    {
      "entropy": 5.759414482116699,
      "epoch": 2.922290591504565,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004155952434110221,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17658353894948958,
      "num_tokens": 64507255.0,
      "step": 29445
    },
    {
      "entropy": 5.643433475494385,
      "epoch": 2.922786820166733,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004155675792140879,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18596506863832474,
      "num_tokens": 64519261.0,
      "step": 29450
    },
    {
      "entropy": 5.63418664932251,
      "epoch": 2.9232830488289006,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004155399115313806,
      "loss": 4.996,
      "mean_token_accuracy": 0.1804897367954254,
      "num_tokens": 64530618.0,
      "step": 29455
    },
    {
      "entropy": 5.709795713424683,
      "epoch": 2.923779277491068,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041551224036358644,
      "loss": 5.0628,
      "mean_token_accuracy": 0.181634059548378,
      "num_tokens": 64543399.0,
      "step": 29460
    },
    {
      "entropy": 5.661111307144165,
      "epoch": 2.9242755061532355,
      "grad_norm": 1.125,
      "learning_rate": 0.00041548456571139145,
      "loss": 4.9842,
      "mean_token_accuracy": 0.19085176438093185,
      "num_tokens": 64553673.0,
      "step": 29465
    },
    {
      "entropy": 5.77988567352295,
      "epoch": 2.924771734815403,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041545688757548184,
      "loss": 5.0978,
      "mean_token_accuracy": 0.18417341411113738,
      "num_tokens": 64563387.0,
      "step": 29470
    },
    {
      "entropy": 5.695212984085083,
      "epoch": 2.9252679634775705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041542920595654423,
      "loss": 4.9691,
      "mean_token_accuracy": 0.19067033380270004,
      "num_tokens": 64573877.0,
      "step": 29475
    },
    {
      "entropy": 5.757127094268799,
      "epoch": 2.925764192139738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041540152085526476,
      "loss": 5.0581,
      "mean_token_accuracy": 0.17866022884845734,
      "num_tokens": 64585222.0,
      "step": 29480
    },
    {
      "entropy": 5.673212385177612,
      "epoch": 2.9262604208019054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041537383227233023,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1851985350251198,
      "num_tokens": 64597303.0,
      "step": 29485
    },
    {
      "entropy": 5.629598760604859,
      "epoch": 2.926756649464073,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004153461402084271,
      "loss": 4.9551,
      "mean_token_accuracy": 0.1896263912320137,
      "num_tokens": 64608855.0,
      "step": 29490
    },
    {
      "entropy": 5.655457496643066,
      "epoch": 2.9272528781262404,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004153184446642421,
      "loss": 5.0297,
      "mean_token_accuracy": 0.18374375849962235,
      "num_tokens": 64617708.0,
      "step": 29495
    },
    {
      "entropy": 5.723411273956299,
      "epoch": 2.927749106788408,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041529074564046204,
      "loss": 5.0149,
      "mean_token_accuracy": 0.18428691774606704,
      "num_tokens": 64628494.0,
      "step": 29500
    },
    {
      "entropy": 5.723449993133545,
      "epoch": 2.9282453354505757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004152630431377738,
      "loss": 5.0352,
      "mean_token_accuracy": 0.18522339016199113,
      "num_tokens": 64639840.0,
      "step": 29505
    },
    {
      "entropy": 5.7318929672241214,
      "epoch": 2.9287415641127432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004152353371568643,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18471959233283997,
      "num_tokens": 64650842.0,
      "step": 29510
    },
    {
      "entropy": 5.697677659988403,
      "epoch": 2.9292377927749107,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041520762769842065,
      "loss": 4.9901,
      "mean_token_accuracy": 0.18673292845487593,
      "num_tokens": 64662383.0,
      "step": 29515
    },
    {
      "entropy": 5.66402153968811,
      "epoch": 2.929734021437078,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041517991476312994,
      "loss": 4.9945,
      "mean_token_accuracy": 0.1854688435792923,
      "num_tokens": 64672840.0,
      "step": 29520
    },
    {
      "entropy": 5.672208213806153,
      "epoch": 2.9302302500992456,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004151521983516794,
      "loss": 5.0056,
      "mean_token_accuracy": 0.1880277767777443,
      "num_tokens": 64684001.0,
      "step": 29525
    },
    {
      "entropy": 5.685013198852539,
      "epoch": 2.930726478761413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004151244784647563,
      "loss": 5.091,
      "mean_token_accuracy": 0.18409111499786376,
      "num_tokens": 64695967.0,
      "step": 29530
    },
    {
      "entropy": 5.671954727172851,
      "epoch": 2.931222707423581,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004150967551030482,
      "loss": 4.9504,
      "mean_token_accuracy": 0.18300808668136598,
      "num_tokens": 64706335.0,
      "step": 29535
    },
    {
      "entropy": 5.634836149215698,
      "epoch": 2.9317189360857485,
      "grad_norm": 1.125,
      "learning_rate": 0.00041506902826724234,
      "loss": 5.017,
      "mean_token_accuracy": 0.1821591377258301,
      "num_tokens": 64717833.0,
      "step": 29540
    },
    {
      "entropy": 5.60970721244812,
      "epoch": 2.932215164747916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004150412979580264,
      "loss": 4.9645,
      "mean_token_accuracy": 0.18852311819791795,
      "num_tokens": 64727378.0,
      "step": 29545
    },
    {
      "entropy": 5.619839286804199,
      "epoch": 2.9327113934100835,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041501356417608816,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1786886841058731,
      "num_tokens": 64738340.0,
      "step": 29550
    },
    {
      "entropy": 5.754138374328614,
      "epoch": 2.933207622072251,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041498582692211513,
      "loss": 4.9693,
      "mean_token_accuracy": 0.1894475743174553,
      "num_tokens": 64748137.0,
      "step": 29555
    },
    {
      "entropy": 5.695917749404908,
      "epoch": 2.9337038507344184,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041495808619679525,
      "loss": 4.9498,
      "mean_token_accuracy": 0.18431950509548187,
      "num_tokens": 64758829.0,
      "step": 29560
    },
    {
      "entropy": 5.688724613189697,
      "epoch": 2.934200079396586,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041493034200081643,
      "loss": 5.0296,
      "mean_token_accuracy": 0.17996067255735398,
      "num_tokens": 64770981.0,
      "step": 29565
    },
    {
      "entropy": 5.663575649261475,
      "epoch": 2.9346963080587534,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004149025943348666,
      "loss": 5.0176,
      "mean_token_accuracy": 0.1816238969564438,
      "num_tokens": 64781475.0,
      "step": 29570
    },
    {
      "entropy": 5.730685806274414,
      "epoch": 2.935192536720921,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041487484319963395,
      "loss": 5.1544,
      "mean_token_accuracy": 0.1802065670490265,
      "num_tokens": 64793478.0,
      "step": 29575
    },
    {
      "entropy": 5.673200607299805,
      "epoch": 2.9356887653830883,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004148470885958066,
      "loss": 4.9623,
      "mean_token_accuracy": 0.18533482551574706,
      "num_tokens": 64803959.0,
      "step": 29580
    },
    {
      "entropy": 5.695508527755737,
      "epoch": 2.936184994045256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004148193305240727,
      "loss": 5.0429,
      "mean_token_accuracy": 0.1807835355401039,
      "num_tokens": 64814426.0,
      "step": 29585
    },
    {
      "entropy": 5.63318510055542,
      "epoch": 2.9366812227074237,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041479156898512076,
      "loss": 4.9431,
      "mean_token_accuracy": 0.19487550109624863,
      "num_tokens": 64826128.0,
      "step": 29590
    },
    {
      "entropy": 5.666909694671631,
      "epoch": 2.937177451369591,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004147638039796391,
      "loss": 5.019,
      "mean_token_accuracy": 0.18675375133752822,
      "num_tokens": 64836285.0,
      "step": 29595
    },
    {
      "entropy": 5.682046699523926,
      "epoch": 2.9376736800317587,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041473603550831636,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18273622244596482,
      "num_tokens": 64848342.0,
      "step": 29600
    },
    {
      "entropy": 5.709911489486695,
      "epoch": 2.938169908693926,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041470826357184096,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1818330004811287,
      "num_tokens": 64859541.0,
      "step": 29605
    },
    {
      "entropy": 5.693556547164917,
      "epoch": 2.9386661373560936,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004146804881709017,
      "loss": 5.0412,
      "mean_token_accuracy": 0.18034396767616273,
      "num_tokens": 64870816.0,
      "step": 29610
    },
    {
      "entropy": 5.7545778274536135,
      "epoch": 2.939162366018261,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004146527093061873,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1793798476457596,
      "num_tokens": 64882604.0,
      "step": 29615
    },
    {
      "entropy": 5.71081805229187,
      "epoch": 2.939658594680429,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004146249269783867,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1921069711446762,
      "num_tokens": 64893861.0,
      "step": 29620
    },
    {
      "entropy": 5.734068155288696,
      "epoch": 2.9401548233425965,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145971411881886,
      "loss": 5.0596,
      "mean_token_accuracy": 0.18073368072509766,
      "num_tokens": 64903863.0,
      "step": 29625
    },
    {
      "entropy": 5.74260835647583,
      "epoch": 2.940651052004764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041456935193628236,
      "loss": 5.0769,
      "mean_token_accuracy": 0.17670133709907532,
      "num_tokens": 64915754.0,
      "step": 29630
    },
    {
      "entropy": 5.631310081481933,
      "epoch": 2.9411472806669314,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004145415592233569,
      "loss": 4.9408,
      "mean_token_accuracy": 0.18613498657941818,
      "num_tokens": 64926217.0,
      "step": 29635
    },
    {
      "entropy": 5.6695951461792,
      "epoch": 2.941643509329099,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041451376305010144,
      "loss": 4.9662,
      "mean_token_accuracy": 0.19089522659778596,
      "num_tokens": 64936478.0,
      "step": 29640
    },
    {
      "entropy": 5.699524307250977,
      "epoch": 2.9421397379912664,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004144859634172053,
      "loss": 5.0731,
      "mean_token_accuracy": 0.18188085854053498,
      "num_tokens": 64948195.0,
      "step": 29645
    },
    {
      "entropy": 5.725248241424561,
      "epoch": 2.942635966653434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041445816032535784,
      "loss": 5.1498,
      "mean_token_accuracy": 0.17725826501846315,
      "num_tokens": 64959019.0,
      "step": 29650
    },
    {
      "entropy": 5.7670985698699955,
      "epoch": 2.9431321953156013,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041443035377524844,
      "loss": 5.0538,
      "mean_token_accuracy": 0.17892611026763916,
      "num_tokens": 64970147.0,
      "step": 29655
    },
    {
      "entropy": 5.718767404556274,
      "epoch": 2.943628423977769,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004144025437675668,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18639361262321472,
      "num_tokens": 64980664.0,
      "step": 29660
    },
    {
      "entropy": 5.669567108154297,
      "epoch": 2.9441246526399363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004143747303030025,
      "loss": 5.0871,
      "mean_token_accuracy": 0.18919861912727357,
      "num_tokens": 64992489.0,
      "step": 29665
    },
    {
      "entropy": 5.633076000213623,
      "epoch": 2.9446208813021038,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041434691338224506,
      "loss": 4.9844,
      "mean_token_accuracy": 0.18284692317247392,
      "num_tokens": 65002861.0,
      "step": 29670
    },
    {
      "entropy": 5.681804656982422,
      "epoch": 2.9451171099642717,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004143190930059846,
      "loss": 5.0593,
      "mean_token_accuracy": 0.18028222173452377,
      "num_tokens": 65013865.0,
      "step": 29675
    },
    {
      "entropy": 5.6997661113739015,
      "epoch": 2.945613338626439,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004142912691749107,
      "loss": 5.0695,
      "mean_token_accuracy": 0.17896425127983093,
      "num_tokens": 65026155.0,
      "step": 29680
    },
    {
      "entropy": 5.72919921875,
      "epoch": 2.9461095672886066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004142634418897135,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17778357863426208,
      "num_tokens": 65037306.0,
      "step": 29685
    },
    {
      "entropy": 5.712626886367798,
      "epoch": 2.946605795950774,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004142356111510831,
      "loss": 5.0576,
      "mean_token_accuracy": 0.1850773051381111,
      "num_tokens": 65047715.0,
      "step": 29690
    },
    {
      "entropy": 5.684674644470215,
      "epoch": 2.9471020246129416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041420777695970945,
      "loss": 4.9978,
      "mean_token_accuracy": 0.18957942426204683,
      "num_tokens": 65057694.0,
      "step": 29695
    },
    {
      "entropy": 5.703702020645141,
      "epoch": 2.947598253275109,
      "grad_norm": 1.203125,
      "learning_rate": 0.000414179939316283,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17763662934303284,
      "num_tokens": 65068131.0,
      "step": 29700
    },
    {
      "entropy": 5.7642427444458,
      "epoch": 2.948094481937277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041415209822149397,
      "loss": 5.121,
      "mean_token_accuracy": 0.17302835434675218,
      "num_tokens": 65081044.0,
      "step": 29705
    },
    {
      "entropy": 5.7396484375,
      "epoch": 2.9485907105994444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004141242536760327,
      "loss": 5.0693,
      "mean_token_accuracy": 0.18047548830509186,
      "num_tokens": 65092151.0,
      "step": 29710
    },
    {
      "entropy": 5.742292737960815,
      "epoch": 2.949086939261612,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004140964056805897,
      "loss": 5.0271,
      "mean_token_accuracy": 0.1862584948539734,
      "num_tokens": 65103787.0,
      "step": 29715
    },
    {
      "entropy": 5.704477739334107,
      "epoch": 2.9495831679237794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004140685542358556,
      "loss": 5.0733,
      "mean_token_accuracy": 0.1812692478299141,
      "num_tokens": 65115157.0,
      "step": 29720
    },
    {
      "entropy": 5.701156902313232,
      "epoch": 2.950079396585947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004140406993425211,
      "loss": 5.0833,
      "mean_token_accuracy": 0.18182234913110734,
      "num_tokens": 65127090.0,
      "step": 29725
    },
    {
      "entropy": 5.700107717514038,
      "epoch": 2.9505756252481143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004140128410012767,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18527742773294448,
      "num_tokens": 65138346.0,
      "step": 29730
    },
    {
      "entropy": 5.767052459716797,
      "epoch": 2.951071853910282,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041398497921281354,
      "loss": 5.135,
      "mean_token_accuracy": 0.17376379668712616,
      "num_tokens": 65149510.0,
      "step": 29735
    },
    {
      "entropy": 5.715350103378296,
      "epoch": 2.9515680825724493,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041395711397782236,
      "loss": 5.0822,
      "mean_token_accuracy": 0.1794881373643875,
      "num_tokens": 65159828.0,
      "step": 29740
    },
    {
      "entropy": 5.739323425292969,
      "epoch": 2.9520643112346168,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004139292452969941,
      "loss": 5.011,
      "mean_token_accuracy": 0.18789291381835938,
      "num_tokens": 65170594.0,
      "step": 29745
    },
    {
      "entropy": 5.716022205352783,
      "epoch": 2.9525605398967842,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041390137317102,
      "loss": 5.0394,
      "mean_token_accuracy": 0.18538489937782288,
      "num_tokens": 65180722.0,
      "step": 29750
    },
    {
      "entropy": 5.736873531341553,
      "epoch": 2.9530567685589517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004138734976005911,
      "loss": 5.0008,
      "mean_token_accuracy": 0.189426951110363,
      "num_tokens": 65191390.0,
      "step": 29755
    },
    {
      "entropy": 5.663537645339966,
      "epoch": 2.9535529972211196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004138456185863986,
      "loss": 5.0016,
      "mean_token_accuracy": 0.1824214816093445,
      "num_tokens": 65202915.0,
      "step": 29760
    },
    {
      "entropy": 5.591426753997803,
      "epoch": 2.954049225883287,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041381773612913413,
      "loss": 4.969,
      "mean_token_accuracy": 0.19300899058580398,
      "num_tokens": 65213098.0,
      "step": 29765
    },
    {
      "entropy": 5.704364967346192,
      "epoch": 2.9545454545454546,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004137898502294889,
      "loss": 5.084,
      "mean_token_accuracy": 0.1826597884297371,
      "num_tokens": 65224003.0,
      "step": 29770
    },
    {
      "entropy": 5.748568201065064,
      "epoch": 2.955041683207622,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004137619608881544,
      "loss": 5.067,
      "mean_token_accuracy": 0.18383711874485015,
      "num_tokens": 65234520.0,
      "step": 29775
    },
    {
      "entropy": 5.68858380317688,
      "epoch": 2.9555379118697895,
      "grad_norm": 1.125,
      "learning_rate": 0.0004137340681058223,
      "loss": 4.9828,
      "mean_token_accuracy": 0.18305407613515853,
      "num_tokens": 65245117.0,
      "step": 29780
    },
    {
      "entropy": 5.566788291931152,
      "epoch": 2.956034140531957,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041370617188318424,
      "loss": 4.902,
      "mean_token_accuracy": 0.19606953412294387,
      "num_tokens": 65254933.0,
      "step": 29785
    },
    {
      "entropy": 5.6885823726654055,
      "epoch": 2.956530369194125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041367827222093195,
      "loss": 5.0907,
      "mean_token_accuracy": 0.18506149500608443,
      "num_tokens": 65266253.0,
      "step": 29790
    },
    {
      "entropy": 5.6579615592956545,
      "epoch": 2.9570265978562924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004136503691197574,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18502936214208604,
      "num_tokens": 65276595.0,
      "step": 29795
    },
    {
      "entropy": 5.721021556854248,
      "epoch": 2.95752282651846,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041362246258035254,
      "loss": 5.1112,
      "mean_token_accuracy": 0.18296367824077606,
      "num_tokens": 65286915.0,
      "step": 29800
    },
    {
      "entropy": 5.6961822509765625,
      "epoch": 2.9580190551806274,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004135945526034091,
      "loss": 4.9797,
      "mean_token_accuracy": 0.19115010350942613,
      "num_tokens": 65297613.0,
      "step": 29805
    },
    {
      "entropy": 5.713046550750732,
      "epoch": 2.958515283842795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041356663918961956,
      "loss": 5.0194,
      "mean_token_accuracy": 0.17870436012744903,
      "num_tokens": 65309273.0,
      "step": 29810
    },
    {
      "entropy": 5.750174808502197,
      "epoch": 2.9590115125049623,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041353872233967595,
      "loss": 5.0694,
      "mean_token_accuracy": 0.18387485295534134,
      "num_tokens": 65321241.0,
      "step": 29815
    },
    {
      "entropy": 5.668472623825073,
      "epoch": 2.95950774116713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004135108020542705,
      "loss": 4.9528,
      "mean_token_accuracy": 0.18806470036506653,
      "num_tokens": 65331520.0,
      "step": 29820
    },
    {
      "entropy": 5.669593477249146,
      "epoch": 2.9600039698292973,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041348287833409557,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18653680980205536,
      "num_tokens": 65341551.0,
      "step": 29825
    },
    {
      "entropy": 5.666712951660156,
      "epoch": 2.9605001984914647,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004134549511798438,
      "loss": 5.0433,
      "mean_token_accuracy": 0.1804463282227516,
      "num_tokens": 65352489.0,
      "step": 29830
    },
    {
      "entropy": 5.641003608703613,
      "epoch": 2.960996427153632,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004134270205922075,
      "loss": 4.9852,
      "mean_token_accuracy": 0.1912762850522995,
      "num_tokens": 65363298.0,
      "step": 29835
    },
    {
      "entropy": 5.786199712753296,
      "epoch": 2.9614926558157997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041339908657187933,
      "loss": 5.1196,
      "mean_token_accuracy": 0.18293869793415068,
      "num_tokens": 65374426.0,
      "step": 29840
    },
    {
      "entropy": 5.749694681167602,
      "epoch": 2.9619888844779676,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004133711491195522,
      "loss": 5.0292,
      "mean_token_accuracy": 0.1860233172774315,
      "num_tokens": 65386007.0,
      "step": 29845
    },
    {
      "entropy": 5.671528053283692,
      "epoch": 2.962485113140135,
      "grad_norm": 1.125,
      "learning_rate": 0.00041334320823591857,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1838751956820488,
      "num_tokens": 65397457.0,
      "step": 29850
    },
    {
      "entropy": 5.66375241279602,
      "epoch": 2.9629813418023025,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041331526392167155,
      "loss": 4.9731,
      "mean_token_accuracy": 0.1898340329527855,
      "num_tokens": 65407905.0,
      "step": 29855
    },
    {
      "entropy": 5.710375642776489,
      "epoch": 2.96347757046447,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000413287316177504,
      "loss": 5.0287,
      "mean_token_accuracy": 0.1834689274430275,
      "num_tokens": 65418946.0,
      "step": 29860
    },
    {
      "entropy": 5.665611028671265,
      "epoch": 2.9639737991266375,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041325936500410905,
      "loss": 4.9486,
      "mean_token_accuracy": 0.19642328917980195,
      "num_tokens": 65428700.0,
      "step": 29865
    },
    {
      "entropy": 5.669977140426636,
      "epoch": 2.964470027788805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041323141040217976,
      "loss": 4.9904,
      "mean_token_accuracy": 0.18810449093580245,
      "num_tokens": 65439473.0,
      "step": 29870
    },
    {
      "entropy": 5.696316289901733,
      "epoch": 2.9649662564509724,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00041320345237240923,
      "loss": 5.0395,
      "mean_token_accuracy": 0.19002858102321624,
      "num_tokens": 65451799.0,
      "step": 29875
    },
    {
      "entropy": 5.723939847946167,
      "epoch": 2.9654624851131404,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000413175490915491,
      "loss": 5.0493,
      "mean_token_accuracy": 0.18040190488100052,
      "num_tokens": 65462890.0,
      "step": 29880
    },
    {
      "entropy": 5.676021385192871,
      "epoch": 2.965958713775308,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004131475260321183,
      "loss": 4.9925,
      "mean_token_accuracy": 0.18191787898540496,
      "num_tokens": 65473503.0,
      "step": 29885
    },
    {
      "entropy": 5.690082645416259,
      "epoch": 2.9664549424374753,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041311955772298464,
      "loss": 5.0934,
      "mean_token_accuracy": 0.1776578202843666,
      "num_tokens": 65485383.0,
      "step": 29890
    },
    {
      "entropy": 5.587419748306274,
      "epoch": 2.966951171099643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004130915859887837,
      "loss": 4.8738,
      "mean_token_accuracy": 0.18874238580465316,
      "num_tokens": 65496044.0,
      "step": 29895
    },
    {
      "entropy": 5.675225353240966,
      "epoch": 2.9674473997618103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004130636108302088,
      "loss": 5.0152,
      "mean_token_accuracy": 0.18653408139944078,
      "num_tokens": 65506673.0,
      "step": 29900
    },
    {
      "entropy": 5.672402620315552,
      "epoch": 2.9679436284239777,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041303563224795393,
      "loss": 5.003,
      "mean_token_accuracy": 0.17707851380109788,
      "num_tokens": 65516492.0,
      "step": 29905
    },
    {
      "entropy": 5.688746309280395,
      "epoch": 2.968439857086145,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004130076502427128,
      "loss": 4.9693,
      "mean_token_accuracy": 0.18719644099473953,
      "num_tokens": 65528240.0,
      "step": 29910
    },
    {
      "entropy": 5.7517022609710695,
      "epoch": 2.9689360857483127,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041297966481517937,
      "loss": 5.0562,
      "mean_token_accuracy": 0.18306194394826888,
      "num_tokens": 65537861.0,
      "step": 29915
    },
    {
      "entropy": 5.663996362686158,
      "epoch": 2.96943231441048,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004129516759660475,
      "loss": 5.0406,
      "mean_token_accuracy": 0.18107558339834212,
      "num_tokens": 65549326.0,
      "step": 29920
    },
    {
      "entropy": 5.668861722946167,
      "epoch": 2.9699285430726476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041292368369601136,
      "loss": 4.9797,
      "mean_token_accuracy": 0.1803616017103195,
      "num_tokens": 65560962.0,
      "step": 29925
    },
    {
      "entropy": 5.7447381019592285,
      "epoch": 2.9704247717348156,
      "grad_norm": 1.0625,
      "learning_rate": 0.000412895688005765,
      "loss": 5.059,
      "mean_token_accuracy": 0.1911727011203766,
      "num_tokens": 65572411.0,
      "step": 29930
    },
    {
      "entropy": 5.715994215011596,
      "epoch": 2.970921000396983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004128676888960028,
      "loss": 5.0695,
      "mean_token_accuracy": 0.18192462921142577,
      "num_tokens": 65582423.0,
      "step": 29935
    },
    {
      "entropy": 5.72514967918396,
      "epoch": 2.9714172290591505,
      "grad_norm": 1.25,
      "learning_rate": 0.00041283968636741903,
      "loss": 5.0458,
      "mean_token_accuracy": 0.18941385746002198,
      "num_tokens": 65594205.0,
      "step": 29940
    },
    {
      "entropy": 5.6393928050994875,
      "epoch": 2.971913457721318,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004128116804207081,
      "loss": 4.9751,
      "mean_token_accuracy": 0.18885877579450608,
      "num_tokens": 65605262.0,
      "step": 29945
    },
    {
      "entropy": 5.6692211627960205,
      "epoch": 2.9724096863834855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004127836710565642,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1814430132508278,
      "num_tokens": 65616175.0,
      "step": 29950
    },
    {
      "entropy": 5.67292537689209,
      "epoch": 2.972905915045653,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041275565827568234,
      "loss": 5.042,
      "mean_token_accuracy": 0.17941778749227524,
      "num_tokens": 65628128.0,
      "step": 29955
    },
    {
      "entropy": 5.715270614624023,
      "epoch": 2.9734021437078204,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000412727642078757,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18745303452014922,
      "num_tokens": 65639605.0,
      "step": 29960
    },
    {
      "entropy": 5.69344048500061,
      "epoch": 2.9738983723699883,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004126996224664829,
      "loss": 5.0754,
      "mean_token_accuracy": 0.18558795899152755,
      "num_tokens": 65652552.0,
      "step": 29965
    },
    {
      "entropy": 5.7475684642791744,
      "epoch": 2.974394601032156,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041267159943955483,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18577490895986556,
      "num_tokens": 65662423.0,
      "step": 29970
    },
    {
      "entropy": 5.715937328338623,
      "epoch": 2.9748908296943233,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004126435729986678,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18530096858739853,
      "num_tokens": 65674582.0,
      "step": 29975
    },
    {
      "entropy": 5.6433769226074215,
      "epoch": 2.9753870583564908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041261554314451674,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18177585452795028,
      "num_tokens": 65685321.0,
      "step": 29980
    },
    {
      "entropy": 5.623621988296509,
      "epoch": 2.9758832870186582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004125875098777967,
      "loss": 4.9891,
      "mean_token_accuracy": 0.1886374235153198,
      "num_tokens": 65696000.0,
      "step": 29985
    },
    {
      "entropy": 5.6865263938903805,
      "epoch": 2.9763795156808257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041255947319920297,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18751096576452256,
      "num_tokens": 65705930.0,
      "step": 29990
    },
    {
      "entropy": 5.650722503662109,
      "epoch": 2.976875744342993,
      "grad_norm": 1.125,
      "learning_rate": 0.00041253143310943064,
      "loss": 4.9785,
      "mean_token_accuracy": 0.19004721343517303,
      "num_tokens": 65717038.0,
      "step": 29995
    },
    {
      "entropy": 5.694242238998413,
      "epoch": 2.9773719730051607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041250338960917515,
      "loss": 5.067,
      "mean_token_accuracy": 0.18459467738866805,
      "num_tokens": 65728915.0,
      "step": 30000
    },
    {
      "epoch": 2.9773719730051607,
      "eval_entropy": 5.467112676240283,
      "eval_loss": 5.134406089782715,
      "eval_mean_token_accuracy": 0.1856139648363196,
      "eval_num_tokens": 65728915.0,
      "eval_runtime": 20.1909,
      "eval_samples_per_second": 1667.529,
      "eval_steps_per_second": 208.46,
      "step": 30000
    }
  ],
  "logging_steps": 5,
  "max_steps": 100750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.637884693389312e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}