jpn-jpan-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed455/checkpoint-24000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.7697634160415463,
  "eval_steps": 3000,
  "global_step": 24000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.512531781196594,
      "epoch": 0.0005770340450086555,
      "grad_norm": 18.125,
      "learning_rate": 2e-06,
      "loss": 14.35,
      "mean_token_accuracy": 0.04723494742065668,
      "num_tokens": 13337.0,
      "step": 5
    },
    {
      "entropy": 2.5895443677902223,
      "epoch": 0.001154068090017311,
      "grad_norm": 17.5,
      "learning_rate": 4.5e-06,
      "loss": 14.2242,
      "mean_token_accuracy": 0.045978688448667523,
      "num_tokens": 26339.0,
      "step": 10
    },
    {
      "entropy": 2.6310792684555055,
      "epoch": 0.0017311021350259665,
      "grad_norm": 19.5,
      "learning_rate": 7e-06,
      "loss": 14.3304,
      "mean_token_accuracy": 0.03695605173707008,
      "num_tokens": 39118.0,
      "step": 15
    },
    {
      "entropy": 2.683561897277832,
      "epoch": 0.002308136180034622,
      "grad_norm": 20.625,
      "learning_rate": 9.5e-06,
      "loss": 14.3454,
      "mean_token_accuracy": 0.029470206797122957,
      "num_tokens": 51367.0,
      "step": 20
    },
    {
      "entropy": 3.0045490741729735,
      "epoch": 0.0028851702250432777,
      "grad_norm": 25.0,
      "learning_rate": 1.2e-05,
      "loss": 13.9558,
      "mean_token_accuracy": 0.037074245233088735,
      "num_tokens": 64639.0,
      "step": 25
    },
    {
      "entropy": 3.8651180982589723,
      "epoch": 0.003462204270051933,
      "grad_norm": 24.5,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.2224,
      "mean_token_accuracy": 0.03461765330284834,
      "num_tokens": 76244.0,
      "step": 30
    },
    {
      "entropy": 5.199391555786133,
      "epoch": 0.004039238315060588,
      "grad_norm": 41.25,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 12.2743,
      "mean_token_accuracy": 0.028778869472444058,
      "num_tokens": 88490.0,
      "step": 35
    },
    {
      "entropy": 8.932403516769408,
      "epoch": 0.004616272360069244,
      "grad_norm": 8.0625,
      "learning_rate": 1.95e-05,
      "loss": 10.7862,
      "mean_token_accuracy": 0.016108644660562277,
      "num_tokens": 102022.0,
      "step": 40
    },
    {
      "entropy": 10.609907341003417,
      "epoch": 0.0051933064050779,
      "grad_norm": 5.03125,
      "learning_rate": 2.2e-05,
      "loss": 10.5807,
      "mean_token_accuracy": 0.008532280148938299,
      "num_tokens": 114442.0,
      "step": 45
    },
    {
      "entropy": 10.433589267730714,
      "epoch": 0.005770340450086555,
      "grad_norm": 2.59375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.4268,
      "mean_token_accuracy": 0.01151183140464127,
      "num_tokens": 128044.0,
      "step": 50
    },
    {
      "entropy": 10.309624099731446,
      "epoch": 0.006347374495095211,
      "grad_norm": 2.5,
      "learning_rate": 2.7e-05,
      "loss": 10.344,
      "mean_token_accuracy": 0.027350079361349346,
      "num_tokens": 141818.0,
      "step": 55
    },
    {
      "entropy": 10.577155780792236,
      "epoch": 0.006924408540103866,
      "grad_norm": 2.265625,
      "learning_rate": 2.95e-05,
      "loss": 10.1314,
      "mean_token_accuracy": 0.03637834042310715,
      "num_tokens": 155109.0,
      "step": 60
    },
    {
      "entropy": 10.506378841400146,
      "epoch": 0.0075014425851125215,
      "grad_norm": 2.09375,
      "learning_rate": 3.2e-05,
      "loss": 9.9578,
      "mean_token_accuracy": 0.04525792058557272,
      "num_tokens": 167036.0,
      "step": 65
    },
    {
      "entropy": 10.556257247924805,
      "epoch": 0.008078476630121177,
      "grad_norm": 2.34375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.9262,
      "mean_token_accuracy": 0.050815035589039326,
      "num_tokens": 181398.0,
      "step": 70
    },
    {
      "entropy": 10.460558032989502,
      "epoch": 0.008655510675129832,
      "grad_norm": 1.7578125,
      "learning_rate": 3.7e-05,
      "loss": 9.7428,
      "mean_token_accuracy": 0.06986093260347843,
      "num_tokens": 193242.0,
      "step": 75
    },
    {
      "entropy": 10.52681770324707,
      "epoch": 0.009232544720138488,
      "grad_norm": 1.7578125,
      "learning_rate": 3.95e-05,
      "loss": 9.7351,
      "mean_token_accuracy": 0.06739085242152214,
      "num_tokens": 206744.0,
      "step": 80
    },
    {
      "entropy": 10.430398750305176,
      "epoch": 0.009809578765147143,
      "grad_norm": 2.046875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6365,
      "mean_token_accuracy": 0.06604859046638012,
      "num_tokens": 219120.0,
      "step": 85
    },
    {
      "entropy": 10.513145351409912,
      "epoch": 0.0103866128101558,
      "grad_norm": 2.21875,
      "learning_rate": 4.45e-05,
      "loss": 9.5987,
      "mean_token_accuracy": 0.06549847275018691,
      "num_tokens": 231942.0,
      "step": 90
    },
    {
      "entropy": 10.511101913452148,
      "epoch": 0.010963646855164455,
      "grad_norm": 1.6796875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5376,
      "mean_token_accuracy": 0.07021979242563248,
      "num_tokens": 244786.0,
      "step": 95
    },
    {
      "entropy": 10.338568496704102,
      "epoch": 0.01154068090017311,
      "grad_norm": 1.859375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3663,
      "mean_token_accuracy": 0.07713823206722736,
      "num_tokens": 257503.0,
      "step": 100
    },
    {
      "entropy": 10.395394325256348,
      "epoch": 0.012117714945181766,
      "grad_norm": 1.6328125,
      "learning_rate": 5.2e-05,
      "loss": 9.3013,
      "mean_token_accuracy": 0.07871861308813095,
      "num_tokens": 270330.0,
      "step": 105
    },
    {
      "entropy": 10.267610359191895,
      "epoch": 0.012694748990190421,
      "grad_norm": 1.3828125,
      "learning_rate": 5.45e-05,
      "loss": 9.289,
      "mean_token_accuracy": 0.08360295034945012,
      "num_tokens": 283866.0,
      "step": 110
    },
    {
      "entropy": 10.299630165100098,
      "epoch": 0.013271783035199077,
      "grad_norm": 1.5078125,
      "learning_rate": 5.7e-05,
      "loss": 9.219,
      "mean_token_accuracy": 0.07119438238441944,
      "num_tokens": 296715.0,
      "step": 115
    },
    {
      "entropy": 10.345106220245361,
      "epoch": 0.013848817080207732,
      "grad_norm": 1.515625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 9.1287,
      "mean_token_accuracy": 0.06990180611610412,
      "num_tokens": 309268.0,
      "step": 120
    },
    {
      "entropy": 10.140081024169922,
      "epoch": 0.014425851125216388,
      "grad_norm": 2.109375,
      "learning_rate": 6.2e-05,
      "loss": 9.0404,
      "mean_token_accuracy": 0.0861529141664505,
      "num_tokens": 322471.0,
      "step": 125
    },
    {
      "entropy": 10.096904373168945,
      "epoch": 0.015002885170225043,
      "grad_norm": 1.625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 9.0344,
      "mean_token_accuracy": 0.07451867423951626,
      "num_tokens": 335728.0,
      "step": 130
    },
    {
      "entropy": 10.11425609588623,
      "epoch": 0.015579919215233698,
      "grad_norm": 1.34375,
      "learning_rate": 6.7e-05,
      "loss": 8.8978,
      "mean_token_accuracy": 0.07459303736686707,
      "num_tokens": 347861.0,
      "step": 135
    },
    {
      "entropy": 9.995181465148926,
      "epoch": 0.016156953260242354,
      "grad_norm": 1.3671875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.868,
      "mean_token_accuracy": 0.08410398475825787,
      "num_tokens": 360275.0,
      "step": 140
    },
    {
      "entropy": 9.862716579437256,
      "epoch": 0.01673398730525101,
      "grad_norm": 1.1171875,
      "learning_rate": 7.2e-05,
      "loss": 8.8007,
      "mean_token_accuracy": 0.07614478804171085,
      "num_tokens": 373948.0,
      "step": 145
    },
    {
      "entropy": 9.893266773223877,
      "epoch": 0.017311021350259664,
      "grad_norm": 1.09375,
      "learning_rate": 7.45e-05,
      "loss": 8.7368,
      "mean_token_accuracy": 0.07509239986538888,
      "num_tokens": 386267.0,
      "step": 150
    },
    {
      "entropy": 9.75681676864624,
      "epoch": 0.01788805539526832,
      "grad_norm": 0.93359375,
      "learning_rate": 7.7e-05,
      "loss": 8.6465,
      "mean_token_accuracy": 0.07349924966692925,
      "num_tokens": 398595.0,
      "step": 155
    },
    {
      "entropy": 9.599973392486572,
      "epoch": 0.018465089440276975,
      "grad_norm": 0.9609375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.5952,
      "mean_token_accuracy": 0.0733072243630886,
      "num_tokens": 410626.0,
      "step": 160
    },
    {
      "entropy": 9.452995586395264,
      "epoch": 0.019042123485285632,
      "grad_norm": 0.921875,
      "learning_rate": 8.2e-05,
      "loss": 8.5567,
      "mean_token_accuracy": 0.07969285026192666,
      "num_tokens": 423176.0,
      "step": 165
    },
    {
      "entropy": 9.233365058898926,
      "epoch": 0.019619157530294286,
      "grad_norm": 1.2890625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.4908,
      "mean_token_accuracy": 0.08393557146191596,
      "num_tokens": 435997.0,
      "step": 170
    },
    {
      "entropy": 9.169007682800293,
      "epoch": 0.020196191575302943,
      "grad_norm": 0.86328125,
      "learning_rate": 8.7e-05,
      "loss": 8.459,
      "mean_token_accuracy": 0.08070328459143639,
      "num_tokens": 449295.0,
      "step": 175
    },
    {
      "entropy": 9.00176544189453,
      "epoch": 0.0207732256203116,
      "grad_norm": 0.81640625,
      "learning_rate": 8.95e-05,
      "loss": 8.4618,
      "mean_token_accuracy": 0.08123850636184216,
      "num_tokens": 463983.0,
      "step": 180
    },
    {
      "entropy": 8.961482334136964,
      "epoch": 0.021350259665320254,
      "grad_norm": 0.83984375,
      "learning_rate": 9.2e-05,
      "loss": 8.4659,
      "mean_token_accuracy": 0.07538431286811828,
      "num_tokens": 476849.0,
      "step": 185
    },
    {
      "entropy": 8.856469058990479,
      "epoch": 0.02192729371032891,
      "grad_norm": 0.9140625,
      "learning_rate": 9.45e-05,
      "loss": 8.4194,
      "mean_token_accuracy": 0.0841868095099926,
      "num_tokens": 489120.0,
      "step": 190
    },
    {
      "entropy": 8.735506343841553,
      "epoch": 0.022504327755337564,
      "grad_norm": 0.9453125,
      "learning_rate": 9.7e-05,
      "loss": 8.338,
      "mean_token_accuracy": 0.08333569169044494,
      "num_tokens": 501085.0,
      "step": 195
    },
    {
      "entropy": 8.774681186676025,
      "epoch": 0.02308136180034622,
      "grad_norm": 0.78515625,
      "learning_rate": 9.95e-05,
      "loss": 8.4191,
      "mean_token_accuracy": 0.07852545715868472,
      "num_tokens": 513207.0,
      "step": 200
    },
    {
      "entropy": 8.802916240692138,
      "epoch": 0.023658395845354875,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000102,
      "loss": 8.3567,
      "mean_token_accuracy": 0.07992818728089332,
      "num_tokens": 524809.0,
      "step": 205
    },
    {
      "entropy": 8.692363357543945,
      "epoch": 0.024235429890363532,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.3929,
      "mean_token_accuracy": 0.0764585092663765,
      "num_tokens": 538950.0,
      "step": 210
    },
    {
      "entropy": 8.760814189910889,
      "epoch": 0.024812463935372186,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000107,
      "loss": 8.367,
      "mean_token_accuracy": 0.08504860773682595,
      "num_tokens": 551199.0,
      "step": 215
    },
    {
      "entropy": 8.599481773376464,
      "epoch": 0.025389497980380843,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0001095,
      "loss": 8.3044,
      "mean_token_accuracy": 0.08574924021959304,
      "num_tokens": 563202.0,
      "step": 220
    },
    {
      "entropy": 8.647394752502441,
      "epoch": 0.025966532025389497,
      "grad_norm": 0.6484375,
      "learning_rate": 0.000112,
      "loss": 8.2907,
      "mean_token_accuracy": 0.0885565310716629,
      "num_tokens": 576716.0,
      "step": 225
    },
    {
      "entropy": 8.612112045288086,
      "epoch": 0.026543566070398154,
      "grad_norm": 0.875,
      "learning_rate": 0.0001145,
      "loss": 8.3039,
      "mean_token_accuracy": 0.08598211780190468,
      "num_tokens": 588413.0,
      "step": 230
    },
    {
      "entropy": 8.698466682434082,
      "epoch": 0.027120600115406807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00011700000000000001,
      "loss": 8.3579,
      "mean_token_accuracy": 0.08150896914303303,
      "num_tokens": 602052.0,
      "step": 235
    },
    {
      "entropy": 8.66174020767212,
      "epoch": 0.027697634160415464,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00011949999999999999,
      "loss": 8.3016,
      "mean_token_accuracy": 0.08273345604538918,
      "num_tokens": 614508.0,
      "step": 240
    },
    {
      "entropy": 8.5486985206604,
      "epoch": 0.02827466820542412,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000122,
      "loss": 8.3261,
      "mean_token_accuracy": 0.08385896608233452,
      "num_tokens": 628758.0,
      "step": 245
    },
    {
      "entropy": 8.613620281219482,
      "epoch": 0.028851702250432775,
      "grad_norm": 0.84375,
      "learning_rate": 0.0001245,
      "loss": 8.2462,
      "mean_token_accuracy": 0.08523005284368992,
      "num_tokens": 641046.0,
      "step": 250
    },
    {
      "entropy": 8.58726749420166,
      "epoch": 0.029428736295441432,
      "grad_norm": 0.76171875,
      "learning_rate": 0.000127,
      "loss": 8.2527,
      "mean_token_accuracy": 0.08660949915647506,
      "num_tokens": 653910.0,
      "step": 255
    },
    {
      "entropy": 8.568908309936523,
      "epoch": 0.030005770340450086,
      "grad_norm": 0.796875,
      "learning_rate": 0.0001295,
      "loss": 8.2246,
      "mean_token_accuracy": 0.08737964332103729,
      "num_tokens": 665862.0,
      "step": 260
    },
    {
      "entropy": 8.553102970123291,
      "epoch": 0.030582804385458743,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000132,
      "loss": 8.2409,
      "mean_token_accuracy": 0.08849976807832718,
      "num_tokens": 678854.0,
      "step": 265
    },
    {
      "entropy": 8.565754413604736,
      "epoch": 0.031159838430467397,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00013450000000000002,
      "loss": 8.2342,
      "mean_token_accuracy": 0.08802064135670662,
      "num_tokens": 690188.0,
      "step": 270
    },
    {
      "entropy": 8.613789081573486,
      "epoch": 0.031736872475476054,
      "grad_norm": 1.078125,
      "learning_rate": 0.00013700000000000002,
      "loss": 8.2052,
      "mean_token_accuracy": 0.09102782309055328,
      "num_tokens": 702041.0,
      "step": 275
    },
    {
      "entropy": 8.503752613067627,
      "epoch": 0.03231390652048471,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0001395,
      "loss": 8.2211,
      "mean_token_accuracy": 0.09037618786096573,
      "num_tokens": 714109.0,
      "step": 280
    },
    {
      "entropy": 8.470402336120605,
      "epoch": 0.03289094056549336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00014199999999999998,
      "loss": 8.231,
      "mean_token_accuracy": 0.08331561237573623,
      "num_tokens": 726120.0,
      "step": 285
    },
    {
      "entropy": 8.582913875579834,
      "epoch": 0.03346797461050202,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0001445,
      "loss": 8.1805,
      "mean_token_accuracy": 0.09305408522486687,
      "num_tokens": 737868.0,
      "step": 290
    },
    {
      "entropy": 8.502965545654297,
      "epoch": 0.034045008655510675,
      "grad_norm": 1.03125,
      "learning_rate": 0.000147,
      "loss": 8.2844,
      "mean_token_accuracy": 0.0913321390748024,
      "num_tokens": 751829.0,
      "step": 295
    },
    {
      "entropy": 8.543883037567138,
      "epoch": 0.03462204270051933,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0001495,
      "loss": 8.2192,
      "mean_token_accuracy": 0.09328029677271843,
      "num_tokens": 764457.0,
      "step": 300
    },
    {
      "entropy": 8.415755367279052,
      "epoch": 0.03519907674552799,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000152,
      "loss": 8.2432,
      "mean_token_accuracy": 0.09327497407793998,
      "num_tokens": 776887.0,
      "step": 305
    },
    {
      "entropy": 8.506604480743409,
      "epoch": 0.03577611079053664,
      "grad_norm": 0.921875,
      "learning_rate": 0.00015450000000000001,
      "loss": 8.2075,
      "mean_token_accuracy": 0.09153351783752442,
      "num_tokens": 789001.0,
      "step": 310
    },
    {
      "entropy": 8.496920394897462,
      "epoch": 0.0363531448355453,
      "grad_norm": 0.9375,
      "learning_rate": 0.000157,
      "loss": 8.2141,
      "mean_token_accuracy": 0.09203747734427452,
      "num_tokens": 802719.0,
      "step": 315
    },
    {
      "entropy": 8.46874942779541,
      "epoch": 0.03693017888055395,
      "grad_norm": 0.9375,
      "learning_rate": 0.0001595,
      "loss": 8.1205,
      "mean_token_accuracy": 0.0977991059422493,
      "num_tokens": 815064.0,
      "step": 320
    },
    {
      "entropy": 8.398346996307373,
      "epoch": 0.03750721292556261,
      "grad_norm": 1.625,
      "learning_rate": 0.000162,
      "loss": 8.164,
      "mean_token_accuracy": 0.10131372883915901,
      "num_tokens": 827778.0,
      "step": 325
    },
    {
      "entropy": 8.450510501861572,
      "epoch": 0.038084246970571264,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00016450000000000001,
      "loss": 8.1331,
      "mean_token_accuracy": 0.09271264746785164,
      "num_tokens": 840691.0,
      "step": 330
    },
    {
      "entropy": 8.408138275146484,
      "epoch": 0.03866128101557992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016700000000000002,
      "loss": 8.182,
      "mean_token_accuracy": 0.08595972582697868,
      "num_tokens": 852818.0,
      "step": 335
    },
    {
      "entropy": 8.384876346588134,
      "epoch": 0.03923831506058857,
      "grad_norm": 1.84375,
      "learning_rate": 0.00016950000000000003,
      "loss": 8.177,
      "mean_token_accuracy": 0.10347938761115075,
      "num_tokens": 866854.0,
      "step": 340
    },
    {
      "entropy": 8.501971244812012,
      "epoch": 0.03981534910559723,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00017199999999999998,
      "loss": 8.1456,
      "mean_token_accuracy": 0.09828434884548187,
      "num_tokens": 879234.0,
      "step": 345
    },
    {
      "entropy": 8.459957504272461,
      "epoch": 0.040392383150605886,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00017449999999999999,
      "loss": 8.0968,
      "mean_token_accuracy": 0.09369696900248528,
      "num_tokens": 890927.0,
      "step": 350
    },
    {
      "entropy": 8.416207027435302,
      "epoch": 0.04096941719561454,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000177,
      "loss": 8.1284,
      "mean_token_accuracy": 0.10050402954220772,
      "num_tokens": 902965.0,
      "step": 355
    },
    {
      "entropy": 8.422626209259032,
      "epoch": 0.0415464512406232,
      "grad_norm": 0.8125,
      "learning_rate": 0.0001795,
      "loss": 8.0804,
      "mean_token_accuracy": 0.10290733128786086,
      "num_tokens": 916384.0,
      "step": 360
    },
    {
      "entropy": 8.409742259979248,
      "epoch": 0.042123485285631854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000182,
      "loss": 8.0126,
      "mean_token_accuracy": 0.10525307431817055,
      "num_tokens": 928618.0,
      "step": 365
    },
    {
      "entropy": 8.309245777130126,
      "epoch": 0.04270051933064051,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001845,
      "loss": 8.0938,
      "mean_token_accuracy": 0.0991098240017891,
      "num_tokens": 940087.0,
      "step": 370
    },
    {
      "entropy": 8.369062232971192,
      "epoch": 0.04327755337564916,
      "grad_norm": 1.65625,
      "learning_rate": 0.000187,
      "loss": 8.0771,
      "mean_token_accuracy": 0.10679042860865592,
      "num_tokens": 953322.0,
      "step": 375
    },
    {
      "entropy": 8.470877838134765,
      "epoch": 0.04385458742065782,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0001895,
      "loss": 8.1176,
      "mean_token_accuracy": 0.1005910687148571,
      "num_tokens": 965815.0,
      "step": 380
    },
    {
      "entropy": 8.311385822296142,
      "epoch": 0.044431621465666475,
      "grad_norm": 1.109375,
      "learning_rate": 0.000192,
      "loss": 8.0945,
      "mean_token_accuracy": 0.09785286337137222,
      "num_tokens": 978731.0,
      "step": 385
    },
    {
      "entropy": 8.387943077087403,
      "epoch": 0.04500865551067513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001945,
      "loss": 7.9726,
      "mean_token_accuracy": 0.11110492274165154,
      "num_tokens": 989505.0,
      "step": 390
    },
    {
      "entropy": 8.284615898132325,
      "epoch": 0.04558568955568378,
      "grad_norm": 1.234375,
      "learning_rate": 0.00019700000000000002,
      "loss": 8.0892,
      "mean_token_accuracy": 0.10689176395535469,
      "num_tokens": 1001888.0,
      "step": 395
    },
    {
      "entropy": 8.30747365951538,
      "epoch": 0.04616272360069244,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00019950000000000002,
      "loss": 8.0217,
      "mean_token_accuracy": 0.11540292128920555,
      "num_tokens": 1015362.0,
      "step": 400
    },
    {
      "entropy": 8.404990005493165,
      "epoch": 0.0467397576457011,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000202,
      "loss": 8.1001,
      "mean_token_accuracy": 0.09940713718533516,
      "num_tokens": 1028700.0,
      "step": 405
    },
    {
      "entropy": 8.34985990524292,
      "epoch": 0.04731679169070975,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00020449999999999998,
      "loss": 8.1157,
      "mean_token_accuracy": 0.10158640816807747,
      "num_tokens": 1040766.0,
      "step": 410
    },
    {
      "entropy": 8.311649894714355,
      "epoch": 0.04789382573571841,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000207,
      "loss": 7.9689,
      "mean_token_accuracy": 0.10854383111000061,
      "num_tokens": 1054321.0,
      "step": 415
    },
    {
      "entropy": 8.336063480377197,
      "epoch": 0.048470859780727064,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002095,
      "loss": 8.0275,
      "mean_token_accuracy": 0.11042613610625267,
      "num_tokens": 1065410.0,
      "step": 420
    },
    {
      "entropy": 8.372344303131104,
      "epoch": 0.04904789382573572,
      "grad_norm": 1.203125,
      "learning_rate": 0.000212,
      "loss": 8.0589,
      "mean_token_accuracy": 0.09765132367610932,
      "num_tokens": 1077369.0,
      "step": 425
    },
    {
      "entropy": 8.32605962753296,
      "epoch": 0.04962492787074437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002145,
      "loss": 8.0076,
      "mean_token_accuracy": 0.10854156166315079,
      "num_tokens": 1089065.0,
      "step": 430
    },
    {
      "entropy": 8.241904735565186,
      "epoch": 0.05020196191575303,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.9173,
      "mean_token_accuracy": 0.11676707863807678,
      "num_tokens": 1101285.0,
      "step": 435
    },
    {
      "entropy": 8.32933588027954,
      "epoch": 0.050778995960761686,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002195,
      "loss": 7.9379,
      "mean_token_accuracy": 0.10726014450192452,
      "num_tokens": 1114001.0,
      "step": 440
    },
    {
      "entropy": 8.260017490386963,
      "epoch": 0.05135603000577034,
      "grad_norm": 0.890625,
      "learning_rate": 0.000222,
      "loss": 8.0005,
      "mean_token_accuracy": 0.1092146560549736,
      "num_tokens": 1126751.0,
      "step": 445
    },
    {
      "entropy": 8.230885124206543,
      "epoch": 0.05193306405077899,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002245,
      "loss": 7.991,
      "mean_token_accuracy": 0.10859929993748665,
      "num_tokens": 1140737.0,
      "step": 450
    },
    {
      "entropy": 8.258237075805663,
      "epoch": 0.052510098095787654,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.9376,
      "mean_token_accuracy": 0.11085548251867294,
      "num_tokens": 1152569.0,
      "step": 455
    },
    {
      "entropy": 8.315269184112548,
      "epoch": 0.05308713214079631,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.9535,
      "mean_token_accuracy": 0.11076531112194062,
      "num_tokens": 1166163.0,
      "step": 460
    },
    {
      "entropy": 8.270787239074707,
      "epoch": 0.05366416618580496,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.9184,
      "mean_token_accuracy": 0.11104324012994767,
      "num_tokens": 1177647.0,
      "step": 465
    },
    {
      "entropy": 8.256643295288086,
      "epoch": 0.054241200230813615,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.9545,
      "mean_token_accuracy": 0.11232023686170578,
      "num_tokens": 1190884.0,
      "step": 470
    },
    {
      "entropy": 8.198192024230957,
      "epoch": 0.054818234275822275,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000237,
      "loss": 7.9617,
      "mean_token_accuracy": 0.11213421821594238,
      "num_tokens": 1203276.0,
      "step": 475
    },
    {
      "entropy": 8.210485458374023,
      "epoch": 0.05539526832083093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002395,
      "loss": 7.8697,
      "mean_token_accuracy": 0.12074657827615738,
      "num_tokens": 1216001.0,
      "step": 480
    },
    {
      "entropy": 8.263874912261963,
      "epoch": 0.05597230236583958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000242,
      "loss": 7.971,
      "mean_token_accuracy": 0.11633578017354011,
      "num_tokens": 1228724.0,
      "step": 485
    },
    {
      "entropy": 8.196341419219971,
      "epoch": 0.05654933641084824,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002445,
      "loss": 7.9923,
      "mean_token_accuracy": 0.11399635374546051,
      "num_tokens": 1242731.0,
      "step": 490
    },
    {
      "entropy": 8.200312662124634,
      "epoch": 0.0571263704558569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000247,
      "loss": 7.9326,
      "mean_token_accuracy": 0.11518245413899422,
      "num_tokens": 1255275.0,
      "step": 495
    },
    {
      "entropy": 8.26755723953247,
      "epoch": 0.05770340450086555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002495,
      "loss": 7.9145,
      "mean_token_accuracy": 0.12056171074509621,
      "num_tokens": 1267719.0,
      "step": 500
    },
    {
      "entropy": 8.105032253265382,
      "epoch": 0.058280438545874204,
      "grad_norm": 1.234375,
      "learning_rate": 0.000252,
      "loss": 7.9041,
      "mean_token_accuracy": 0.11403353810310364,
      "num_tokens": 1279960.0,
      "step": 505
    },
    {
      "entropy": 8.281104183197021,
      "epoch": 0.058857472590882864,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002545,
      "loss": 7.8882,
      "mean_token_accuracy": 0.11356199756264687,
      "num_tokens": 1292570.0,
      "step": 510
    },
    {
      "entropy": 8.16637954711914,
      "epoch": 0.05943450663589152,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000257,
      "loss": 7.8851,
      "mean_token_accuracy": 0.11174787804484368,
      "num_tokens": 1305548.0,
      "step": 515
    },
    {
      "entropy": 8.07060809135437,
      "epoch": 0.06001154068090017,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002595,
      "loss": 7.828,
      "mean_token_accuracy": 0.12017364054918289,
      "num_tokens": 1318904.0,
      "step": 520
    },
    {
      "entropy": 8.264026546478272,
      "epoch": 0.060588574725908825,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000262,
      "loss": 7.8875,
      "mean_token_accuracy": 0.11706614196300506,
      "num_tokens": 1332841.0,
      "step": 525
    },
    {
      "entropy": 8.177509117126466,
      "epoch": 0.061165608770917486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.8888,
      "mean_token_accuracy": 0.11371787264943123,
      "num_tokens": 1345708.0,
      "step": 530
    },
    {
      "entropy": 8.115266609191895,
      "epoch": 0.06174264281592614,
      "grad_norm": 1.25,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.8416,
      "mean_token_accuracy": 0.12110820636153222,
      "num_tokens": 1358611.0,
      "step": 535
    },
    {
      "entropy": 8.071786689758301,
      "epoch": 0.06231967686093479,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.8965,
      "mean_token_accuracy": 0.11819371730089187,
      "num_tokens": 1372753.0,
      "step": 540
    },
    {
      "entropy": 8.169945526123048,
      "epoch": 0.06289671090594345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.8013,
      "mean_token_accuracy": 0.12582628801465034,
      "num_tokens": 1383811.0,
      "step": 545
    },
    {
      "entropy": 8.11972336769104,
      "epoch": 0.06347374495095211,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002745,
      "loss": 7.8802,
      "mean_token_accuracy": 0.11965988725423812,
      "num_tokens": 1395228.0,
      "step": 550
    },
    {
      "entropy": 8.023060703277588,
      "epoch": 0.06405077899596076,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000277,
      "loss": 7.8824,
      "mean_token_accuracy": 0.11617207005620003,
      "num_tokens": 1408869.0,
      "step": 555
    },
    {
      "entropy": 8.129542350769043,
      "epoch": 0.06462781304096941,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002795,
      "loss": 7.7987,
      "mean_token_accuracy": 0.12237567305564881,
      "num_tokens": 1420926.0,
      "step": 560
    },
    {
      "entropy": 8.091229438781738,
      "epoch": 0.06520484708597807,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.7791,
      "mean_token_accuracy": 0.12508851885795594,
      "num_tokens": 1432435.0,
      "step": 565
    },
    {
      "entropy": 8.181858110427857,
      "epoch": 0.06578188113098672,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002845,
      "loss": 7.9104,
      "mean_token_accuracy": 0.11593875512480736,
      "num_tokens": 1444716.0,
      "step": 570
    },
    {
      "entropy": 8.023220300674438,
      "epoch": 0.06635891517599539,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000287,
      "loss": 7.8024,
      "mean_token_accuracy": 0.12172861248254777,
      "num_tokens": 1458129.0,
      "step": 575
    },
    {
      "entropy": 8.007987642288208,
      "epoch": 0.06693594922100404,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002895,
      "loss": 7.7706,
      "mean_token_accuracy": 0.1253301203250885,
      "num_tokens": 1472076.0,
      "step": 580
    },
    {
      "entropy": 8.011878776550294,
      "epoch": 0.0675129832660127,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000292,
      "loss": 7.7542,
      "mean_token_accuracy": 0.13274379968643188,
      "num_tokens": 1486026.0,
      "step": 585
    },
    {
      "entropy": 8.01979365348816,
      "epoch": 0.06809001731102135,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002945,
      "loss": 7.772,
      "mean_token_accuracy": 0.13351956158876419,
      "num_tokens": 1498735.0,
      "step": 590
    },
    {
      "entropy": 8.010276079177856,
      "epoch": 0.06866705135603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000297,
      "loss": 7.7535,
      "mean_token_accuracy": 0.1248227261006832,
      "num_tokens": 1510755.0,
      "step": 595
    },
    {
      "entropy": 8.088212156295777,
      "epoch": 0.06924408540103866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002995,
      "loss": 7.7748,
      "mean_token_accuracy": 0.12387159541249275,
      "num_tokens": 1522386.0,
      "step": 600
    },
    {
      "entropy": 8.043047952651978,
      "epoch": 0.06982111944604731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000302,
      "loss": 7.8151,
      "mean_token_accuracy": 0.12131175622344018,
      "num_tokens": 1534757.0,
      "step": 605
    },
    {
      "entropy": 8.028788948059082,
      "epoch": 0.07039815349105598,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003045,
      "loss": 7.7821,
      "mean_token_accuracy": 0.12405301108956338,
      "num_tokens": 1547684.0,
      "step": 610
    },
    {
      "entropy": 8.01380696296692,
      "epoch": 0.07097518753606463,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000307,
      "loss": 7.7925,
      "mean_token_accuracy": 0.11749588996171952,
      "num_tokens": 1560583.0,
      "step": 615
    },
    {
      "entropy": 8.04894208908081,
      "epoch": 0.07155222158107329,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003095,
      "loss": 7.787,
      "mean_token_accuracy": 0.11909917891025543,
      "num_tokens": 1573174.0,
      "step": 620
    },
    {
      "entropy": 8.029009103775024,
      "epoch": 0.07212925562608194,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000312,
      "loss": 7.7789,
      "mean_token_accuracy": 0.12413554638624191,
      "num_tokens": 1585746.0,
      "step": 625
    },
    {
      "entropy": 8.037794351577759,
      "epoch": 0.0727062896710906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003145,
      "loss": 7.7608,
      "mean_token_accuracy": 0.12169933542609215,
      "num_tokens": 1598538.0,
      "step": 630
    },
    {
      "entropy": 7.993807029724121,
      "epoch": 0.07328332371609925,
      "grad_norm": 0.890625,
      "learning_rate": 0.000317,
      "loss": 7.7304,
      "mean_token_accuracy": 0.12405971959233283,
      "num_tokens": 1612598.0,
      "step": 635
    },
    {
      "entropy": 7.991437578201294,
      "epoch": 0.0738603577611079,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003195,
      "loss": 7.6674,
      "mean_token_accuracy": 0.12977527305483819,
      "num_tokens": 1624418.0,
      "step": 640
    },
    {
      "entropy": 7.903521156311035,
      "epoch": 0.07443739180611655,
      "grad_norm": 0.953125,
      "learning_rate": 0.000322,
      "loss": 7.7507,
      "mean_token_accuracy": 0.13144003003835678,
      "num_tokens": 1637512.0,
      "step": 645
    },
    {
      "entropy": 7.92955265045166,
      "epoch": 0.07501442585112522,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.6976,
      "mean_token_accuracy": 0.12809038534760475,
      "num_tokens": 1650413.0,
      "step": 650
    },
    {
      "entropy": 7.944208335876465,
      "epoch": 0.07559145989613388,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.6683,
      "mean_token_accuracy": 0.12944014593958855,
      "num_tokens": 1662225.0,
      "step": 655
    },
    {
      "entropy": 7.995383930206299,
      "epoch": 0.07616849394114253,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.7236,
      "mean_token_accuracy": 0.12496703863143921,
      "num_tokens": 1674458.0,
      "step": 660
    },
    {
      "entropy": 7.944713544845581,
      "epoch": 0.07674552798615118,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.7476,
      "mean_token_accuracy": 0.12143125832080841,
      "num_tokens": 1686366.0,
      "step": 665
    },
    {
      "entropy": 8.000255727767945,
      "epoch": 0.07732256203115984,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.7442,
      "mean_token_accuracy": 0.12491990998387337,
      "num_tokens": 1699654.0,
      "step": 670
    },
    {
      "entropy": 7.9440826892852785,
      "epoch": 0.07789959607616849,
      "grad_norm": 1.09375,
      "learning_rate": 0.000337,
      "loss": 7.6926,
      "mean_token_accuracy": 0.12900558561086656,
      "num_tokens": 1711923.0,
      "step": 675
    },
    {
      "entropy": 7.972041606903076,
      "epoch": 0.07847663012117714,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003395,
      "loss": 7.6978,
      "mean_token_accuracy": 0.12197642624378205,
      "num_tokens": 1723413.0,
      "step": 680
    },
    {
      "entropy": 7.880360507965088,
      "epoch": 0.07905366416618581,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000342,
      "loss": 7.6585,
      "mean_token_accuracy": 0.12834955081343652,
      "num_tokens": 1735975.0,
      "step": 685
    },
    {
      "entropy": 7.841808462142945,
      "epoch": 0.07963069821119446,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.6589,
      "mean_token_accuracy": 0.12771303430199624,
      "num_tokens": 1748563.0,
      "step": 690
    },
    {
      "entropy": 7.920938873291016,
      "epoch": 0.08020773225620312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000347,
      "loss": 7.685,
      "mean_token_accuracy": 0.1225068636238575,
      "num_tokens": 1762914.0,
      "step": 695
    },
    {
      "entropy": 7.946118593215942,
      "epoch": 0.08078476630121177,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003495,
      "loss": 7.6416,
      "mean_token_accuracy": 0.13206669613718985,
      "num_tokens": 1776084.0,
      "step": 700
    },
    {
      "entropy": 7.8522233963012695,
      "epoch": 0.08136180034622043,
      "grad_norm": 1.109375,
      "learning_rate": 0.000352,
      "loss": 7.6597,
      "mean_token_accuracy": 0.12902271673083304,
      "num_tokens": 1789316.0,
      "step": 705
    },
    {
      "entropy": 7.830107116699219,
      "epoch": 0.08193883439122908,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003545,
      "loss": 7.6724,
      "mean_token_accuracy": 0.1287533812224865,
      "num_tokens": 1801576.0,
      "step": 710
    },
    {
      "entropy": 7.857487392425537,
      "epoch": 0.08251586843623773,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000357,
      "loss": 7.6286,
      "mean_token_accuracy": 0.12918446362018585,
      "num_tokens": 1814039.0,
      "step": 715
    },
    {
      "entropy": 7.842514276504517,
      "epoch": 0.0830929024812464,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003595,
      "loss": 7.5683,
      "mean_token_accuracy": 0.13093318343162536,
      "num_tokens": 1825193.0,
      "step": 720
    },
    {
      "entropy": 7.7918308734893795,
      "epoch": 0.08366993652625505,
      "grad_norm": 1.125,
      "learning_rate": 0.000362,
      "loss": 7.653,
      "mean_token_accuracy": 0.1273672729730606,
      "num_tokens": 1837983.0,
      "step": 725
    },
    {
      "entropy": 7.921112680435181,
      "epoch": 0.08424697057126371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003645,
      "loss": 7.5584,
      "mean_token_accuracy": 0.13639293611049652,
      "num_tokens": 1849830.0,
      "step": 730
    },
    {
      "entropy": 7.83430438041687,
      "epoch": 0.08482400461627236,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000367,
      "loss": 7.6309,
      "mean_token_accuracy": 0.1251987598836422,
      "num_tokens": 1862577.0,
      "step": 735
    },
    {
      "entropy": 7.849653959274292,
      "epoch": 0.08540103866128101,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003695,
      "loss": 7.6367,
      "mean_token_accuracy": 0.13423554003238677,
      "num_tokens": 1874888.0,
      "step": 740
    },
    {
      "entropy": 7.746623182296753,
      "epoch": 0.08597807270628967,
      "grad_norm": 1.609375,
      "learning_rate": 0.000372,
      "loss": 7.5754,
      "mean_token_accuracy": 0.13472723960876465,
      "num_tokens": 1888400.0,
      "step": 745
    },
    {
      "entropy": 7.872568368911743,
      "epoch": 0.08655510675129832,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003745,
      "loss": 7.6565,
      "mean_token_accuracy": 0.12168999016284943,
      "num_tokens": 1900961.0,
      "step": 750
    },
    {
      "entropy": 7.816582107543946,
      "epoch": 0.08713214079630698,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000377,
      "loss": 7.5822,
      "mean_token_accuracy": 0.1359177239239216,
      "num_tokens": 1913410.0,
      "step": 755
    },
    {
      "entropy": 7.774113464355469,
      "epoch": 0.08770917484131564,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003795,
      "loss": 7.632,
      "mean_token_accuracy": 0.12596142962574958,
      "num_tokens": 1926602.0,
      "step": 760
    },
    {
      "entropy": 7.874756813049316,
      "epoch": 0.0882862088863243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000382,
      "loss": 7.6001,
      "mean_token_accuracy": 0.1248251162469387,
      "num_tokens": 1938904.0,
      "step": 765
    },
    {
      "entropy": 7.8404929637908936,
      "epoch": 0.08886324293133295,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003845,
      "loss": 7.6437,
      "mean_token_accuracy": 0.12104213982820511,
      "num_tokens": 1952827.0,
      "step": 770
    },
    {
      "entropy": 7.740968370437622,
      "epoch": 0.0894402769763416,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.6081,
      "mean_token_accuracy": 0.13230291232466698,
      "num_tokens": 1965182.0,
      "step": 775
    },
    {
      "entropy": 7.757106733322144,
      "epoch": 0.09001731102135026,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.5679,
      "mean_token_accuracy": 0.12761696726083754,
      "num_tokens": 1976953.0,
      "step": 780
    },
    {
      "entropy": 7.908223438262939,
      "epoch": 0.09059434506635891,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.6523,
      "mean_token_accuracy": 0.12661396712064743,
      "num_tokens": 1990019.0,
      "step": 785
    },
    {
      "entropy": 7.842516183853149,
      "epoch": 0.09117137911136756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.5967,
      "mean_token_accuracy": 0.1285574607551098,
      "num_tokens": 2002976.0,
      "step": 790
    },
    {
      "entropy": 7.783304500579834,
      "epoch": 0.09174841315637623,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.606,
      "mean_token_accuracy": 0.12812672331929206,
      "num_tokens": 2017368.0,
      "step": 795
    },
    {
      "entropy": 7.74679274559021,
      "epoch": 0.09232544720138489,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003995,
      "loss": 7.4697,
      "mean_token_accuracy": 0.13386220484972,
      "num_tokens": 2029838.0,
      "step": 800
    },
    {
      "entropy": 7.811276197433472,
      "epoch": 0.09290248124639354,
      "grad_norm": 1.015625,
      "learning_rate": 0.000402,
      "loss": 7.5942,
      "mean_token_accuracy": 0.12542516514658927,
      "num_tokens": 2042085.0,
      "step": 805
    },
    {
      "entropy": 7.784253311157227,
      "epoch": 0.0934795152914022,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004045,
      "loss": 7.6123,
      "mean_token_accuracy": 0.12910280898213386,
      "num_tokens": 2053268.0,
      "step": 810
    },
    {
      "entropy": 7.732159948348999,
      "epoch": 0.09405654933641085,
      "grad_norm": 1.0,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.5886,
      "mean_token_accuracy": 0.12081153169274331,
      "num_tokens": 2066621.0,
      "step": 815
    },
    {
      "entropy": 7.798933219909668,
      "epoch": 0.0946335833814195,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004095,
      "loss": 7.4894,
      "mean_token_accuracy": 0.12757847532629968,
      "num_tokens": 2079590.0,
      "step": 820
    },
    {
      "entropy": 7.734460258483887,
      "epoch": 0.09521061742642815,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412,
      "loss": 7.5833,
      "mean_token_accuracy": 0.12265915721654892,
      "num_tokens": 2091729.0,
      "step": 825
    },
    {
      "entropy": 7.6462846279144285,
      "epoch": 0.09578765147143682,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004145,
      "loss": 7.547,
      "mean_token_accuracy": 0.1345345124602318,
      "num_tokens": 2104254.0,
      "step": 830
    },
    {
      "entropy": 7.8504674434661865,
      "epoch": 0.09636468551644548,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000417,
      "loss": 7.6223,
      "mean_token_accuracy": 0.12215886563062668,
      "num_tokens": 2116327.0,
      "step": 835
    },
    {
      "entropy": 7.772087621688843,
      "epoch": 0.09694171956145413,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004195,
      "loss": 7.5887,
      "mean_token_accuracy": 0.12261896207928658,
      "num_tokens": 2129153.0,
      "step": 840
    },
    {
      "entropy": 7.678793621063233,
      "epoch": 0.09751875360646278,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000422,
      "loss": 7.5453,
      "mean_token_accuracy": 0.13327833712100984,
      "num_tokens": 2141705.0,
      "step": 845
    },
    {
      "entropy": 7.719631814956665,
      "epoch": 0.09809578765147144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004245,
      "loss": 7.4725,
      "mean_token_accuracy": 0.13811284452676773,
      "num_tokens": 2153760.0,
      "step": 850
    },
    {
      "entropy": 7.7237812042236325,
      "epoch": 0.09867282169648009,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000427,
      "loss": 7.5732,
      "mean_token_accuracy": 0.13227060809731483,
      "num_tokens": 2168206.0,
      "step": 855
    },
    {
      "entropy": 7.767458248138428,
      "epoch": 0.09924985574148874,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004295,
      "loss": 7.5144,
      "mean_token_accuracy": 0.13358384743332863,
      "num_tokens": 2180779.0,
      "step": 860
    },
    {
      "entropy": 7.586296081542969,
      "epoch": 0.0998268897864974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000432,
      "loss": 7.4646,
      "mean_token_accuracy": 0.1356838047504425,
      "num_tokens": 2194075.0,
      "step": 865
    },
    {
      "entropy": 7.7125969409942625,
      "epoch": 0.10040392383150606,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004345,
      "loss": 7.4718,
      "mean_token_accuracy": 0.13414136543869973,
      "num_tokens": 2206424.0,
      "step": 870
    },
    {
      "entropy": 7.709120893478394,
      "epoch": 0.10098095787651472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000437,
      "loss": 7.516,
      "mean_token_accuracy": 0.12676840350031854,
      "num_tokens": 2218550.0,
      "step": 875
    },
    {
      "entropy": 7.638567209243774,
      "epoch": 0.10155799192152337,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004395,
      "loss": 7.5389,
      "mean_token_accuracy": 0.13040511906147004,
      "num_tokens": 2231857.0,
      "step": 880
    },
    {
      "entropy": 7.743336486816406,
      "epoch": 0.10213502596653203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000442,
      "loss": 7.5447,
      "mean_token_accuracy": 0.13032453507184982,
      "num_tokens": 2244057.0,
      "step": 885
    },
    {
      "entropy": 7.807929277420044,
      "epoch": 0.10271206001154068,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004445,
      "loss": 7.5759,
      "mean_token_accuracy": 0.1243269257247448,
      "num_tokens": 2257361.0,
      "step": 890
    },
    {
      "entropy": 7.6737877368927006,
      "epoch": 0.10328909405654933,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000447,
      "loss": 7.5513,
      "mean_token_accuracy": 0.1263903483748436,
      "num_tokens": 2269599.0,
      "step": 895
    },
    {
      "entropy": 7.745085287094116,
      "epoch": 0.10386612810155799,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.576,
      "mean_token_accuracy": 0.1303069844841957,
      "num_tokens": 2282791.0,
      "step": 900
    },
    {
      "entropy": 7.7095636367797855,
      "epoch": 0.10444316214656665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.5207,
      "mean_token_accuracy": 0.13242215067148208,
      "num_tokens": 2295336.0,
      "step": 905
    },
    {
      "entropy": 7.752670526504517,
      "epoch": 0.10502019619157531,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.5386,
      "mean_token_accuracy": 0.12832471057772638,
      "num_tokens": 2308813.0,
      "step": 910
    },
    {
      "entropy": 7.666492176055908,
      "epoch": 0.10559723023658396,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.5173,
      "mean_token_accuracy": 0.12629655599594117,
      "num_tokens": 2321374.0,
      "step": 915
    },
    {
      "entropy": 7.706280326843261,
      "epoch": 0.10617426428159261,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.5314,
      "mean_token_accuracy": 0.12404570654034615,
      "num_tokens": 2333698.0,
      "step": 920
    },
    {
      "entropy": 7.732450342178344,
      "epoch": 0.10675129832660127,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000462,
      "loss": 7.5129,
      "mean_token_accuracy": 0.13543608784675598,
      "num_tokens": 2346557.0,
      "step": 925
    },
    {
      "entropy": 7.497116231918335,
      "epoch": 0.10732833237160992,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004645,
      "loss": 7.3586,
      "mean_token_accuracy": 0.13911109417676926,
      "num_tokens": 2360003.0,
      "step": 930
    },
    {
      "entropy": 7.704750299453735,
      "epoch": 0.10790536641661858,
      "grad_norm": 1.046875,
      "learning_rate": 0.000467,
      "loss": 7.4169,
      "mean_token_accuracy": 0.13651928678154945,
      "num_tokens": 2372765.0,
      "step": 935
    },
    {
      "entropy": 7.554190254211425,
      "epoch": 0.10848240046162723,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004695,
      "loss": 7.4229,
      "mean_token_accuracy": 0.13624904453754424,
      "num_tokens": 2385154.0,
      "step": 940
    },
    {
      "entropy": 7.579919815063477,
      "epoch": 0.1090594345066359,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472,
      "loss": 7.4283,
      "mean_token_accuracy": 0.13729820549488067,
      "num_tokens": 2397264.0,
      "step": 945
    },
    {
      "entropy": 7.707980585098267,
      "epoch": 0.10963646855164455,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004745,
      "loss": 7.5236,
      "mean_token_accuracy": 0.1310661919414997,
      "num_tokens": 2408652.0,
      "step": 950
    },
    {
      "entropy": 7.583171081542969,
      "epoch": 0.1102135025966532,
      "grad_norm": 0.953125,
      "learning_rate": 0.000477,
      "loss": 7.4921,
      "mean_token_accuracy": 0.12758150473237037,
      "num_tokens": 2421683.0,
      "step": 955
    },
    {
      "entropy": 7.651930141448974,
      "epoch": 0.11079053664166186,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004795,
      "loss": 7.4028,
      "mean_token_accuracy": 0.13190027326345444,
      "num_tokens": 2434622.0,
      "step": 960
    },
    {
      "entropy": 7.659194564819336,
      "epoch": 0.11136757068667051,
      "grad_norm": 1.015625,
      "learning_rate": 0.000482,
      "loss": 7.4263,
      "mean_token_accuracy": 0.14118838533759118,
      "num_tokens": 2448157.0,
      "step": 965
    },
    {
      "entropy": 7.696449565887451,
      "epoch": 0.11194460473167916,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845,
      "loss": 7.4789,
      "mean_token_accuracy": 0.12850709185004233,
      "num_tokens": 2460284.0,
      "step": 970
    },
    {
      "entropy": 7.52245602607727,
      "epoch": 0.11252163877668782,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000487,
      "loss": 7.476,
      "mean_token_accuracy": 0.1310298591852188,
      "num_tokens": 2473031.0,
      "step": 975
    },
    {
      "entropy": 7.670017910003662,
      "epoch": 0.11309867282169649,
      "grad_norm": 0.875,
      "learning_rate": 0.0004895,
      "loss": 7.4819,
      "mean_token_accuracy": 0.13639922589063644,
      "num_tokens": 2486870.0,
      "step": 980
    },
    {
      "entropy": 7.584562635421753,
      "epoch": 0.11367570686670514,
      "grad_norm": 1.0,
      "learning_rate": 0.000492,
      "loss": 7.3807,
      "mean_token_accuracy": 0.13501427620649337,
      "num_tokens": 2498921.0,
      "step": 985
    },
    {
      "entropy": 7.616317844390869,
      "epoch": 0.1142527409117138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945,
      "loss": 7.4674,
      "mean_token_accuracy": 0.12648973166942595,
      "num_tokens": 2511122.0,
      "step": 990
    },
    {
      "entropy": 7.670611572265625,
      "epoch": 0.11482977495672245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497,
      "loss": 7.4794,
      "mean_token_accuracy": 0.1284182198345661,
      "num_tokens": 2523726.0,
      "step": 995
    },
    {
      "entropy": 7.538252162933349,
      "epoch": 0.1154068090017311,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995,
      "loss": 7.3524,
      "mean_token_accuracy": 0.13616370856761933,
      "num_tokens": 2536192.0,
      "step": 1000
    },
    {
      "entropy": 7.569291305541992,
      "epoch": 0.11598384304673975,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999999975783157,
      "loss": 7.4319,
      "mean_token_accuracy": 0.12918952256441116,
      "num_tokens": 2550200.0,
      "step": 1005
    },
    {
      "entropy": 7.612759971618653,
      "epoch": 0.11656087709174841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999999877402236,
      "loss": 7.3793,
      "mean_token_accuracy": 0.133512731641531,
      "num_tokens": 2564045.0,
      "step": 1010
    },
    {
      "entropy": 7.557199764251709,
      "epoch": 0.11713791113675708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999703343686,
      "loss": 7.3972,
      "mean_token_accuracy": 0.12909132093191147,
      "num_tokens": 2576210.0,
      "step": 1015
    },
    {
      "entropy": 7.435172700881958,
      "epoch": 0.11771494518176573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999999453607515,
      "loss": 7.378,
      "mean_token_accuracy": 0.1307467371225357,
      "num_tokens": 2589325.0,
      "step": 1020
    },
    {
      "entropy": 7.680232429504395,
      "epoch": 0.11829197922677438,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999128193729,
      "loss": 7.4734,
      "mean_token_accuracy": 0.13209090158343315,
      "num_tokens": 2601522.0,
      "step": 1025
    },
    {
      "entropy": 7.629494333267212,
      "epoch": 0.11886901327178304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499999872710234,
      "loss": 7.4398,
      "mean_token_accuracy": 0.1303465448319912,
      "num_tokens": 2615364.0,
      "step": 1030
    },
    {
      "entropy": 7.538928461074829,
      "epoch": 0.11944604731679169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999998250333361,
      "loss": 7.3623,
      "mean_token_accuracy": 0.13469313830137253,
      "num_tokens": 2628528.0,
      "step": 1035
    },
    {
      "entropy": 7.55066442489624,
      "epoch": 0.12002308136180034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499999769788681,
      "loss": 7.3676,
      "mean_token_accuracy": 0.13092967718839646,
      "num_tokens": 2641877.0,
      "step": 1040
    },
    {
      "entropy": 7.518403100967407,
      "epoch": 0.120600115406809,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999997069762703,
      "loss": 7.3645,
      "mean_token_accuracy": 0.1341833934187889,
      "num_tokens": 2655419.0,
      "step": 1045
    },
    {
      "entropy": 7.514628076553345,
      "epoch": 0.12117714945181765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999996365961062,
      "loss": 7.3271,
      "mean_token_accuracy": 0.1362943857908249,
      "num_tokens": 2667551.0,
      "step": 1050
    },
    {
      "entropy": 7.557382869720459,
      "epoch": 0.12175418349682632,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999995586481912,
      "loss": 7.3534,
      "mean_token_accuracy": 0.13944169133901596,
      "num_tokens": 2679948.0,
      "step": 1055
    },
    {
      "entropy": 7.471993780136108,
      "epoch": 0.12233121754183497,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999994731325276,
      "loss": 7.3972,
      "mean_token_accuracy": 0.1320507399737835,
      "num_tokens": 2691501.0,
      "step": 1060
    },
    {
      "entropy": 7.50092716217041,
      "epoch": 0.12290825158684363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999993800491187,
      "loss": 7.3195,
      "mean_token_accuracy": 0.1353978231549263,
      "num_tokens": 2702689.0,
      "step": 1065
    },
    {
      "entropy": 7.551953458786011,
      "epoch": 0.12348528563185228,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999992793979672,
      "loss": 7.3338,
      "mean_token_accuracy": 0.1352718137204647,
      "num_tokens": 2715023.0,
      "step": 1070
    },
    {
      "entropy": 7.482572031021118,
      "epoch": 0.12406231967686093,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999991711790769,
      "loss": 7.25,
      "mean_token_accuracy": 0.13727134317159653,
      "num_tokens": 2729517.0,
      "step": 1075
    },
    {
      "entropy": 7.496352291107177,
      "epoch": 0.12463935372186959,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999990553924511,
      "loss": 7.3561,
      "mean_token_accuracy": 0.13238862976431848,
      "num_tokens": 2740984.0,
      "step": 1080
    },
    {
      "entropy": 7.511165189743042,
      "epoch": 0.12521638776687824,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999989320380939,
      "loss": 7.3101,
      "mean_token_accuracy": 0.1363685481250286,
      "num_tokens": 2752943.0,
      "step": 1085
    },
    {
      "entropy": 7.534621715545654,
      "epoch": 0.1257934218118869,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004999988011160094,
      "loss": 7.3589,
      "mean_token_accuracy": 0.13417960107326507,
      "num_tokens": 2765613.0,
      "step": 1090
    },
    {
      "entropy": 7.450796127319336,
      "epoch": 0.12637045585689555,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499998662626202,
      "loss": 7.2979,
      "mean_token_accuracy": 0.1321871429681778,
      "num_tokens": 2776820.0,
      "step": 1095
    },
    {
      "entropy": 7.463679599761963,
      "epoch": 0.12694748990190421,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999985165686764,
      "loss": 7.3429,
      "mean_token_accuracy": 0.13456639721989633,
      "num_tokens": 2788981.0,
      "step": 1100
    },
    {
      "entropy": 7.459660530090332,
      "epoch": 0.12752452394691285,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999983629434373,
      "loss": 7.1963,
      "mean_token_accuracy": 0.1402936853468418,
      "num_tokens": 2801975.0,
      "step": 1105
    },
    {
      "entropy": 7.450168704986572,
      "epoch": 0.12810155799192152,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999982017504901,
      "loss": 7.2817,
      "mean_token_accuracy": 0.1321284957230091,
      "num_tokens": 2814708.0,
      "step": 1110
    },
    {
      "entropy": 7.456075811386109,
      "epoch": 0.1286785920369302,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999980329898401,
      "loss": 7.302,
      "mean_token_accuracy": 0.13727107644081116,
      "num_tokens": 2827903.0,
      "step": 1115
    },
    {
      "entropy": 7.500937414169312,
      "epoch": 0.12925562608193883,
      "grad_norm": 1.0,
      "learning_rate": 0.000499997856661493,
      "loss": 7.3277,
      "mean_token_accuracy": 0.13589581474661827,
      "num_tokens": 2841077.0,
      "step": 1120
    },
    {
      "entropy": 7.48737940788269,
      "epoch": 0.1298326601269475,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499997672765455,
      "loss": 7.3185,
      "mean_token_accuracy": 0.13034305721521378,
      "num_tokens": 2853235.0,
      "step": 1125
    },
    {
      "entropy": 7.418439817428589,
      "epoch": 0.13040969417195614,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999974813017318,
      "loss": 7.288,
      "mean_token_accuracy": 0.1318283796310425,
      "num_tokens": 2864937.0,
      "step": 1130
    },
    {
      "entropy": 7.438732004165649,
      "epoch": 0.1309867282169648,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999972822703301,
      "loss": 7.2431,
      "mean_token_accuracy": 0.14729652926325798,
      "num_tokens": 2877791.0,
      "step": 1135
    },
    {
      "entropy": 7.374364995956421,
      "epoch": 0.13156376226197344,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999970756712567,
      "loss": 7.2388,
      "mean_token_accuracy": 0.13121830448508262,
      "num_tokens": 2889897.0,
      "step": 1140
    },
    {
      "entropy": 7.502627754211426,
      "epoch": 0.1321407963069821,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999968615045183,
      "loss": 7.2464,
      "mean_token_accuracy": 0.13865444660186768,
      "num_tokens": 2902806.0,
      "step": 1145
    },
    {
      "entropy": 7.364279270172119,
      "epoch": 0.13271783035199078,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999966397701222,
      "loss": 7.1478,
      "mean_token_accuracy": 0.14386670887470246,
      "num_tokens": 2915703.0,
      "step": 1150
    },
    {
      "entropy": 7.446257495880127,
      "epoch": 0.13329486439699942,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999964104680759,
      "loss": 7.3489,
      "mean_token_accuracy": 0.12771031260490417,
      "num_tokens": 2928171.0,
      "step": 1155
    },
    {
      "entropy": 7.4577220439910885,
      "epoch": 0.13387189844200809,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999961735983871,
      "loss": 7.1758,
      "mean_token_accuracy": 0.14841930717229843,
      "num_tokens": 2939909.0,
      "step": 1160
    },
    {
      "entropy": 7.482584476470947,
      "epoch": 0.13444893248701673,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999959291610639,
      "loss": 7.3014,
      "mean_token_accuracy": 0.13288074284791945,
      "num_tokens": 2951510.0,
      "step": 1165
    },
    {
      "entropy": 7.388064241409301,
      "epoch": 0.1350259665320254,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999956771561143,
      "loss": 7.2356,
      "mean_token_accuracy": 0.1302425242960453,
      "num_tokens": 2963326.0,
      "step": 1170
    },
    {
      "entropy": 7.376373052597046,
      "epoch": 0.13560300057703403,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999954175835469,
      "loss": 7.2405,
      "mean_token_accuracy": 0.14472833275794983,
      "num_tokens": 2976226.0,
      "step": 1175
    },
    {
      "entropy": 7.442576217651367,
      "epoch": 0.1361800346220427,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999951504433703,
      "loss": 7.255,
      "mean_token_accuracy": 0.13099167943000795,
      "num_tokens": 2988180.0,
      "step": 1180
    },
    {
      "entropy": 7.3849358558654785,
      "epoch": 0.13675706866705137,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999948757355935,
      "loss": 7.2402,
      "mean_token_accuracy": 0.13745544701814652,
      "num_tokens": 3000215.0,
      "step": 1185
    },
    {
      "entropy": 7.460414743423462,
      "epoch": 0.13733410271206,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499994593460226,
      "loss": 7.2419,
      "mean_token_accuracy": 0.13573380932211876,
      "num_tokens": 3012516.0,
      "step": 1190
    },
    {
      "entropy": 7.335642337799072,
      "epoch": 0.13791113675706868,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999943036172771,
      "loss": 7.2407,
      "mean_token_accuracy": 0.13760888651013375,
      "num_tokens": 3024786.0,
      "step": 1195
    },
    {
      "entropy": 7.41163215637207,
      "epoch": 0.13848817080207732,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999940062067565,
      "loss": 7.1789,
      "mean_token_accuracy": 0.1398928850889206,
      "num_tokens": 3037179.0,
      "step": 1200
    },
    {
      "entropy": 7.380322027206421,
      "epoch": 0.13906520484708598,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999937012286742,
      "loss": 7.2651,
      "mean_token_accuracy": 0.1392878621816635,
      "num_tokens": 3049443.0,
      "step": 1205
    },
    {
      "entropy": 7.35688591003418,
      "epoch": 0.13964223889209462,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999933886830405,
      "loss": 7.1713,
      "mean_token_accuracy": 0.133247160166502,
      "num_tokens": 3061857.0,
      "step": 1210
    },
    {
      "entropy": 7.40545539855957,
      "epoch": 0.1402192729371033,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499993068569866,
      "loss": 7.2008,
      "mean_token_accuracy": 0.14127044454216958,
      "num_tokens": 3075185.0,
      "step": 1215
    },
    {
      "entropy": 7.398829460144043,
      "epoch": 0.14079630698211196,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999927408891614,
      "loss": 7.2143,
      "mean_token_accuracy": 0.14073001742362976,
      "num_tokens": 3087645.0,
      "step": 1220
    },
    {
      "entropy": 7.389093589782715,
      "epoch": 0.1413733410271206,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999924056409378,
      "loss": 7.2218,
      "mean_token_accuracy": 0.13467969745397568,
      "num_tokens": 3101185.0,
      "step": 1225
    },
    {
      "entropy": 7.341273593902588,
      "epoch": 0.14195037507212926,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999920628252063,
      "loss": 7.1938,
      "mean_token_accuracy": 0.13926867693662642,
      "num_tokens": 3113694.0,
      "step": 1230
    },
    {
      "entropy": 7.389741516113281,
      "epoch": 0.1425274091171379,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999917124419785,
      "loss": 7.2361,
      "mean_token_accuracy": 0.1366308718919754,
      "num_tokens": 3128225.0,
      "step": 1235
    },
    {
      "entropy": 7.346083259582519,
      "epoch": 0.14310444316214657,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999913544912664,
      "loss": 7.1621,
      "mean_token_accuracy": 0.14138347059488296,
      "num_tokens": 3140118.0,
      "step": 1240
    },
    {
      "entropy": 7.316737031936645,
      "epoch": 0.1436814772071552,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999909889730817,
      "loss": 7.2285,
      "mean_token_accuracy": 0.13803385123610495,
      "num_tokens": 3153479.0,
      "step": 1245
    },
    {
      "entropy": 7.433447456359863,
      "epoch": 0.14425851125216388,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999906158874368,
      "loss": 7.2226,
      "mean_token_accuracy": 0.14075266495347022,
      "num_tokens": 3167062.0,
      "step": 1250
    },
    {
      "entropy": 7.32918004989624,
      "epoch": 0.14483554529717255,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999902352343444,
      "loss": 7.181,
      "mean_token_accuracy": 0.13708969652652742,
      "num_tokens": 3179061.0,
      "step": 1255
    },
    {
      "entropy": 7.340552043914795,
      "epoch": 0.1454125793421812,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999898470138171,
      "loss": 7.1299,
      "mean_token_accuracy": 0.137362752109766,
      "num_tokens": 3191333.0,
      "step": 1260
    },
    {
      "entropy": 7.366755676269531,
      "epoch": 0.14598961338718985,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999894512258681,
      "loss": 7.2347,
      "mean_token_accuracy": 0.1339644305408001,
      "num_tokens": 3204487.0,
      "step": 1265
    },
    {
      "entropy": 7.3037999153137205,
      "epoch": 0.1465666474321985,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999890478705107,
      "loss": 7.2018,
      "mean_token_accuracy": 0.1373523458838463,
      "num_tokens": 3216733.0,
      "step": 1270
    },
    {
      "entropy": 7.35508165359497,
      "epoch": 0.14714368147720716,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999886369477585,
      "loss": 7.1258,
      "mean_token_accuracy": 0.14343418851494788,
      "num_tokens": 3229605.0,
      "step": 1275
    },
    {
      "entropy": 7.294068384170532,
      "epoch": 0.1477207155222158,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999882184576251,
      "loss": 7.1764,
      "mean_token_accuracy": 0.13913899287581444,
      "num_tokens": 3241023.0,
      "step": 1280
    },
    {
      "entropy": 7.356681680679321,
      "epoch": 0.14829774956722447,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999877924001248,
      "loss": 7.1466,
      "mean_token_accuracy": 0.1401684083044529,
      "num_tokens": 3252435.0,
      "step": 1285
    },
    {
      "entropy": 7.2978212356567385,
      "epoch": 0.1488747836122331,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999873587752718,
      "loss": 7.1465,
      "mean_token_accuracy": 0.14340650513768197,
      "num_tokens": 3265663.0,
      "step": 1290
    },
    {
      "entropy": 7.384179496765137,
      "epoch": 0.14945181765724178,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999869175830808,
      "loss": 7.2055,
      "mean_token_accuracy": 0.13356677517294885,
      "num_tokens": 3278309.0,
      "step": 1295
    },
    {
      "entropy": 7.367056083679199,
      "epoch": 0.15002885170225044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999864688235666,
      "loss": 7.2115,
      "mean_token_accuracy": 0.1325480066239834,
      "num_tokens": 3290899.0,
      "step": 1300
    },
    {
      "entropy": 7.298659467697144,
      "epoch": 0.15060588574725908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999860124967442,
      "loss": 7.1556,
      "mean_token_accuracy": 0.13973417952656747,
      "num_tokens": 3304383.0,
      "step": 1305
    },
    {
      "entropy": 7.280820751190186,
      "epoch": 0.15118291979226775,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999855486026291,
      "loss": 7.1489,
      "mean_token_accuracy": 0.13906904980540274,
      "num_tokens": 3316310.0,
      "step": 1310
    },
    {
      "entropy": 7.3616657733917235,
      "epoch": 0.1517599538372764,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999850771412369,
      "loss": 7.2426,
      "mean_token_accuracy": 0.13819348588585853,
      "num_tokens": 3330018.0,
      "step": 1315
    },
    {
      "entropy": 7.392954874038696,
      "epoch": 0.15233698788228506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999845981125832,
      "loss": 7.3005,
      "mean_token_accuracy": 0.1357297696173191,
      "num_tokens": 3343134.0,
      "step": 1320
    },
    {
      "entropy": 7.2494100570678714,
      "epoch": 0.1529140219272937,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999841115166844,
      "loss": 7.0969,
      "mean_token_accuracy": 0.14118720814585686,
      "num_tokens": 3356262.0,
      "step": 1325
    },
    {
      "entropy": 7.343341732025147,
      "epoch": 0.15349105597230236,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999836173535568,
      "loss": 7.2429,
      "mean_token_accuracy": 0.12794803753495215,
      "num_tokens": 3369672.0,
      "step": 1330
    },
    {
      "entropy": 7.342515563964843,
      "epoch": 0.15406809001731103,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999831156232169,
      "loss": 7.1474,
      "mean_token_accuracy": 0.14077327102422715,
      "num_tokens": 3381677.0,
      "step": 1335
    },
    {
      "entropy": 7.262819242477417,
      "epoch": 0.15464512406231967,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999826063256818,
      "loss": 7.0818,
      "mean_token_accuracy": 0.14118852764368056,
      "num_tokens": 3394662.0,
      "step": 1340
    },
    {
      "entropy": 7.284024095535278,
      "epoch": 0.15522215810732834,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999820894609683,
      "loss": 7.0937,
      "mean_token_accuracy": 0.14023175537586213,
      "num_tokens": 3407286.0,
      "step": 1345
    },
    {
      "entropy": 7.391539764404297,
      "epoch": 0.15579919215233698,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999815650290941,
      "loss": 7.2068,
      "mean_token_accuracy": 0.13057266250252725,
      "num_tokens": 3419478.0,
      "step": 1350
    },
    {
      "entropy": 7.242104721069336,
      "epoch": 0.15637622619734565,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999810330300766,
      "loss": 7.1625,
      "mean_token_accuracy": 0.13775511607527732,
      "num_tokens": 3433188.0,
      "step": 1355
    },
    {
      "entropy": 7.366675519943238,
      "epoch": 0.1569532602423543,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499980493463934,
      "loss": 7.0901,
      "mean_token_accuracy": 0.13610174432396888,
      "num_tokens": 3445806.0,
      "step": 1360
    },
    {
      "entropy": 7.216462755203247,
      "epoch": 0.15753029428736295,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999799463306841,
      "loss": 7.1471,
      "mean_token_accuracy": 0.13988320752978325,
      "num_tokens": 3458931.0,
      "step": 1365
    },
    {
      "entropy": 7.320661640167236,
      "epoch": 0.15810732833237162,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999793916303455,
      "loss": 7.1803,
      "mean_token_accuracy": 0.13539275228977204,
      "num_tokens": 3471338.0,
      "step": 1370
    },
    {
      "entropy": 7.3501372814178465,
      "epoch": 0.15868436237738026,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999788293629368,
      "loss": 7.2315,
      "mean_token_accuracy": 0.13646726906299592,
      "num_tokens": 3484865.0,
      "step": 1375
    },
    {
      "entropy": 7.3266260623931885,
      "epoch": 0.15926139642238893,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499978259528477,
      "loss": 7.0678,
      "mean_token_accuracy": 0.14280698746442794,
      "num_tokens": 3496676.0,
      "step": 1380
    },
    {
      "entropy": 7.195064067840576,
      "epoch": 0.15983843046739757,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999776821269852,
      "loss": 7.0447,
      "mean_token_accuracy": 0.1451018936932087,
      "num_tokens": 3508748.0,
      "step": 1385
    },
    {
      "entropy": 7.267773008346557,
      "epoch": 0.16041546451240624,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999770971584807,
      "loss": 7.1826,
      "mean_token_accuracy": 0.13283163756132127,
      "num_tokens": 3521490.0,
      "step": 1390
    },
    {
      "entropy": 7.2999695301055905,
      "epoch": 0.16099249855741488,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999765046229834,
      "loss": 7.1451,
      "mean_token_accuracy": 0.13972139805555345,
      "num_tokens": 3534092.0,
      "step": 1395
    },
    {
      "entropy": 7.30112590789795,
      "epoch": 0.16156953260242354,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499975904520513,
      "loss": 7.1229,
      "mean_token_accuracy": 0.13600729033350945,
      "num_tokens": 3546583.0,
      "step": 1400
    },
    {
      "entropy": 7.261274099349976,
      "epoch": 0.1621465666474322,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049997529685109,
      "loss": 7.0986,
      "mean_token_accuracy": 0.13902345672249794,
      "num_tokens": 3559409.0,
      "step": 1405
    },
    {
      "entropy": 7.244414520263672,
      "epoch": 0.16272360069244085,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999746816147344,
      "loss": 7.1063,
      "mean_token_accuracy": 0.13943156972527504,
      "num_tokens": 3572362.0,
      "step": 1410
    },
    {
      "entropy": 7.205703496932983,
      "epoch": 0.16330063473744952,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999740588114673,
      "loss": 7.0916,
      "mean_token_accuracy": 0.14903789088129998,
      "num_tokens": 3585505.0,
      "step": 1415
    },
    {
      "entropy": 7.34771957397461,
      "epoch": 0.16387766878245816,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999734284413095,
      "loss": 7.1202,
      "mean_token_accuracy": 0.14055844619870186,
      "num_tokens": 3597322.0,
      "step": 1420
    },
    {
      "entropy": 7.172052621841431,
      "epoch": 0.16445470282746683,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499972790504282,
      "loss": 7.1121,
      "mean_token_accuracy": 0.13925491869449616,
      "num_tokens": 3609941.0,
      "step": 1425
    },
    {
      "entropy": 7.33465166091919,
      "epoch": 0.16503173687247547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999721450004067,
      "loss": 7.1586,
      "mean_token_accuracy": 0.13594435006380082,
      "num_tokens": 3621751.0,
      "step": 1430
    },
    {
      "entropy": 7.223167848587036,
      "epoch": 0.16560877091748413,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999714919297049,
      "loss": 7.1104,
      "mean_token_accuracy": 0.13295586705207824,
      "num_tokens": 3633070.0,
      "step": 1435
    },
    {
      "entropy": 7.267610120773315,
      "epoch": 0.1661858049624928,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999708312921987,
      "loss": 7.0962,
      "mean_token_accuracy": 0.13987760469317437,
      "num_tokens": 3646730.0,
      "step": 1440
    },
    {
      "entropy": 7.284721183776855,
      "epoch": 0.16676283900750144,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999701630879103,
      "loss": 7.1188,
      "mean_token_accuracy": 0.1329401448369026,
      "num_tokens": 3659976.0,
      "step": 1445
    },
    {
      "entropy": 7.247362232208252,
      "epoch": 0.1673398730525101,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999694873168623,
      "loss": 7.0619,
      "mean_token_accuracy": 0.13823318034410476,
      "num_tokens": 3672607.0,
      "step": 1450
    },
    {
      "entropy": 7.093194961547852,
      "epoch": 0.16791690709751875,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999688039790773,
      "loss": 6.9973,
      "mean_token_accuracy": 0.1486719325184822,
      "num_tokens": 3686338.0,
      "step": 1455
    },
    {
      "entropy": 7.252864933013916,
      "epoch": 0.16849394114252741,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999681130745784,
      "loss": 7.001,
      "mean_token_accuracy": 0.1449768289923668,
      "num_tokens": 3698986.0,
      "step": 1460
    },
    {
      "entropy": 7.1841700077056885,
      "epoch": 0.16907097518753605,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999674146033888,
      "loss": 7.0495,
      "mean_token_accuracy": 0.14095961302518845,
      "num_tokens": 3712453.0,
      "step": 1465
    },
    {
      "entropy": 7.133856296539307,
      "epoch": 0.16964800923254472,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999667085655318,
      "loss": 7.0579,
      "mean_token_accuracy": 0.13971312269568442,
      "num_tokens": 3724769.0,
      "step": 1470
    },
    {
      "entropy": 7.322906827926635,
      "epoch": 0.1702250432775534,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999659949610313,
      "loss": 7.105,
      "mean_token_accuracy": 0.13217086791992189,
      "num_tokens": 3738052.0,
      "step": 1475
    },
    {
      "entropy": 7.270902395248413,
      "epoch": 0.17080207732256203,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999652737899114,
      "loss": 7.0903,
      "mean_token_accuracy": 0.13719962015748025,
      "num_tokens": 3751279.0,
      "step": 1480
    },
    {
      "entropy": 7.262608242034912,
      "epoch": 0.1713791113675707,
      "grad_norm": 0.75,
      "learning_rate": 0.0004999645450521963,
      "loss": 7.1323,
      "mean_token_accuracy": 0.13691732063889503,
      "num_tokens": 3764688.0,
      "step": 1485
    },
    {
      "entropy": 7.236039876937866,
      "epoch": 0.17195614541257934,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999638087479104,
      "loss": 7.0961,
      "mean_token_accuracy": 0.13691733703017234,
      "num_tokens": 3776892.0,
      "step": 1490
    },
    {
      "entropy": 7.258961296081543,
      "epoch": 0.172533179457588,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999630648770786,
      "loss": 7.0148,
      "mean_token_accuracy": 0.14263081848621367,
      "num_tokens": 3789287.0,
      "step": 1495
    },
    {
      "entropy": 7.17362151145935,
      "epoch": 0.17311021350259664,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999623134397257,
      "loss": 7.0039,
      "mean_token_accuracy": 0.14118167236447335,
      "num_tokens": 3801467.0,
      "step": 1500
    },
    {
      "entropy": 7.180401945114136,
      "epoch": 0.1736872475476053,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999615544358774,
      "loss": 7.0941,
      "mean_token_accuracy": 0.1388135813176632,
      "num_tokens": 3814641.0,
      "step": 1505
    },
    {
      "entropy": 7.323696327209473,
      "epoch": 0.17426428159261395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999607878655587,
      "loss": 7.1379,
      "mean_token_accuracy": 0.13087100833654403,
      "num_tokens": 3827187.0,
      "step": 1510
    },
    {
      "entropy": 7.094374799728394,
      "epoch": 0.17484131563762262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999600137287958,
      "loss": 6.972,
      "mean_token_accuracy": 0.14806395322084426,
      "num_tokens": 3839832.0,
      "step": 1515
    },
    {
      "entropy": 7.095919132232666,
      "epoch": 0.17541834968263129,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999592320256146,
      "loss": 6.9735,
      "mean_token_accuracy": 0.1458645448088646,
      "num_tokens": 3852026.0,
      "step": 1520
    },
    {
      "entropy": 7.2352509021759035,
      "epoch": 0.17599538372763993,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999584427560412,
      "loss": 7.017,
      "mean_token_accuracy": 0.14108823537826537,
      "num_tokens": 3864481.0,
      "step": 1525
    },
    {
      "entropy": 7.192263603210449,
      "epoch": 0.1765724177726486,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999576459201024,
      "loss": 7.0276,
      "mean_token_accuracy": 0.1382262274622917,
      "num_tokens": 3876896.0,
      "step": 1530
    },
    {
      "entropy": 7.149485349655151,
      "epoch": 0.17714945181765723,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499956841517825,
      "loss": 6.9683,
      "mean_token_accuracy": 0.14003728628158568,
      "num_tokens": 3890230.0,
      "step": 1535
    },
    {
      "entropy": 7.214767408370972,
      "epoch": 0.1777264858626659,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499956029549236,
      "loss": 7.0839,
      "mean_token_accuracy": 0.13458969220519065,
      "num_tokens": 3903138.0,
      "step": 1540
    },
    {
      "entropy": 7.140475034713745,
      "epoch": 0.17830351990767454,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999552100143625,
      "loss": 6.9988,
      "mean_token_accuracy": 0.1441511981189251,
      "num_tokens": 3915709.0,
      "step": 1545
    },
    {
      "entropy": 7.149213743209839,
      "epoch": 0.1788805539526832,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004999543829132324,
      "loss": 7.0617,
      "mean_token_accuracy": 0.14265220016241073,
      "num_tokens": 3929265.0,
      "step": 1550
    },
    {
      "entropy": 7.208163404464722,
      "epoch": 0.17945758799769188,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999535482458734,
      "loss": 7.0482,
      "mean_token_accuracy": 0.14286358505487443,
      "num_tokens": 3940416.0,
      "step": 1555
    },
    {
      "entropy": 7.217859220504761,
      "epoch": 0.18003462204270052,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999527060123135,
      "loss": 7.0464,
      "mean_token_accuracy": 0.13893593847751617,
      "num_tokens": 3952999.0,
      "step": 1560
    },
    {
      "entropy": 7.1100293636322025,
      "epoch": 0.18061165608770918,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999518562125811,
      "loss": 7.0657,
      "mean_token_accuracy": 0.1427687995135784,
      "num_tokens": 3966852.0,
      "step": 1565
    },
    {
      "entropy": 7.210110139846802,
      "epoch": 0.18118869013271782,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999509988467047,
      "loss": 6.9639,
      "mean_token_accuracy": 0.14725383818149568,
      "num_tokens": 3978242.0,
      "step": 1570
    },
    {
      "entropy": 7.188309288024902,
      "epoch": 0.1817657241777265,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999501339147132,
      "loss": 7.0899,
      "mean_token_accuracy": 0.131745096296072,
      "num_tokens": 3993122.0,
      "step": 1575
    },
    {
      "entropy": 7.261885738372802,
      "epoch": 0.18234275822273513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999492614166357,
      "loss": 7.0259,
      "mean_token_accuracy": 0.13976342976093292,
      "num_tokens": 4005230.0,
      "step": 1580
    },
    {
      "entropy": 7.037604188919067,
      "epoch": 0.1829197922677438,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999483813525014,
      "loss": 6.978,
      "mean_token_accuracy": 0.14762265235185623,
      "num_tokens": 4017413.0,
      "step": 1585
    },
    {
      "entropy": 7.191242218017578,
      "epoch": 0.18349682631275246,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999474937223403,
      "loss": 6.991,
      "mean_token_accuracy": 0.13611105009913443,
      "num_tokens": 4029679.0,
      "step": 1590
    },
    {
      "entropy": 7.1627381324768065,
      "epoch": 0.1840738603577611,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999465985261818,
      "loss": 6.9535,
      "mean_token_accuracy": 0.14157049655914306,
      "num_tokens": 4041857.0,
      "step": 1595
    },
    {
      "entropy": 7.140171051025391,
      "epoch": 0.18465089440276977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999456957640562,
      "loss": 7.0374,
      "mean_token_accuracy": 0.13944343626499175,
      "num_tokens": 4054743.0,
      "step": 1600
    },
    {
      "entropy": 7.072859477996826,
      "epoch": 0.1852279284477784,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999447854359939,
      "loss": 6.9562,
      "mean_token_accuracy": 0.14270277544856072,
      "num_tokens": 4067554.0,
      "step": 1605
    },
    {
      "entropy": 7.180684185028076,
      "epoch": 0.18580496249278708,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999438675420255,
      "loss": 6.9519,
      "mean_token_accuracy": 0.1420460894703865,
      "num_tokens": 4079912.0,
      "step": 1610
    },
    {
      "entropy": 7.153562593460083,
      "epoch": 0.18638199653779572,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999429420821818,
      "loss": 7.0764,
      "mean_token_accuracy": 0.13569277077913283,
      "num_tokens": 4093217.0,
      "step": 1615
    },
    {
      "entropy": 7.13769006729126,
      "epoch": 0.1869590305828044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499942009056494,
      "loss": 7.0162,
      "mean_token_accuracy": 0.1355166032910347,
      "num_tokens": 4104328.0,
      "step": 1620
    },
    {
      "entropy": 7.216564655303955,
      "epoch": 0.18753606462781305,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999410684649935,
      "loss": 6.9849,
      "mean_token_accuracy": 0.13703312277793883,
      "num_tokens": 4116711.0,
      "step": 1625
    },
    {
      "entropy": 7.153830718994141,
      "epoch": 0.1881130986728217,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999401203077119,
      "loss": 7.0056,
      "mean_token_accuracy": 0.14228765815496444,
      "num_tokens": 4129066.0,
      "step": 1630
    },
    {
      "entropy": 7.08653039932251,
      "epoch": 0.18869013271783036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499939164584681,
      "loss": 6.9143,
      "mean_token_accuracy": 0.14499093145132064,
      "num_tokens": 4141312.0,
      "step": 1635
    },
    {
      "entropy": 7.0617640018463135,
      "epoch": 0.189267166762839,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999382012959331,
      "loss": 6.961,
      "mean_token_accuracy": 0.1401074267923832,
      "num_tokens": 4154687.0,
      "step": 1640
    },
    {
      "entropy": 7.193846559524536,
      "epoch": 0.18984420080784767,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999372304415005,
      "loss": 7.0702,
      "mean_token_accuracy": 0.1319998621940613,
      "num_tokens": 4168401.0,
      "step": 1645
    },
    {
      "entropy": 7.144888019561767,
      "epoch": 0.1904212348528563,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999362520214159,
      "loss": 6.9787,
      "mean_token_accuracy": 0.143699299544096,
      "num_tokens": 4180707.0,
      "step": 1650
    },
    {
      "entropy": 7.093651580810547,
      "epoch": 0.19099826889786498,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999352660357122,
      "loss": 6.9224,
      "mean_token_accuracy": 0.1452721819281578,
      "num_tokens": 4192939.0,
      "step": 1655
    },
    {
      "entropy": 7.1100091457366945,
      "epoch": 0.19157530294287364,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999342724844226,
      "loss": 6.9571,
      "mean_token_accuracy": 0.14119631350040435,
      "num_tokens": 4206277.0,
      "step": 1660
    },
    {
      "entropy": 7.124865531921387,
      "epoch": 0.19215233698788228,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999332713675804,
      "loss": 6.9803,
      "mean_token_accuracy": 0.14399617463350295,
      "num_tokens": 4218672.0,
      "step": 1665
    },
    {
      "entropy": 7.107157850265503,
      "epoch": 0.19272937103289095,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999322626852193,
      "loss": 6.9265,
      "mean_token_accuracy": 0.15203145295381545,
      "num_tokens": 4229002.0,
      "step": 1670
    },
    {
      "entropy": 7.030970716476441,
      "epoch": 0.1933064050778996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999312464373734,
      "loss": 6.8427,
      "mean_token_accuracy": 0.15255002826452255,
      "num_tokens": 4241986.0,
      "step": 1675
    },
    {
      "entropy": 7.03170862197876,
      "epoch": 0.19388343912290826,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999302226240767,
      "loss": 6.9614,
      "mean_token_accuracy": 0.1456663116812706,
      "num_tokens": 4254670.0,
      "step": 1680
    },
    {
      "entropy": 7.147547388076783,
      "epoch": 0.1944604731679169,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999291912453637,
      "loss": 6.9988,
      "mean_token_accuracy": 0.1436385042965412,
      "num_tokens": 4267561.0,
      "step": 1685
    },
    {
      "entropy": 7.182132863998413,
      "epoch": 0.19503750721292556,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999281523012691,
      "loss": 6.9331,
      "mean_token_accuracy": 0.14853112250566483,
      "num_tokens": 4280145.0,
      "step": 1690
    },
    {
      "entropy": 7.027525091171265,
      "epoch": 0.1956145412579342,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999271057918278,
      "loss": 6.9788,
      "mean_token_accuracy": 0.14283846467733383,
      "num_tokens": 4294089.0,
      "step": 1695
    },
    {
      "entropy": 7.107568359375,
      "epoch": 0.19619157530294287,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999260517170751,
      "loss": 6.9167,
      "mean_token_accuracy": 0.14075717329978943,
      "num_tokens": 4305926.0,
      "step": 1700
    },
    {
      "entropy": 7.086518144607544,
      "epoch": 0.19676860934795154,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999249900770463,
      "loss": 6.9481,
      "mean_token_accuracy": 0.14459511563181876,
      "num_tokens": 4319319.0,
      "step": 1705
    },
    {
      "entropy": 7.104380512237549,
      "epoch": 0.19734564339296018,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999239208717772,
      "loss": 6.9396,
      "mean_token_accuracy": 0.1490633636713028,
      "num_tokens": 4331650.0,
      "step": 1710
    },
    {
      "entropy": 6.978099775314331,
      "epoch": 0.19792267743796885,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999228441013037,
      "loss": 6.9182,
      "mean_token_accuracy": 0.14964553862810134,
      "num_tokens": 4344289.0,
      "step": 1715
    },
    {
      "entropy": 7.099493741989136,
      "epoch": 0.1984997114829775,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499921759765662,
      "loss": 6.9221,
      "mean_token_accuracy": 0.14225320667028427,
      "num_tokens": 4357316.0,
      "step": 1720
    },
    {
      "entropy": 7.004121541976929,
      "epoch": 0.19907674552798615,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999206678648888,
      "loss": 6.9264,
      "mean_token_accuracy": 0.1489520326256752,
      "num_tokens": 4369383.0,
      "step": 1725
    },
    {
      "entropy": 7.094613456726075,
      "epoch": 0.1996537795729948,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999195683990206,
      "loss": 6.9235,
      "mean_token_accuracy": 0.14189272969961167,
      "num_tokens": 4383509.0,
      "step": 1730
    },
    {
      "entropy": 7.103254556655884,
      "epoch": 0.20023081361800346,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999184613680945,
      "loss": 6.9059,
      "mean_token_accuracy": 0.14382569789886473,
      "num_tokens": 4395105.0,
      "step": 1735
    },
    {
      "entropy": 7.039672946929931,
      "epoch": 0.20080784766301213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999173467721476,
      "loss": 6.9091,
      "mean_token_accuracy": 0.15155962854623795,
      "num_tokens": 4407473.0,
      "step": 1740
    },
    {
      "entropy": 7.116544246673584,
      "epoch": 0.20138488170802077,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999162246112175,
      "loss": 6.991,
      "mean_token_accuracy": 0.13716431334614754,
      "num_tokens": 4420186.0,
      "step": 1745
    },
    {
      "entropy": 7.053312683105469,
      "epoch": 0.20196191575302944,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999150948853419,
      "loss": 6.8901,
      "mean_token_accuracy": 0.14453017637133597,
      "num_tokens": 4432944.0,
      "step": 1750
    },
    {
      "entropy": 7.120892095565796,
      "epoch": 0.20253894979803808,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999139575945587,
      "loss": 6.9558,
      "mean_token_accuracy": 0.14354105591773986,
      "num_tokens": 4445784.0,
      "step": 1755
    },
    {
      "entropy": 6.965320730209351,
      "epoch": 0.20311598384304674,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999128127389065,
      "loss": 6.8445,
      "mean_token_accuracy": 0.15029353350400926,
      "num_tokens": 4458338.0,
      "step": 1760
    },
    {
      "entropy": 7.06877818107605,
      "epoch": 0.20369301788805538,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999116603184233,
      "loss": 6.8774,
      "mean_token_accuracy": 0.13867350146174431,
      "num_tokens": 4471058.0,
      "step": 1765
    },
    {
      "entropy": 6.997467470169068,
      "epoch": 0.20427005193306405,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004999105003331482,
      "loss": 6.8973,
      "mean_token_accuracy": 0.14823642373085022,
      "num_tokens": 4485561.0,
      "step": 1770
    },
    {
      "entropy": 7.019530820846557,
      "epoch": 0.20484708597807272,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999093327831202,
      "loss": 6.9028,
      "mean_token_accuracy": 0.14461316913366318,
      "num_tokens": 4497865.0,
      "step": 1775
    },
    {
      "entropy": 6.937532806396485,
      "epoch": 0.20542412002308136,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999081576683784,
      "loss": 6.8082,
      "mean_token_accuracy": 0.15614131838083267,
      "num_tokens": 4510291.0,
      "step": 1780
    },
    {
      "entropy": 7.044717454910279,
      "epoch": 0.20600115406809003,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999069749889626,
      "loss": 6.9073,
      "mean_token_accuracy": 0.1450217232108116,
      "num_tokens": 4521855.0,
      "step": 1785
    },
    {
      "entropy": 7.052458572387695,
      "epoch": 0.20657818811309867,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999057847449123,
      "loss": 6.99,
      "mean_token_accuracy": 0.1421843633055687,
      "num_tokens": 4534123.0,
      "step": 1790
    },
    {
      "entropy": 7.197088527679443,
      "epoch": 0.20715522215810733,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999045869362678,
      "loss": 6.9418,
      "mean_token_accuracy": 0.13784723579883576,
      "num_tokens": 4547298.0,
      "step": 1795
    },
    {
      "entropy": 6.992271852493286,
      "epoch": 0.20773225620311597,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499903381563069,
      "loss": 6.8698,
      "mean_token_accuracy": 0.1466771975159645,
      "num_tokens": 4560945.0,
      "step": 1800
    },
    {
      "entropy": 7.020458364486695,
      "epoch": 0.20830929024812464,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999021686253568,
      "loss": 6.9329,
      "mean_token_accuracy": 0.14652197062969208,
      "num_tokens": 4572696.0,
      "step": 1805
    },
    {
      "entropy": 6.95990252494812,
      "epoch": 0.2088863242931333,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999009481231719,
      "loss": 6.8845,
      "mean_token_accuracy": 0.14989398419857025,
      "num_tokens": 4585670.0,
      "step": 1810
    },
    {
      "entropy": 7.115705442428589,
      "epoch": 0.20946335833814195,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998997200565553,
      "loss": 6.8657,
      "mean_token_accuracy": 0.15091128051280975,
      "num_tokens": 4597694.0,
      "step": 1815
    },
    {
      "entropy": 7.001132869720459,
      "epoch": 0.21004039238315061,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998984844255483,
      "loss": 6.8746,
      "mean_token_accuracy": 0.14744569212198258,
      "num_tokens": 4609924.0,
      "step": 1820
    },
    {
      "entropy": 7.056229448318481,
      "epoch": 0.21061742642815925,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998972412301925,
      "loss": 6.885,
      "mean_token_accuracy": 0.1468400090932846,
      "num_tokens": 4621668.0,
      "step": 1825
    },
    {
      "entropy": 7.011569261550903,
      "epoch": 0.21119446047316792,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004998959904705297,
      "loss": 6.9449,
      "mean_token_accuracy": 0.14132919982075692,
      "num_tokens": 4635970.0,
      "step": 1830
    },
    {
      "entropy": 7.080425167083741,
      "epoch": 0.21177149451817656,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998947321466021,
      "loss": 6.9183,
      "mean_token_accuracy": 0.14053108170628548,
      "num_tokens": 4649557.0,
      "step": 1835
    },
    {
      "entropy": 7.013520288467407,
      "epoch": 0.21234852856318523,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998934662584518,
      "loss": 6.9063,
      "mean_token_accuracy": 0.14713118821382523,
      "num_tokens": 4662572.0,
      "step": 1840
    },
    {
      "entropy": 7.0543595314025875,
      "epoch": 0.2129255626081939,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998921928061214,
      "loss": 6.8599,
      "mean_token_accuracy": 0.14088837131857873,
      "num_tokens": 4674783.0,
      "step": 1845
    },
    {
      "entropy": 6.977069759368897,
      "epoch": 0.21350259665320254,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998909117896539,
      "loss": 6.8811,
      "mean_token_accuracy": 0.1488935634493828,
      "num_tokens": 4687543.0,
      "step": 1850
    },
    {
      "entropy": 7.099917268753051,
      "epoch": 0.2140796306982112,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998896232090922,
      "loss": 6.8471,
      "mean_token_accuracy": 0.14590785950422286,
      "num_tokens": 4700225.0,
      "step": 1855
    },
    {
      "entropy": 6.8859778881073,
      "epoch": 0.21465666474321984,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004998883270644798,
      "loss": 6.8981,
      "mean_token_accuracy": 0.1461702950298786,
      "num_tokens": 4712950.0,
      "step": 1860
    },
    {
      "entropy": 7.103154993057251,
      "epoch": 0.2152336987882285,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998870233558602,
      "loss": 6.9823,
      "mean_token_accuracy": 0.14054031670093536,
      "num_tokens": 4725563.0,
      "step": 1865
    },
    {
      "entropy": 7.021540546417237,
      "epoch": 0.21581073283323715,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998857120832774,
      "loss": 6.8994,
      "mean_token_accuracy": 0.14567926824092864,
      "num_tokens": 4739057.0,
      "step": 1870
    },
    {
      "entropy": 7.013464498519897,
      "epoch": 0.21638776687824582,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998843932467751,
      "loss": 6.8557,
      "mean_token_accuracy": 0.14868776947259904,
      "num_tokens": 4752782.0,
      "step": 1875
    },
    {
      "entropy": 7.032290697097778,
      "epoch": 0.21696480092325446,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998830668463982,
      "loss": 6.8122,
      "mean_token_accuracy": 0.14852888584136964,
      "num_tokens": 4765740.0,
      "step": 1880
    },
    {
      "entropy": 6.985332727432251,
      "epoch": 0.21754183496826313,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998817328821909,
      "loss": 6.8821,
      "mean_token_accuracy": 0.14597213715314866,
      "num_tokens": 4778711.0,
      "step": 1885
    },
    {
      "entropy": 7.0917829990386965,
      "epoch": 0.2181188690132718,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998803913541983,
      "loss": 6.8686,
      "mean_token_accuracy": 0.14363647997379303,
      "num_tokens": 4791402.0,
      "step": 1890
    },
    {
      "entropy": 7.007903146743774,
      "epoch": 0.21869590305828043,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998790422624654,
      "loss": 6.8535,
      "mean_token_accuracy": 0.13729466870427132,
      "num_tokens": 4804256.0,
      "step": 1895
    },
    {
      "entropy": 6.964800500869751,
      "epoch": 0.2192729371032891,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998776856070376,
      "loss": 6.8066,
      "mean_token_accuracy": 0.1438037261366844,
      "num_tokens": 4817424.0,
      "step": 1900
    },
    {
      "entropy": 7.044483804702759,
      "epoch": 0.21984997114829774,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004998763213879606,
      "loss": 6.8492,
      "mean_token_accuracy": 0.1408962205052376,
      "num_tokens": 4829222.0,
      "step": 1905
    },
    {
      "entropy": 6.946775960922241,
      "epoch": 0.2204270051933064,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998749496052803,
      "loss": 6.8166,
      "mean_token_accuracy": 0.1510320410132408,
      "num_tokens": 4844011.0,
      "step": 1910
    },
    {
      "entropy": 7.0013550281524655,
      "epoch": 0.22100403923831505,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998735702590426,
      "loss": 6.8753,
      "mean_token_accuracy": 0.14439389556646348,
      "num_tokens": 4857116.0,
      "step": 1915
    },
    {
      "entropy": 6.983421802520752,
      "epoch": 0.22158107328332372,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998721833492942,
      "loss": 6.7985,
      "mean_token_accuracy": 0.15554276704788209,
      "num_tokens": 4867680.0,
      "step": 1920
    },
    {
      "entropy": 6.959651088714599,
      "epoch": 0.22215810732833238,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998707888760816,
      "loss": 6.87,
      "mean_token_accuracy": 0.14348941370844842,
      "num_tokens": 4880553.0,
      "step": 1925
    },
    {
      "entropy": 7.035657167434692,
      "epoch": 0.22273514137334102,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998693868394516,
      "loss": 6.8344,
      "mean_token_accuracy": 0.14735666215419768,
      "num_tokens": 4893706.0,
      "step": 1930
    },
    {
      "entropy": 6.914802122116089,
      "epoch": 0.2233121754183497,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998679772394516,
      "loss": 6.9102,
      "mean_token_accuracy": 0.1454026609659195,
      "num_tokens": 4906016.0,
      "step": 1935
    },
    {
      "entropy": 6.97300386428833,
      "epoch": 0.22388920946335833,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998665600761289,
      "loss": 6.7939,
      "mean_token_accuracy": 0.14941516369581223,
      "num_tokens": 4918628.0,
      "step": 1940
    },
    {
      "entropy": 7.029558706283569,
      "epoch": 0.224466243508367,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499865135349531,
      "loss": 6.8703,
      "mean_token_accuracy": 0.13900768756866455,
      "num_tokens": 4930525.0,
      "step": 1945
    },
    {
      "entropy": 7.005332565307617,
      "epoch": 0.22504327755337564,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998637030597061,
      "loss": 6.8806,
      "mean_token_accuracy": 0.14249326959252356,
      "num_tokens": 4942510.0,
      "step": 1950
    },
    {
      "entropy": 7.0068284034729,
      "epoch": 0.2256203115983843,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998622632067022,
      "loss": 6.8463,
      "mean_token_accuracy": 0.1456683486700058,
      "num_tokens": 4954865.0,
      "step": 1955
    },
    {
      "entropy": 6.9775794506072994,
      "epoch": 0.22619734564339297,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998608157905678,
      "loss": 6.95,
      "mean_token_accuracy": 0.1380724899470806,
      "num_tokens": 4968078.0,
      "step": 1960
    },
    {
      "entropy": 7.020229148864746,
      "epoch": 0.2267743796884016,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998593608113515,
      "loss": 6.7892,
      "mean_token_accuracy": 0.14944704324007035,
      "num_tokens": 4980273.0,
      "step": 1965
    },
    {
      "entropy": 6.9938514709472654,
      "epoch": 0.22735141373341028,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998578982691024,
      "loss": 6.8542,
      "mean_token_accuracy": 0.14151277467608453,
      "num_tokens": 4992372.0,
      "step": 1970
    },
    {
      "entropy": 6.939938163757324,
      "epoch": 0.22792844777841892,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998564281638694,
      "loss": 6.8316,
      "mean_token_accuracy": 0.14077244997024535,
      "num_tokens": 5004766.0,
      "step": 1975
    },
    {
      "entropy": 7.018451166152954,
      "epoch": 0.2285054818234276,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998549504957023,
      "loss": 6.7877,
      "mean_token_accuracy": 0.14108608216047286,
      "num_tokens": 5017007.0,
      "step": 1980
    },
    {
      "entropy": 6.921030187606812,
      "epoch": 0.22908251586843623,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998534652646506,
      "loss": 6.8458,
      "mean_token_accuracy": 0.14241134524345397,
      "num_tokens": 5029506.0,
      "step": 1985
    },
    {
      "entropy": 6.955693626403809,
      "epoch": 0.2296595499134449,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998519724707642,
      "loss": 6.8155,
      "mean_token_accuracy": 0.14788847863674165,
      "num_tokens": 5042209.0,
      "step": 1990
    },
    {
      "entropy": 6.97462306022644,
      "epoch": 0.23023658395845356,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998504721140934,
      "loss": 6.8718,
      "mean_token_accuracy": 0.13877833858132363,
      "num_tokens": 5056026.0,
      "step": 1995
    },
    {
      "entropy": 6.990704870223999,
      "epoch": 0.2308136180034622,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998489641946887,
      "loss": 6.8118,
      "mean_token_accuracy": 0.14473072141408921,
      "num_tokens": 5068194.0,
      "step": 2000
    },
    {
      "entropy": 6.874226713180542,
      "epoch": 0.23139065204847087,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998474487126009,
      "loss": 6.7544,
      "mean_token_accuracy": 0.14470188170671464,
      "num_tokens": 5080468.0,
      "step": 2005
    },
    {
      "entropy": 7.0825056552886965,
      "epoch": 0.2319676860934795,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004998459256678806,
      "loss": 6.9486,
      "mean_token_accuracy": 0.13459489569067956,
      "num_tokens": 5093104.0,
      "step": 2010
    },
    {
      "entropy": 7.020360994338989,
      "epoch": 0.23254472013848818,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004998443950605796,
      "loss": 6.8025,
      "mean_token_accuracy": 0.14722402542829513,
      "num_tokens": 5105710.0,
      "step": 2015
    },
    {
      "entropy": 6.992463827133179,
      "epoch": 0.23312175418349682,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998428568907488,
      "loss": 6.8825,
      "mean_token_accuracy": 0.14219225868582724,
      "num_tokens": 5118946.0,
      "step": 2020
    },
    {
      "entropy": 6.929893112182617,
      "epoch": 0.23369878822850548,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998413111584403,
      "loss": 6.7236,
      "mean_token_accuracy": 0.15229955613613128,
      "num_tokens": 5132567.0,
      "step": 2025
    },
    {
      "entropy": 6.87170090675354,
      "epoch": 0.23427582227351415,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998397578637059,
      "loss": 6.7848,
      "mean_token_accuracy": 0.1508283868432045,
      "num_tokens": 5145669.0,
      "step": 2030
    },
    {
      "entropy": 6.9173017024993895,
      "epoch": 0.2348528563185228,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499838197006598,
      "loss": 6.7264,
      "mean_token_accuracy": 0.1511531800031662,
      "num_tokens": 5157753.0,
      "step": 2035
    },
    {
      "entropy": 6.895808982849121,
      "epoch": 0.23542989036353146,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499836628587169,
      "loss": 6.7534,
      "mean_token_accuracy": 0.1487403929233551,
      "num_tokens": 5170167.0,
      "step": 2040
    },
    {
      "entropy": 6.922911071777344,
      "epoch": 0.2360069244085401,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998350526054716,
      "loss": 6.7484,
      "mean_token_accuracy": 0.14699087888002396,
      "num_tokens": 5182143.0,
      "step": 2045
    },
    {
      "entropy": 6.920343017578125,
      "epoch": 0.23658395845354876,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998334690615591,
      "loss": 6.7837,
      "mean_token_accuracy": 0.14400450214743615,
      "num_tokens": 5195586.0,
      "step": 2050
    },
    {
      "entropy": 6.898791360855102,
      "epoch": 0.2371609924985574,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998318779554844,
      "loss": 6.7823,
      "mean_token_accuracy": 0.14565831422805786,
      "num_tokens": 5207774.0,
      "step": 2055
    },
    {
      "entropy": 6.857323026657104,
      "epoch": 0.23773802654356607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998302792873012,
      "loss": 6.7961,
      "mean_token_accuracy": 0.14450274109840394,
      "num_tokens": 5219330.0,
      "step": 2060
    },
    {
      "entropy": 7.0199343204498295,
      "epoch": 0.23831506058857474,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998286730570631,
      "loss": 6.8601,
      "mean_token_accuracy": 0.13632812798023225,
      "num_tokens": 5233817.0,
      "step": 2065
    },
    {
      "entropy": 6.916458797454834,
      "epoch": 0.23889209463358338,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998270592648245,
      "loss": 6.8575,
      "mean_token_accuracy": 0.14362055882811547,
      "num_tokens": 5246483.0,
      "step": 2070
    },
    {
      "entropy": 6.994282245635986,
      "epoch": 0.23946912867859205,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998254379106394,
      "loss": 6.839,
      "mean_token_accuracy": 0.14538607597351075,
      "num_tokens": 5258695.0,
      "step": 2075
    },
    {
      "entropy": 6.908141565322876,
      "epoch": 0.2400461627236007,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998238089945622,
      "loss": 6.8047,
      "mean_token_accuracy": 0.14092738628387452,
      "num_tokens": 5271355.0,
      "step": 2080
    },
    {
      "entropy": 7.025559520721435,
      "epoch": 0.24062319676860935,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004998221725166479,
      "loss": 6.9123,
      "mean_token_accuracy": 0.13938132375478746,
      "num_tokens": 5284905.0,
      "step": 2085
    },
    {
      "entropy": 6.899095153808593,
      "epoch": 0.241200230813618,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998205284769516,
      "loss": 6.8295,
      "mean_token_accuracy": 0.1434583306312561,
      "num_tokens": 5298266.0,
      "step": 2090
    },
    {
      "entropy": 6.973186731338501,
      "epoch": 0.24177726485862666,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998188768755285,
      "loss": 6.8747,
      "mean_token_accuracy": 0.13531978204846382,
      "num_tokens": 5311148.0,
      "step": 2095
    },
    {
      "entropy": 6.945560312271118,
      "epoch": 0.2423542989036353,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998172177124341,
      "loss": 6.7786,
      "mean_token_accuracy": 0.1462605893611908,
      "num_tokens": 5322638.0,
      "step": 2100
    },
    {
      "entropy": 6.831422948837281,
      "epoch": 0.24293133294864397,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998155509877242,
      "loss": 6.6679,
      "mean_token_accuracy": 0.15973918437957763,
      "num_tokens": 5336116.0,
      "step": 2105
    },
    {
      "entropy": 6.87315354347229,
      "epoch": 0.24350836699365264,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499813876701455,
      "loss": 6.7941,
      "mean_token_accuracy": 0.14872830137610435,
      "num_tokens": 5347561.0,
      "step": 2110
    },
    {
      "entropy": 7.002074909210205,
      "epoch": 0.24408540103866128,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998121948536828,
      "loss": 6.8244,
      "mean_token_accuracy": 0.1419316753745079,
      "num_tokens": 5360055.0,
      "step": 2115
    },
    {
      "entropy": 6.937633228302002,
      "epoch": 0.24466243508366994,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998105054444639,
      "loss": 6.7812,
      "mean_token_accuracy": 0.15047216340899466,
      "num_tokens": 5372722.0,
      "step": 2120
    },
    {
      "entropy": 6.960194253921509,
      "epoch": 0.24523946912867858,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998088084738555,
      "loss": 6.7418,
      "mean_token_accuracy": 0.14477523565292358,
      "num_tokens": 5385586.0,
      "step": 2125
    },
    {
      "entropy": 6.831562471389771,
      "epoch": 0.24581650317368725,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004998071039419144,
      "loss": 6.7828,
      "mean_token_accuracy": 0.1484777107834816,
      "num_tokens": 5399266.0,
      "step": 2130
    },
    {
      "entropy": 6.943821907043457,
      "epoch": 0.2463935372186959,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499805391848698,
      "loss": 6.7836,
      "mean_token_accuracy": 0.1484945073723793,
      "num_tokens": 5412093.0,
      "step": 2135
    },
    {
      "entropy": 6.854267024993897,
      "epoch": 0.24697057126370456,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499803672194264,
      "loss": 6.7768,
      "mean_token_accuracy": 0.14299426227808,
      "num_tokens": 5425181.0,
      "step": 2140
    },
    {
      "entropy": 6.8708819389343265,
      "epoch": 0.24754760530871323,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998019449786701,
      "loss": 6.6947,
      "mean_token_accuracy": 0.15227821469306946,
      "num_tokens": 5437392.0,
      "step": 2145
    },
    {
      "entropy": 6.947283506393433,
      "epoch": 0.24812463935372187,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998002102019744,
      "loss": 6.7657,
      "mean_token_accuracy": 0.15049567073583603,
      "num_tokens": 5449455.0,
      "step": 2150
    },
    {
      "entropy": 6.891529560089111,
      "epoch": 0.24870167339873053,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997984678642354,
      "loss": 6.7946,
      "mean_token_accuracy": 0.14755706638097763,
      "num_tokens": 5463095.0,
      "step": 2155
    },
    {
      "entropy": 6.925650119781494,
      "epoch": 0.24927870744373917,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997967179655115,
      "loss": 6.7416,
      "mean_token_accuracy": 0.15180930644273757,
      "num_tokens": 5474440.0,
      "step": 2160
    },
    {
      "entropy": 6.880334043502808,
      "epoch": 0.24985574148874784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997949605058617,
      "loss": 6.7436,
      "mean_token_accuracy": 0.15553607046604156,
      "num_tokens": 5487365.0,
      "step": 2165
    },
    {
      "entropy": 6.943732452392578,
      "epoch": 0.2504327755337565,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997931954853451,
      "loss": 6.8102,
      "mean_token_accuracy": 0.1437340646982193,
      "num_tokens": 5502356.0,
      "step": 2170
    },
    {
      "entropy": 6.863589477539063,
      "epoch": 0.2510098095787651,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499791422904021,
      "loss": 6.814,
      "mean_token_accuracy": 0.14547016769647597,
      "num_tokens": 5514504.0,
      "step": 2175
    },
    {
      "entropy": 7.013930749893189,
      "epoch": 0.2515868436237738,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997896427619491,
      "loss": 6.8368,
      "mean_token_accuracy": 0.14100218415260315,
      "num_tokens": 5526415.0,
      "step": 2180
    },
    {
      "entropy": 6.841579246520996,
      "epoch": 0.25216387766878245,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997878550591892,
      "loss": 6.6874,
      "mean_token_accuracy": 0.15348772406578065,
      "num_tokens": 5538484.0,
      "step": 2185
    },
    {
      "entropy": 6.90242280960083,
      "epoch": 0.2527409117137911,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997860597958013,
      "loss": 6.764,
      "mean_token_accuracy": 0.15220421701669692,
      "num_tokens": 5551315.0,
      "step": 2190
    },
    {
      "entropy": 6.826332235336304,
      "epoch": 0.2533179457587998,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997842569718461,
      "loss": 6.6467,
      "mean_token_accuracy": 0.15433922857046128,
      "num_tokens": 5564179.0,
      "step": 2195
    },
    {
      "entropy": 6.796852874755859,
      "epoch": 0.25389497980380843,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499782446587384,
      "loss": 6.7843,
      "mean_token_accuracy": 0.14296953305602073,
      "num_tokens": 5577060.0,
      "step": 2200
    },
    {
      "entropy": 6.819638252258301,
      "epoch": 0.25447201384881707,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499780628642476,
      "loss": 6.6023,
      "mean_token_accuracy": 0.16436567083001136,
      "num_tokens": 5589170.0,
      "step": 2205
    },
    {
      "entropy": 6.853410863876343,
      "epoch": 0.2550490478938257,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997788031371834,
      "loss": 6.7762,
      "mean_token_accuracy": 0.1478083088994026,
      "num_tokens": 5601747.0,
      "step": 2210
    },
    {
      "entropy": 6.910475969314575,
      "epoch": 0.2556260819388344,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997769700715672,
      "loss": 6.8117,
      "mean_token_accuracy": 0.14246365278959275,
      "num_tokens": 5614308.0,
      "step": 2215
    },
    {
      "entropy": 6.882238912582397,
      "epoch": 0.25620311598384304,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004997751294456892,
      "loss": 6.7753,
      "mean_token_accuracy": 0.13991254717111587,
      "num_tokens": 5626351.0,
      "step": 2220
    },
    {
      "entropy": 6.902230548858642,
      "epoch": 0.2567801500288517,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997732812596114,
      "loss": 6.7473,
      "mean_token_accuracy": 0.1505092367529869,
      "num_tokens": 5638339.0,
      "step": 2225
    },
    {
      "entropy": 6.901751756668091,
      "epoch": 0.2573571840738604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997714255133961,
      "loss": 6.8054,
      "mean_token_accuracy": 0.14526830166578292,
      "num_tokens": 5651210.0,
      "step": 2230
    },
    {
      "entropy": 6.877227878570556,
      "epoch": 0.257934218118869,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997695622071054,
      "loss": 6.7704,
      "mean_token_accuracy": 0.14814679324626923,
      "num_tokens": 5664150.0,
      "step": 2235
    },
    {
      "entropy": 6.898557186126709,
      "epoch": 0.25851125216387766,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997676913408021,
      "loss": 6.7121,
      "mean_token_accuracy": 0.15328368097543715,
      "num_tokens": 5676009.0,
      "step": 2240
    },
    {
      "entropy": 6.965892267227173,
      "epoch": 0.2590882862088863,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997658129145493,
      "loss": 6.8838,
      "mean_token_accuracy": 0.14226163998246194,
      "num_tokens": 5688557.0,
      "step": 2245
    },
    {
      "entropy": 6.861143636703491,
      "epoch": 0.259665320253895,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004997639269284099,
      "loss": 6.8308,
      "mean_token_accuracy": 0.1465721994638443,
      "num_tokens": 5702032.0,
      "step": 2250
    },
    {
      "entropy": 6.920127868652344,
      "epoch": 0.26024235429890363,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997620333824476,
      "loss": 6.6695,
      "mean_token_accuracy": 0.15270446389913558,
      "num_tokens": 5714623.0,
      "step": 2255
    },
    {
      "entropy": 6.813676214218139,
      "epoch": 0.2608193883439123,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997601322767258,
      "loss": 6.6871,
      "mean_token_accuracy": 0.1490221619606018,
      "num_tokens": 5727415.0,
      "step": 2260
    },
    {
      "entropy": 6.875842475891114,
      "epoch": 0.26139642238892097,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997582236113087,
      "loss": 6.7316,
      "mean_token_accuracy": 0.14756003841757775,
      "num_tokens": 5740129.0,
      "step": 2265
    },
    {
      "entropy": 6.854972076416016,
      "epoch": 0.2619734564339296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997563073862603,
      "loss": 6.6885,
      "mean_token_accuracy": 0.15406904220581055,
      "num_tokens": 5752469.0,
      "step": 2270
    },
    {
      "entropy": 6.81607403755188,
      "epoch": 0.26255049047893825,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997543836016453,
      "loss": 6.7811,
      "mean_token_accuracy": 0.15202534720301628,
      "num_tokens": 5765408.0,
      "step": 2275
    },
    {
      "entropy": 6.929606485366821,
      "epoch": 0.2631275245239469,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997524522575281,
      "loss": 6.705,
      "mean_token_accuracy": 0.1446349136531353,
      "num_tokens": 5777450.0,
      "step": 2280
    },
    {
      "entropy": 6.859155130386353,
      "epoch": 0.2637045585689556,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997505133539738,
      "loss": 6.76,
      "mean_token_accuracy": 0.14340940490365028,
      "num_tokens": 5790435.0,
      "step": 2285
    },
    {
      "entropy": 6.930182981491089,
      "epoch": 0.2642815926139642,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997485668910476,
      "loss": 6.7483,
      "mean_token_accuracy": 0.14692381769418716,
      "num_tokens": 5803517.0,
      "step": 2290
    },
    {
      "entropy": 6.785097646713257,
      "epoch": 0.26485862665897286,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997466128688151,
      "loss": 6.7016,
      "mean_token_accuracy": 0.15636104047298433,
      "num_tokens": 5815802.0,
      "step": 2295
    },
    {
      "entropy": 6.829444026947021,
      "epoch": 0.26543566070398156,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499744651287342,
      "loss": 6.7371,
      "mean_token_accuracy": 0.1466741144657135,
      "num_tokens": 5827807.0,
      "step": 2300
    },
    {
      "entropy": 6.863865900039673,
      "epoch": 0.2660126947489902,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499742682146694,
      "loss": 6.7391,
      "mean_token_accuracy": 0.14884627610445023,
      "num_tokens": 5840759.0,
      "step": 2305
    },
    {
      "entropy": 6.915482997894287,
      "epoch": 0.26658972879399884,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997407054469377,
      "loss": 6.7772,
      "mean_token_accuracy": 0.14534752815961838,
      "num_tokens": 5853117.0,
      "step": 2310
    },
    {
      "entropy": 6.74487190246582,
      "epoch": 0.2671667628390075,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997387211881392,
      "loss": 6.6864,
      "mean_token_accuracy": 0.15353853702545167,
      "num_tokens": 5866717.0,
      "step": 2315
    },
    {
      "entropy": 6.883811712265015,
      "epoch": 0.26774379688401617,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997367293703656,
      "loss": 6.7412,
      "mean_token_accuracy": 0.14999326467514038,
      "num_tokens": 5879017.0,
      "step": 2320
    },
    {
      "entropy": 6.83184404373169,
      "epoch": 0.2683208309290248,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997347299936837,
      "loss": 6.7185,
      "mean_token_accuracy": 0.14858803004026414,
      "num_tokens": 5891481.0,
      "step": 2325
    },
    {
      "entropy": 6.848265933990478,
      "epoch": 0.26889786497403345,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997327230581608,
      "loss": 6.7359,
      "mean_token_accuracy": 0.1505471758544445,
      "num_tokens": 5903389.0,
      "step": 2330
    },
    {
      "entropy": 6.828605127334595,
      "epoch": 0.26947489901904215,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997307085638644,
      "loss": 6.6471,
      "mean_token_accuracy": 0.14687670022249222,
      "num_tokens": 5916084.0,
      "step": 2335
    },
    {
      "entropy": 6.8207825183868405,
      "epoch": 0.2700519330640508,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004997286865108621,
      "loss": 6.6947,
      "mean_token_accuracy": 0.14965061396360396,
      "num_tokens": 5927797.0,
      "step": 2340
    },
    {
      "entropy": 6.725537252426148,
      "epoch": 0.2706289671090594,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997266568992222,
      "loss": 6.602,
      "mean_token_accuracy": 0.15474189966917037,
      "num_tokens": 5941886.0,
      "step": 2345
    },
    {
      "entropy": 6.837874221801758,
      "epoch": 0.27120600115406807,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004997246197290128,
      "loss": 6.7189,
      "mean_token_accuracy": 0.1519598752260208,
      "num_tokens": 5953655.0,
      "step": 2350
    },
    {
      "entropy": 6.989822483062744,
      "epoch": 0.27178303519907676,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997225750003024,
      "loss": 6.7316,
      "mean_token_accuracy": 0.14648340493440629,
      "num_tokens": 5967818.0,
      "step": 2355
    },
    {
      "entropy": 6.777784395217895,
      "epoch": 0.2723600692440854,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997205227131599,
      "loss": 6.6877,
      "mean_token_accuracy": 0.15397463142871856,
      "num_tokens": 5980742.0,
      "step": 2360
    },
    {
      "entropy": 6.8580399513244625,
      "epoch": 0.27293710328909404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997184628676542,
      "loss": 6.7001,
      "mean_token_accuracy": 0.1460026979446411,
      "num_tokens": 5993263.0,
      "step": 2365
    },
    {
      "entropy": 6.879456043243408,
      "epoch": 0.27351413733410274,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997163954638546,
      "loss": 6.794,
      "mean_token_accuracy": 0.13860531225800515,
      "num_tokens": 6007820.0,
      "step": 2370
    },
    {
      "entropy": 6.88253002166748,
      "epoch": 0.2740911713791114,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997143205018306,
      "loss": 6.736,
      "mean_token_accuracy": 0.14723660051822662,
      "num_tokens": 6020665.0,
      "step": 2375
    },
    {
      "entropy": 6.853304862976074,
      "epoch": 0.27466820542412,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997122379816523,
      "loss": 6.6706,
      "mean_token_accuracy": 0.15280609726905822,
      "num_tokens": 6032847.0,
      "step": 2380
    },
    {
      "entropy": 6.813856410980224,
      "epoch": 0.27524523946912866,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997101479033893,
      "loss": 6.7285,
      "mean_token_accuracy": 0.14313747137784957,
      "num_tokens": 6045291.0,
      "step": 2385
    },
    {
      "entropy": 6.841581058502197,
      "epoch": 0.27582227351413735,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997080502671122,
      "loss": 6.7033,
      "mean_token_accuracy": 0.14931570738554,
      "num_tokens": 6057872.0,
      "step": 2390
    },
    {
      "entropy": 6.79990816116333,
      "epoch": 0.276399307559146,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997059450728913,
      "loss": 6.6866,
      "mean_token_accuracy": 0.15096329003572465,
      "num_tokens": 6070252.0,
      "step": 2395
    },
    {
      "entropy": 6.918078327178955,
      "epoch": 0.27697634160415463,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997038323207977,
      "loss": 6.7697,
      "mean_token_accuracy": 0.14049090892076493,
      "num_tokens": 6083728.0,
      "step": 2400
    },
    {
      "entropy": 6.816164302825928,
      "epoch": 0.2775533756491633,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997017120109022,
      "loss": 6.7161,
      "mean_token_accuracy": 0.14372968971729277,
      "num_tokens": 6097140.0,
      "step": 2405
    },
    {
      "entropy": 6.773523044586182,
      "epoch": 0.27813040969417197,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996995841432762,
      "loss": 6.6932,
      "mean_token_accuracy": 0.14606269001960753,
      "num_tokens": 6109518.0,
      "step": 2410
    },
    {
      "entropy": 6.870539903640747,
      "epoch": 0.2787074437391806,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996974487179915,
      "loss": 6.6722,
      "mean_token_accuracy": 0.1518884226679802,
      "num_tokens": 6122160.0,
      "step": 2415
    },
    {
      "entropy": 6.702911806106568,
      "epoch": 0.27928447778418924,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996953057351195,
      "loss": 6.7244,
      "mean_token_accuracy": 0.14826845824718476,
      "num_tokens": 6134638.0,
      "step": 2420
    },
    {
      "entropy": 6.88034405708313,
      "epoch": 0.27986151182919794,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996931551947327,
      "loss": 6.7612,
      "mean_token_accuracy": 0.14381736442446708,
      "num_tokens": 6147142.0,
      "step": 2425
    },
    {
      "entropy": 6.940686750411987,
      "epoch": 0.2804385458742066,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996909970969031,
      "loss": 6.7526,
      "mean_token_accuracy": 0.1503518268465996,
      "num_tokens": 6159137.0,
      "step": 2430
    },
    {
      "entropy": 6.719251680374145,
      "epoch": 0.2810155799192152,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996888314417034,
      "loss": 6.7423,
      "mean_token_accuracy": 0.1468128316104412,
      "num_tokens": 6171663.0,
      "step": 2435
    },
    {
      "entropy": 6.958425569534302,
      "epoch": 0.2815926139642239,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996866582292067,
      "loss": 6.7465,
      "mean_token_accuracy": 0.15097531378269197,
      "num_tokens": 6184282.0,
      "step": 2440
    },
    {
      "entropy": 6.830920457839966,
      "epoch": 0.28216964800923255,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996844774594856,
      "loss": 6.7386,
      "mean_token_accuracy": 0.14542510211467743,
      "num_tokens": 6197243.0,
      "step": 2445
    },
    {
      "entropy": 6.791044044494629,
      "epoch": 0.2827466820542412,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996822891326138,
      "loss": 6.6572,
      "mean_token_accuracy": 0.14865050166845323,
      "num_tokens": 6209061.0,
      "step": 2450
    },
    {
      "entropy": 6.817491722106934,
      "epoch": 0.28332371609924983,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004996800932486648,
      "loss": 6.641,
      "mean_token_accuracy": 0.1513649955391884,
      "num_tokens": 6223497.0,
      "step": 2455
    },
    {
      "entropy": 6.7723153591156,
      "epoch": 0.28390075014425853,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996778898077126,
      "loss": 6.6669,
      "mean_token_accuracy": 0.15665484070777894,
      "num_tokens": 6236395.0,
      "step": 2460
    },
    {
      "entropy": 6.745643424987793,
      "epoch": 0.28447778418926717,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996756788098309,
      "loss": 6.6425,
      "mean_token_accuracy": 0.1522977739572525,
      "num_tokens": 6248700.0,
      "step": 2465
    },
    {
      "entropy": 6.87175989151001,
      "epoch": 0.2850548182342758,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004996734602550945,
      "loss": 6.6899,
      "mean_token_accuracy": 0.14756119549274443,
      "num_tokens": 6262728.0,
      "step": 2470
    },
    {
      "entropy": 6.705410718917847,
      "epoch": 0.2856318522792845,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996712341435779,
      "loss": 6.6553,
      "mean_token_accuracy": 0.14883957505226136,
      "num_tokens": 6276007.0,
      "step": 2475
    },
    {
      "entropy": 6.898572158813477,
      "epoch": 0.28620888632429314,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996690004753559,
      "loss": 6.7871,
      "mean_token_accuracy": 0.14304396212100984,
      "num_tokens": 6289033.0,
      "step": 2480
    },
    {
      "entropy": 6.798679208755493,
      "epoch": 0.2867859203693018,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996667592505037,
      "loss": 6.7183,
      "mean_token_accuracy": 0.14170958548784257,
      "num_tokens": 6300994.0,
      "step": 2485
    },
    {
      "entropy": 6.767846488952637,
      "epoch": 0.2873629544143104,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996645104690967,
      "loss": 6.6763,
      "mean_token_accuracy": 0.15046601742506027,
      "num_tokens": 6313447.0,
      "step": 2490
    },
    {
      "entropy": 6.8288219451904295,
      "epoch": 0.2879399884593191,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996622541312103,
      "loss": 6.6737,
      "mean_token_accuracy": 0.15016402155160904,
      "num_tokens": 6326622.0,
      "step": 2495
    },
    {
      "entropy": 6.799957323074341,
      "epoch": 0.28851702250432776,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996599902369207,
      "loss": 6.6845,
      "mean_token_accuracy": 0.14732487350702286,
      "num_tokens": 6339286.0,
      "step": 2500
    },
    {
      "entropy": 6.783337116241455,
      "epoch": 0.2890940565493364,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996577187863039,
      "loss": 6.6882,
      "mean_token_accuracy": 0.14488342702388762,
      "num_tokens": 6351869.0,
      "step": 2505
    },
    {
      "entropy": 6.761418199539184,
      "epoch": 0.2896710905943451,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996554397794364,
      "loss": 6.623,
      "mean_token_accuracy": 0.1587457537651062,
      "num_tokens": 6363974.0,
      "step": 2510
    },
    {
      "entropy": 6.773473787307739,
      "epoch": 0.29024812463935373,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996531532163947,
      "loss": 6.6119,
      "mean_token_accuracy": 0.1547771155834198,
      "num_tokens": 6376569.0,
      "step": 2515
    },
    {
      "entropy": 6.77434344291687,
      "epoch": 0.2908251586843624,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996508590972558,
      "loss": 6.6397,
      "mean_token_accuracy": 0.15693849921226502,
      "num_tokens": 6388830.0,
      "step": 2520
    },
    {
      "entropy": 6.750413084030152,
      "epoch": 0.291402192729371,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996485574220969,
      "loss": 6.6441,
      "mean_token_accuracy": 0.14794425517320633,
      "num_tokens": 6401542.0,
      "step": 2525
    },
    {
      "entropy": 6.694020843505859,
      "epoch": 0.2919792267743797,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996462481909953,
      "loss": 6.6248,
      "mean_token_accuracy": 0.1506824940443039,
      "num_tokens": 6414123.0,
      "step": 2530
    },
    {
      "entropy": 6.8416389465332035,
      "epoch": 0.29255626081938835,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004996439314040287,
      "loss": 6.6858,
      "mean_token_accuracy": 0.14844510704278946,
      "num_tokens": 6427269.0,
      "step": 2535
    },
    {
      "entropy": 6.79074239730835,
      "epoch": 0.293133294864397,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499641607061275,
      "loss": 6.758,
      "mean_token_accuracy": 0.1483464166522026,
      "num_tokens": 6442296.0,
      "step": 2540
    },
    {
      "entropy": 6.821115922927857,
      "epoch": 0.2937103289094057,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996392751628126,
      "loss": 6.7614,
      "mean_token_accuracy": 0.14189203828573227,
      "num_tokens": 6456954.0,
      "step": 2545
    },
    {
      "entropy": 6.866867971420288,
      "epoch": 0.2942873629544143,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996369357087196,
      "loss": 6.7358,
      "mean_token_accuracy": 0.14389990866184235,
      "num_tokens": 6470139.0,
      "step": 2550
    },
    {
      "entropy": 6.869349193572998,
      "epoch": 0.29486439699942296,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499634588699075,
      "loss": 6.6721,
      "mean_token_accuracy": 0.15112345814704894,
      "num_tokens": 6483106.0,
      "step": 2555
    },
    {
      "entropy": 6.740794134140015,
      "epoch": 0.2954414310444316,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004996322341339575,
      "loss": 6.7031,
      "mean_token_accuracy": 0.1481093369424343,
      "num_tokens": 6495921.0,
      "step": 2560
    },
    {
      "entropy": 6.757239103317261,
      "epoch": 0.2960184650894403,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996298720134465,
      "loss": 6.5961,
      "mean_token_accuracy": 0.1495403155684471,
      "num_tokens": 6508408.0,
      "step": 2565
    },
    {
      "entropy": 6.712530660629272,
      "epoch": 0.29659549913444894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996275023376213,
      "loss": 6.6843,
      "mean_token_accuracy": 0.14357175007462503,
      "num_tokens": 6522081.0,
      "step": 2570
    },
    {
      "entropy": 6.83108491897583,
      "epoch": 0.2971725331794576,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996251251065615,
      "loss": 6.667,
      "mean_token_accuracy": 0.14881591349840165,
      "num_tokens": 6535065.0,
      "step": 2575
    },
    {
      "entropy": 6.745174551010132,
      "epoch": 0.2977495672244662,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996227403203473,
      "loss": 6.6262,
      "mean_token_accuracy": 0.15182467997074128,
      "num_tokens": 6547972.0,
      "step": 2580
    },
    {
      "entropy": 6.76308388710022,
      "epoch": 0.2983266012694749,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004996203479790589,
      "loss": 6.6084,
      "mean_token_accuracy": 0.15488889515399934,
      "num_tokens": 6559935.0,
      "step": 2585
    },
    {
      "entropy": 6.758035564422608,
      "epoch": 0.29890363531448355,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004996179480827764,
      "loss": 6.6027,
      "mean_token_accuracy": 0.15274570137262344,
      "num_tokens": 6572373.0,
      "step": 2590
    },
    {
      "entropy": 6.786394119262695,
      "epoch": 0.2994806693594922,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499615540631581,
      "loss": 6.6898,
      "mean_token_accuracy": 0.15085428357124328,
      "num_tokens": 6585846.0,
      "step": 2595
    },
    {
      "entropy": 6.833403587341309,
      "epoch": 0.3000577034045009,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996131256255534,
      "loss": 6.7514,
      "mean_token_accuracy": 0.14574786573648452,
      "num_tokens": 6600387.0,
      "step": 2600
    },
    {
      "entropy": 6.733997964859009,
      "epoch": 0.3006347374495095,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004996107030647749,
      "loss": 6.6546,
      "mean_token_accuracy": 0.1498239278793335,
      "num_tokens": 6613142.0,
      "step": 2605
    },
    {
      "entropy": 6.719230985641479,
      "epoch": 0.30121177149451817,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996082729493269,
      "loss": 6.576,
      "mean_token_accuracy": 0.15462903380393983,
      "num_tokens": 6626273.0,
      "step": 2610
    },
    {
      "entropy": 6.811122417449951,
      "epoch": 0.3017888055395268,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004996058352792913,
      "loss": 6.6689,
      "mean_token_accuracy": 0.14554012417793274,
      "num_tokens": 6639831.0,
      "step": 2615
    },
    {
      "entropy": 6.734864568710327,
      "epoch": 0.3023658395845355,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049960339005475,
      "loss": 6.7325,
      "mean_token_accuracy": 0.14520045071840287,
      "num_tokens": 6652776.0,
      "step": 2620
    },
    {
      "entropy": 6.930051994323731,
      "epoch": 0.30294287362954414,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996009372757852,
      "loss": 6.6974,
      "mean_token_accuracy": 0.14814985245466233,
      "num_tokens": 6665465.0,
      "step": 2625
    },
    {
      "entropy": 6.608390140533447,
      "epoch": 0.3035199076745528,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004995984769424795,
      "loss": 6.5622,
      "mean_token_accuracy": 0.16428658366203308,
      "num_tokens": 6677472.0,
      "step": 2630
    },
    {
      "entropy": 6.680166530609131,
      "epoch": 0.3040969417195615,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995960090549155,
      "loss": 6.5869,
      "mean_token_accuracy": 0.15190582871437072,
      "num_tokens": 6689145.0,
      "step": 2635
    },
    {
      "entropy": 6.766489934921265,
      "epoch": 0.3046739757645701,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995935336131764,
      "loss": 6.5804,
      "mean_token_accuracy": 0.15143152326345444,
      "num_tokens": 6701520.0,
      "step": 2640
    },
    {
      "entropy": 6.769906234741211,
      "epoch": 0.30525100980957875,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004995910506173452,
      "loss": 6.6662,
      "mean_token_accuracy": 0.1508184403181076,
      "num_tokens": 6715615.0,
      "step": 2645
    },
    {
      "entropy": 6.837305879592895,
      "epoch": 0.3058280438545874,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995885600675059,
      "loss": 6.6889,
      "mean_token_accuracy": 0.14585935696959496,
      "num_tokens": 6728772.0,
      "step": 2650
    },
    {
      "entropy": 6.781810855865478,
      "epoch": 0.3064050778995961,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995860619637415,
      "loss": 6.6319,
      "mean_token_accuracy": 0.14923317804932595,
      "num_tokens": 6741454.0,
      "step": 2655
    },
    {
      "entropy": 6.709044456481934,
      "epoch": 0.30698211194460473,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004995835563061367,
      "loss": 6.6654,
      "mean_token_accuracy": 0.14947767108678817,
      "num_tokens": 6753983.0,
      "step": 2660
    },
    {
      "entropy": 6.855918884277344,
      "epoch": 0.30755914598961337,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995810430947756,
      "loss": 6.7013,
      "mean_token_accuracy": 0.14999711960554124,
      "num_tokens": 6767377.0,
      "step": 2665
    },
    {
      "entropy": 6.7711738586425785,
      "epoch": 0.30813618003462206,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995785223297426,
      "loss": 6.6499,
      "mean_token_accuracy": 0.15468411445617675,
      "num_tokens": 6779220.0,
      "step": 2670
    },
    {
      "entropy": 6.8247597217559814,
      "epoch": 0.3087132140796307,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004995759940111225,
      "loss": 6.6286,
      "mean_token_accuracy": 0.15276417881250381,
      "num_tokens": 6792202.0,
      "step": 2675
    },
    {
      "entropy": 6.6788218975067135,
      "epoch": 0.30929024812463934,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995734581390005,
      "loss": 6.6376,
      "mean_token_accuracy": 0.15422565937042237,
      "num_tokens": 6804988.0,
      "step": 2680
    },
    {
      "entropy": 6.736396455764771,
      "epoch": 0.309867282169648,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995709147134617,
      "loss": 6.6538,
      "mean_token_accuracy": 0.14916067719459533,
      "num_tokens": 6817311.0,
      "step": 2685
    },
    {
      "entropy": 6.789905023574829,
      "epoch": 0.3104443162146567,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004995683637345919,
      "loss": 6.6367,
      "mean_token_accuracy": 0.1507313832640648,
      "num_tokens": 6829875.0,
      "step": 2690
    },
    {
      "entropy": 6.814420080184936,
      "epoch": 0.3110213502596653,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004995658052024765,
      "loss": 6.6156,
      "mean_token_accuracy": 0.15454477518796922,
      "num_tokens": 6843080.0,
      "step": 2695
    },
    {
      "entropy": 6.697454023361206,
      "epoch": 0.31159838430467396,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995632391172019,
      "loss": 6.6061,
      "mean_token_accuracy": 0.1518676370382309,
      "num_tokens": 6857202.0,
      "step": 2700
    },
    {
      "entropy": 6.774089813232422,
      "epoch": 0.31217541834968265,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004995606654788543,
      "loss": 6.6958,
      "mean_token_accuracy": 0.14331221878528594,
      "num_tokens": 6870801.0,
      "step": 2705
    },
    {
      "entropy": 6.790430974960327,
      "epoch": 0.3127524523946913,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004995580842875202,
      "loss": 6.642,
      "mean_token_accuracy": 0.1496342271566391,
      "num_tokens": 6883084.0,
      "step": 2710
    },
    {
      "entropy": 6.821096277236938,
      "epoch": 0.31332948643969993,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995554955432866,
      "loss": 6.6693,
      "mean_token_accuracy": 0.14794540405273438,
      "num_tokens": 6895309.0,
      "step": 2715
    },
    {
      "entropy": 6.7353767395019535,
      "epoch": 0.3139065204847086,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995528992462404,
      "loss": 6.6427,
      "mean_token_accuracy": 0.15779228061437606,
      "num_tokens": 6907646.0,
      "step": 2720
    },
    {
      "entropy": 6.766403102874756,
      "epoch": 0.31448355452971727,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499550295396469,
      "loss": 6.6185,
      "mean_token_accuracy": 0.1583697885274887,
      "num_tokens": 6919769.0,
      "step": 2725
    },
    {
      "entropy": 6.680029249191284,
      "epoch": 0.3150605885747259,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995476839940598,
      "loss": 6.581,
      "mean_token_accuracy": 0.15339329093694687,
      "num_tokens": 6933348.0,
      "step": 2730
    },
    {
      "entropy": 6.785920238494873,
      "epoch": 0.31563762261973455,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000499545065039101,
      "loss": 6.6585,
      "mean_token_accuracy": 0.14711489379405976,
      "num_tokens": 6945105.0,
      "step": 2735
    },
    {
      "entropy": 6.736616373062134,
      "epoch": 0.31621465666474324,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995424385316803,
      "loss": 6.6561,
      "mean_token_accuracy": 0.14591162502765656,
      "num_tokens": 6958009.0,
      "step": 2740
    },
    {
      "entropy": 6.748479509353638,
      "epoch": 0.3167916907097519,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995398044718863,
      "loss": 6.5594,
      "mean_token_accuracy": 0.15630290359258653,
      "num_tokens": 6970156.0,
      "step": 2745
    },
    {
      "entropy": 6.710201978683472,
      "epoch": 0.3173687247547605,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004995371628598074,
      "loss": 6.6478,
      "mean_token_accuracy": 0.15026513785123824,
      "num_tokens": 6983431.0,
      "step": 2750
    },
    {
      "entropy": 6.806408214569092,
      "epoch": 0.31794575879976916,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995345136955328,
      "loss": 6.6938,
      "mean_token_accuracy": 0.14544053599238396,
      "num_tokens": 6997470.0,
      "step": 2755
    },
    {
      "entropy": 6.746524286270142,
      "epoch": 0.31852279284477786,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995318569791512,
      "loss": 6.6371,
      "mean_token_accuracy": 0.15571504682302476,
      "num_tokens": 7010337.0,
      "step": 2760
    },
    {
      "entropy": 6.8036500930786135,
      "epoch": 0.3190998268897865,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004995291927107522,
      "loss": 6.6618,
      "mean_token_accuracy": 0.1454496204853058,
      "num_tokens": 7023950.0,
      "step": 2765
    },
    {
      "entropy": 6.689703798294067,
      "epoch": 0.31967686093479514,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995265208904252,
      "loss": 6.6302,
      "mean_token_accuracy": 0.15477116107940675,
      "num_tokens": 7036409.0,
      "step": 2770
    },
    {
      "entropy": 6.779079484939575,
      "epoch": 0.32025389497980383,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995238415182603,
      "loss": 6.6425,
      "mean_token_accuracy": 0.1460179179906845,
      "num_tokens": 7049409.0,
      "step": 2775
    },
    {
      "entropy": 6.785402870178222,
      "epoch": 0.3208309290248125,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995211545943476,
      "loss": 6.6862,
      "mean_token_accuracy": 0.15174451917409898,
      "num_tokens": 7062120.0,
      "step": 2780
    },
    {
      "entropy": 6.77389817237854,
      "epoch": 0.3214079630698211,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995184601187772,
      "loss": 6.6715,
      "mean_token_accuracy": 0.1446915090084076,
      "num_tokens": 7073150.0,
      "step": 2785
    },
    {
      "entropy": 6.756207847595215,
      "epoch": 0.32198499711482975,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049951575809164,
      "loss": 6.6123,
      "mean_token_accuracy": 0.14350593835115433,
      "num_tokens": 7085927.0,
      "step": 2790
    },
    {
      "entropy": 6.743623447418213,
      "epoch": 0.32256203115983845,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995130485130269,
      "loss": 6.6799,
      "mean_token_accuracy": 0.1505696654319763,
      "num_tokens": 7097607.0,
      "step": 2795
    },
    {
      "entropy": 6.801728105545044,
      "epoch": 0.3231390652048471,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995103313830289,
      "loss": 6.591,
      "mean_token_accuracy": 0.1520095705986023,
      "num_tokens": 7109771.0,
      "step": 2800
    },
    {
      "entropy": 6.725235414505005,
      "epoch": 0.3237160992498557,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995076067017373,
      "loss": 6.6574,
      "mean_token_accuracy": 0.14673417136073114,
      "num_tokens": 7122757.0,
      "step": 2805
    },
    {
      "entropy": 6.721699571609497,
      "epoch": 0.3242931332948644,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004995048744692439,
      "loss": 6.6415,
      "mean_token_accuracy": 0.14944978207349777,
      "num_tokens": 7135516.0,
      "step": 2810
    },
    {
      "entropy": 6.755308294296265,
      "epoch": 0.32487016733987306,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995021346856405,
      "loss": 6.7226,
      "mean_token_accuracy": 0.1461216315627098,
      "num_tokens": 7149125.0,
      "step": 2815
    },
    {
      "entropy": 6.745944166183472,
      "epoch": 0.3254472013848817,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994993873510196,
      "loss": 6.6727,
      "mean_token_accuracy": 0.14473605155944824,
      "num_tokens": 7160723.0,
      "step": 2820
    },
    {
      "entropy": 6.771013927459717,
      "epoch": 0.32602423542989034,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994966324654731,
      "loss": 6.6407,
      "mean_token_accuracy": 0.14543287828564644,
      "num_tokens": 7174551.0,
      "step": 2825
    },
    {
      "entropy": 6.802860164642334,
      "epoch": 0.32660126947489904,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499493870029094,
      "loss": 6.6193,
      "mean_token_accuracy": 0.14886424392461778,
      "num_tokens": 7186764.0,
      "step": 2830
    },
    {
      "entropy": 6.7293703079223635,
      "epoch": 0.3271783035199077,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994911000419749,
      "loss": 6.6135,
      "mean_token_accuracy": 0.1525777131319046,
      "num_tokens": 7199645.0,
      "step": 2835
    },
    {
      "entropy": 6.685914325714111,
      "epoch": 0.3277553375649163,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994883225042093,
      "loss": 6.5341,
      "mean_token_accuracy": 0.16299870759248733,
      "num_tokens": 7212579.0,
      "step": 2840
    },
    {
      "entropy": 6.696114015579224,
      "epoch": 0.328332371609925,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994855374158905,
      "loss": 6.5716,
      "mean_token_accuracy": 0.15522724390029907,
      "num_tokens": 7225139.0,
      "step": 2845
    },
    {
      "entropy": 6.791204595565796,
      "epoch": 0.32890940565493365,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000499482744777112,
      "loss": 6.7228,
      "mean_token_accuracy": 0.142743781208992,
      "num_tokens": 7239220.0,
      "step": 2850
    },
    {
      "entropy": 6.822061014175415,
      "epoch": 0.3294864396999423,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499479944587968,
      "loss": 6.7418,
      "mean_token_accuracy": 0.1484747812151909,
      "num_tokens": 7252613.0,
      "step": 2855
    },
    {
      "entropy": 6.678646373748779,
      "epoch": 0.33006347374495093,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994771368485526,
      "loss": 6.5736,
      "mean_token_accuracy": 0.15565900206565858,
      "num_tokens": 7263658.0,
      "step": 2860
    },
    {
      "entropy": 6.7811792373657225,
      "epoch": 0.3306405077899596,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994743215589602,
      "loss": 6.5885,
      "mean_token_accuracy": 0.1597794234752655,
      "num_tokens": 7275782.0,
      "step": 2865
    },
    {
      "entropy": 6.614157962799072,
      "epoch": 0.33121754183496827,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994714987192854,
      "loss": 6.6449,
      "mean_token_accuracy": 0.15721748024225235,
      "num_tokens": 7288536.0,
      "step": 2870
    },
    {
      "entropy": 6.73630838394165,
      "epoch": 0.3317945758799769,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994686683296234,
      "loss": 6.6077,
      "mean_token_accuracy": 0.15456028878688813,
      "num_tokens": 7300480.0,
      "step": 2875
    },
    {
      "entropy": 6.709596300125122,
      "epoch": 0.3323716099249856,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004994658303900691,
      "loss": 6.6456,
      "mean_token_accuracy": 0.15098604261875154,
      "num_tokens": 7313168.0,
      "step": 2880
    },
    {
      "entropy": 6.666839075088501,
      "epoch": 0.33294864396999424,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004994629849007182,
      "loss": 6.6031,
      "mean_token_accuracy": 0.15539678037166596,
      "num_tokens": 7327347.0,
      "step": 2885
    },
    {
      "entropy": 6.813637113571167,
      "epoch": 0.3335256780150029,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994601318616663,
      "loss": 6.6433,
      "mean_token_accuracy": 0.1485406205058098,
      "num_tokens": 7339955.0,
      "step": 2890
    },
    {
      "entropy": 6.741922521591187,
      "epoch": 0.3341027120600115,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994572712730092,
      "loss": 6.6214,
      "mean_token_accuracy": 0.15295621007680893,
      "num_tokens": 7353586.0,
      "step": 2895
    },
    {
      "entropy": 6.6835249900817875,
      "epoch": 0.3346797461050202,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994544031348434,
      "loss": 6.6024,
      "mean_token_accuracy": 0.152432219684124,
      "num_tokens": 7365802.0,
      "step": 2900
    },
    {
      "entropy": 6.729006338119507,
      "epoch": 0.33525678015002885,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994515274472653,
      "loss": 6.5619,
      "mean_token_accuracy": 0.14667018949985505,
      "num_tokens": 7377864.0,
      "step": 2905
    },
    {
      "entropy": 6.67462158203125,
      "epoch": 0.3358338141950375,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994486442103715,
      "loss": 6.5129,
      "mean_token_accuracy": 0.15633828938007355,
      "num_tokens": 7389921.0,
      "step": 2910
    },
    {
      "entropy": 6.713885641098022,
      "epoch": 0.3364108482400462,
      "grad_norm": 0.78125,
      "learning_rate": 0.000499445753424259,
      "loss": 6.5734,
      "mean_token_accuracy": 0.15145887583494186,
      "num_tokens": 7402765.0,
      "step": 2915
    },
    {
      "entropy": 6.767501354217529,
      "epoch": 0.33698788228505483,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004994428550890251,
      "loss": 6.5914,
      "mean_token_accuracy": 0.14778004884719848,
      "num_tokens": 7416330.0,
      "step": 2920
    },
    {
      "entropy": 6.617673254013061,
      "epoch": 0.33756491633006347,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004994399492047672,
      "loss": 6.5484,
      "mean_token_accuracy": 0.16363891512155532,
      "num_tokens": 7428701.0,
      "step": 2925
    },
    {
      "entropy": 6.790898227691651,
      "epoch": 0.3381419503750721,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994370357715832,
      "loss": 6.5492,
      "mean_token_accuracy": 0.1558838188648224,
      "num_tokens": 7440687.0,
      "step": 2930
    },
    {
      "entropy": 6.691900014877319,
      "epoch": 0.3387189844200808,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499434114789571,
      "loss": 6.6532,
      "mean_token_accuracy": 0.14500061124563218,
      "num_tokens": 7454448.0,
      "step": 2935
    },
    {
      "entropy": 6.795220136642456,
      "epoch": 0.33929601846508944,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994311862588286,
      "loss": 6.5717,
      "mean_token_accuracy": 0.15572112798690796,
      "num_tokens": 7466722.0,
      "step": 2940
    },
    {
      "entropy": 6.66687707901001,
      "epoch": 0.3398730525100981,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994282501794551,
      "loss": 6.5761,
      "mean_token_accuracy": 0.1562015950679779,
      "num_tokens": 7478574.0,
      "step": 2945
    },
    {
      "entropy": 6.6353167533874515,
      "epoch": 0.3404500865551068,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994253065515486,
      "loss": 6.4752,
      "mean_token_accuracy": 0.15674795657396318,
      "num_tokens": 7490424.0,
      "step": 2950
    },
    {
      "entropy": 6.738166666030883,
      "epoch": 0.3410271206001154,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994223553752084,
      "loss": 6.6116,
      "mean_token_accuracy": 0.1547369971871376,
      "num_tokens": 7502832.0,
      "step": 2955
    },
    {
      "entropy": 6.6372476577758786,
      "epoch": 0.34160415464512406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994193966505339,
      "loss": 6.5617,
      "mean_token_accuracy": 0.15487824976444245,
      "num_tokens": 7514151.0,
      "step": 2960
    },
    {
      "entropy": 6.712488985061645,
      "epoch": 0.3421811886901327,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994164303776244,
      "loss": 6.5853,
      "mean_token_accuracy": 0.15818527191877366,
      "num_tokens": 7526877.0,
      "step": 2965
    },
    {
      "entropy": 6.701138973236084,
      "epoch": 0.3427582227351414,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994134565565797,
      "loss": 6.551,
      "mean_token_accuracy": 0.15412139296531677,
      "num_tokens": 7538858.0,
      "step": 2970
    },
    {
      "entropy": 6.767053508758545,
      "epoch": 0.34333525678015003,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994104751875,
      "loss": 6.5921,
      "mean_token_accuracy": 0.154334969073534,
      "num_tokens": 7550804.0,
      "step": 2975
    },
    {
      "entropy": 6.637975883483887,
      "epoch": 0.3439122908251587,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994074862704854,
      "loss": 6.5323,
      "mean_token_accuracy": 0.1576222226023674,
      "num_tokens": 7562269.0,
      "step": 2980
    },
    {
      "entropy": 6.650318717956543,
      "epoch": 0.3444893248701673,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994044898056363,
      "loss": 6.6225,
      "mean_token_accuracy": 0.14792083203792572,
      "num_tokens": 7573695.0,
      "step": 2985
    },
    {
      "entropy": 6.7255530834198,
      "epoch": 0.345066358915176,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994014857930538,
      "loss": 6.6233,
      "mean_token_accuracy": 0.14724487364292144,
      "num_tokens": 7585943.0,
      "step": 2990
    },
    {
      "entropy": 6.695814752578736,
      "epoch": 0.34564339296018465,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993984742328387,
      "loss": 6.6071,
      "mean_token_accuracy": 0.15086225420236588,
      "num_tokens": 7598326.0,
      "step": 2995
    },
    {
      "entropy": 6.731552076339722,
      "epoch": 0.3462204270051933,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993954551250925,
      "loss": 6.6033,
      "mean_token_accuracy": 0.148460191488266,
      "num_tokens": 7609950.0,
      "step": 3000
    },
    {
      "epoch": 0.3462204270051933,
      "eval_entropy": 6.558329563233064,
      "eval_loss": 6.633069038391113,
      "eval_mean_token_accuracy": 0.15427329348342508,
      "eval_num_tokens": 7609950.0,
      "eval_runtime": 17.9419,
      "eval_samples_per_second": 1568.171,
      "eval_steps_per_second": 196.021,
      "step": 3000
    },
    {
      "entropy": 6.737070798873901,
      "epoch": 0.346797461050202,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993924284699165,
      "loss": 6.5596,
      "mean_token_accuracy": 0.15286604613065718,
      "num_tokens": 7621987.0,
      "step": 3005
    },
    {
      "entropy": 6.6897971630096436,
      "epoch": 0.3473744950952106,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993893942674127,
      "loss": 6.6225,
      "mean_token_accuracy": 0.1466008722782135,
      "num_tokens": 7634801.0,
      "step": 3010
    },
    {
      "entropy": 6.778344202041626,
      "epoch": 0.34795152914021926,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499386352517683,
      "loss": 6.5567,
      "mean_token_accuracy": 0.15949935764074324,
      "num_tokens": 7647942.0,
      "step": 3015
    },
    {
      "entropy": 6.760732507705688,
      "epoch": 0.3485285631852279,
      "grad_norm": 0.828125,
      "learning_rate": 0.00049938330322083,
      "loss": 6.5854,
      "mean_token_accuracy": 0.1525387793779373,
      "num_tokens": 7662361.0,
      "step": 3020
    },
    {
      "entropy": 6.656428384780884,
      "epoch": 0.3491055972302366,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993802463769558,
      "loss": 6.5719,
      "mean_token_accuracy": 0.15370852500200272,
      "num_tokens": 7674161.0,
      "step": 3025
    },
    {
      "entropy": 6.739625692367554,
      "epoch": 0.34968263127524524,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993771819861636,
      "loss": 6.5816,
      "mean_token_accuracy": 0.15462501496076583,
      "num_tokens": 7686957.0,
      "step": 3030
    },
    {
      "entropy": 6.640616703033447,
      "epoch": 0.3502596653202539,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004993741100485564,
      "loss": 6.5504,
      "mean_token_accuracy": 0.15630083084106444,
      "num_tokens": 7699436.0,
      "step": 3035
    },
    {
      "entropy": 6.641919851303101,
      "epoch": 0.35083669936526257,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993710305642374,
      "loss": 6.5503,
      "mean_token_accuracy": 0.15547990351915358,
      "num_tokens": 7711794.0,
      "step": 3040
    },
    {
      "entropy": 6.740817403793335,
      "epoch": 0.3514137334102712,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993679435333102,
      "loss": 6.5236,
      "mean_token_accuracy": 0.15344731360673905,
      "num_tokens": 7723496.0,
      "step": 3045
    },
    {
      "entropy": 6.6541423320770265,
      "epoch": 0.35199076745527985,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993648489558788,
      "loss": 6.5355,
      "mean_token_accuracy": 0.15317789018154143,
      "num_tokens": 7735980.0,
      "step": 3050
    },
    {
      "entropy": 6.707408475875854,
      "epoch": 0.3525678015002885,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993617468320471,
      "loss": 6.5927,
      "mean_token_accuracy": 0.15256803184747697,
      "num_tokens": 7750145.0,
      "step": 3055
    },
    {
      "entropy": 6.686557340621948,
      "epoch": 0.3531448355452972,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993586371619193,
      "loss": 6.5914,
      "mean_token_accuracy": 0.15120215117931365,
      "num_tokens": 7762284.0,
      "step": 3060
    },
    {
      "entropy": 6.598657512664795,
      "epoch": 0.3537218695903058,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993555199456004,
      "loss": 6.5485,
      "mean_token_accuracy": 0.16331221014261246,
      "num_tokens": 7776225.0,
      "step": 3065
    },
    {
      "entropy": 6.651039886474609,
      "epoch": 0.35429890363531447,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993523951831949,
      "loss": 6.4857,
      "mean_token_accuracy": 0.16369962245225905,
      "num_tokens": 7789408.0,
      "step": 3070
    },
    {
      "entropy": 6.728009128570557,
      "epoch": 0.35487593768032316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993492628748081,
      "loss": 6.5714,
      "mean_token_accuracy": 0.14850344136357307,
      "num_tokens": 7801824.0,
      "step": 3075
    },
    {
      "entropy": 6.632408380508423,
      "epoch": 0.3554529717253318,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993461230205453,
      "loss": 6.5623,
      "mean_token_accuracy": 0.15774839520454406,
      "num_tokens": 7814167.0,
      "step": 3080
    },
    {
      "entropy": 6.721304559707642,
      "epoch": 0.35603000577034044,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004993429756205121,
      "loss": 6.5776,
      "mean_token_accuracy": 0.154751193523407,
      "num_tokens": 7826736.0,
      "step": 3085
    },
    {
      "entropy": 6.653947114944458,
      "epoch": 0.3566070398153491,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993398206748142,
      "loss": 6.5623,
      "mean_token_accuracy": 0.15578778833150864,
      "num_tokens": 7838626.0,
      "step": 3090
    },
    {
      "entropy": 6.703815984725952,
      "epoch": 0.3571840738603578,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993366581835581,
      "loss": 6.5489,
      "mean_token_accuracy": 0.15328158885240556,
      "num_tokens": 7850608.0,
      "step": 3095
    },
    {
      "entropy": 6.684204244613648,
      "epoch": 0.3577611079053664,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993334881468498,
      "loss": 6.5392,
      "mean_token_accuracy": 0.15394985526800156,
      "num_tokens": 7862800.0,
      "step": 3100
    },
    {
      "entropy": 6.7322910785675045,
      "epoch": 0.35833814195037506,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993303105647961,
      "loss": 6.6049,
      "mean_token_accuracy": 0.153105728328228,
      "num_tokens": 7876198.0,
      "step": 3105
    },
    {
      "entropy": 6.737527465820312,
      "epoch": 0.35891517599538375,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993271254375038,
      "loss": 6.6142,
      "mean_token_accuracy": 0.1486457869410515,
      "num_tokens": 7888129.0,
      "step": 3110
    },
    {
      "entropy": 6.712254238128662,
      "epoch": 0.3594922100403924,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00049932393276508,
      "loss": 6.5859,
      "mean_token_accuracy": 0.15163826197385788,
      "num_tokens": 7900453.0,
      "step": 3115
    },
    {
      "entropy": 6.707119369506836,
      "epoch": 0.36006924408540103,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993207325476323,
      "loss": 6.5885,
      "mean_token_accuracy": 0.14777861386537552,
      "num_tokens": 7913538.0,
      "step": 3120
    },
    {
      "entropy": 6.7163959503173825,
      "epoch": 0.36064627813040967,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993175247852681,
      "loss": 6.5554,
      "mean_token_accuracy": 0.15100172013044358,
      "num_tokens": 7927744.0,
      "step": 3125
    },
    {
      "entropy": 6.628836631774902,
      "epoch": 0.36122331217541837,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993143094780954,
      "loss": 6.5067,
      "mean_token_accuracy": 0.15232598334550856,
      "num_tokens": 7939975.0,
      "step": 3130
    },
    {
      "entropy": 6.690788125991821,
      "epoch": 0.361800346220427,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993110866262224,
      "loss": 6.5662,
      "mean_token_accuracy": 0.15564843118190766,
      "num_tokens": 7951692.0,
      "step": 3135
    },
    {
      "entropy": 6.703995895385742,
      "epoch": 0.36237738026543564,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993078562297573,
      "loss": 6.5669,
      "mean_token_accuracy": 0.15299160033464432,
      "num_tokens": 7964887.0,
      "step": 3140
    },
    {
      "entropy": 6.669623613357544,
      "epoch": 0.36295441431044434,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993046182888089,
      "loss": 6.5759,
      "mean_token_accuracy": 0.1529490426182747,
      "num_tokens": 7977608.0,
      "step": 3145
    },
    {
      "entropy": 6.640576410293579,
      "epoch": 0.363531448355453,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993013728034861,
      "loss": 6.5009,
      "mean_token_accuracy": 0.15490880310535432,
      "num_tokens": 7990094.0,
      "step": 3150
    },
    {
      "entropy": 6.738632535934448,
      "epoch": 0.3641084824004616,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000499298119773898,
      "loss": 6.5557,
      "mean_token_accuracy": 0.1515592098236084,
      "num_tokens": 8003477.0,
      "step": 3155
    },
    {
      "entropy": 6.614168548583985,
      "epoch": 0.36468551644547026,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992948592001541,
      "loss": 6.5111,
      "mean_token_accuracy": 0.15605866014957429,
      "num_tokens": 8016241.0,
      "step": 3160
    },
    {
      "entropy": 6.603486204147339,
      "epoch": 0.36526255049047895,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499291591082364,
      "loss": 6.4867,
      "mean_token_accuracy": 0.16270006597042083,
      "num_tokens": 8028352.0,
      "step": 3165
    },
    {
      "entropy": 6.679505729675293,
      "epoch": 0.3658395845354876,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992883154206377,
      "loss": 6.5717,
      "mean_token_accuracy": 0.15018778592348098,
      "num_tokens": 8040215.0,
      "step": 3170
    },
    {
      "entropy": 6.650276947021484,
      "epoch": 0.36641661858049623,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992850322150853,
      "loss": 6.4888,
      "mean_token_accuracy": 0.16015321314334868,
      "num_tokens": 8051633.0,
      "step": 3175
    },
    {
      "entropy": 6.693320226669312,
      "epoch": 0.36699365262550493,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992817414658172,
      "loss": 6.5516,
      "mean_token_accuracy": 0.15389626622200012,
      "num_tokens": 8063390.0,
      "step": 3180
    },
    {
      "entropy": 6.6833244323730465,
      "epoch": 0.36757068667051357,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992784431729442,
      "loss": 6.5321,
      "mean_token_accuracy": 0.15152984708547593,
      "num_tokens": 8077561.0,
      "step": 3185
    },
    {
      "entropy": 6.611824989318848,
      "epoch": 0.3681477207155222,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004992751373365771,
      "loss": 6.595,
      "mean_token_accuracy": 0.15470331758260727,
      "num_tokens": 8091345.0,
      "step": 3190
    },
    {
      "entropy": 6.65063967704773,
      "epoch": 0.36872475476053085,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992718239568273,
      "loss": 6.4521,
      "mean_token_accuracy": 0.1627359613776207,
      "num_tokens": 8104179.0,
      "step": 3195
    },
    {
      "entropy": 6.691599035263062,
      "epoch": 0.36930178880553954,
      "grad_norm": 0.875,
      "learning_rate": 0.000499268503033806,
      "loss": 6.6218,
      "mean_token_accuracy": 0.14577734395861625,
      "num_tokens": 8115831.0,
      "step": 3200
    },
    {
      "entropy": 6.670457363128662,
      "epoch": 0.3698788228505482,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0004992651745676249,
      "loss": 6.5415,
      "mean_token_accuracy": 0.15876328498125075,
      "num_tokens": 8129048.0,
      "step": 3205
    },
    {
      "entropy": 6.703500890731812,
      "epoch": 0.3704558568955568,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992618385583962,
      "loss": 6.5535,
      "mean_token_accuracy": 0.15128162652254104,
      "num_tokens": 8141679.0,
      "step": 3210
    },
    {
      "entropy": 6.644980478286743,
      "epoch": 0.3710328909405655,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992584950062319,
      "loss": 6.5271,
      "mean_token_accuracy": 0.15173182934522628,
      "num_tokens": 8155004.0,
      "step": 3215
    },
    {
      "entropy": 6.607394123077393,
      "epoch": 0.37160992498557416,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992551439112446,
      "loss": 6.4594,
      "mean_token_accuracy": 0.15869889855384828,
      "num_tokens": 8166718.0,
      "step": 3220
    },
    {
      "entropy": 6.69423246383667,
      "epoch": 0.3721869590305828,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992517852735469,
      "loss": 6.6532,
      "mean_token_accuracy": 0.14730547666549682,
      "num_tokens": 8180011.0,
      "step": 3225
    },
    {
      "entropy": 6.697569942474365,
      "epoch": 0.37276399307559144,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992484190932517,
      "loss": 6.4638,
      "mean_token_accuracy": 0.15984451472759248,
      "num_tokens": 8194800.0,
      "step": 3230
    },
    {
      "entropy": 6.6850013256073,
      "epoch": 0.37334102712060013,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992450453704723,
      "loss": 6.5909,
      "mean_token_accuracy": 0.14907579496502876,
      "num_tokens": 8207153.0,
      "step": 3235
    },
    {
      "entropy": 6.701785898208618,
      "epoch": 0.3739180611656088,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992416641053223,
      "loss": 6.5702,
      "mean_token_accuracy": 0.15299123525619507,
      "num_tokens": 8221135.0,
      "step": 3240
    },
    {
      "entropy": 6.574634313583374,
      "epoch": 0.3744950952106174,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992382752979153,
      "loss": 6.4426,
      "mean_token_accuracy": 0.16063585728406907,
      "num_tokens": 8234271.0,
      "step": 3245
    },
    {
      "entropy": 6.676979637145996,
      "epoch": 0.3750721292556261,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992348789483651,
      "loss": 6.5782,
      "mean_token_accuracy": 0.15047473162412645,
      "num_tokens": 8246710.0,
      "step": 3250
    },
    {
      "entropy": 6.649428367614746,
      "epoch": 0.37564916330063475,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004992314750567864,
      "loss": 6.6366,
      "mean_token_accuracy": 0.15452072620391846,
      "num_tokens": 8259817.0,
      "step": 3255
    },
    {
      "entropy": 6.723262643814087,
      "epoch": 0.3762261973456434,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992280636232933,
      "loss": 6.5672,
      "mean_token_accuracy": 0.16001370698213577,
      "num_tokens": 8274359.0,
      "step": 3260
    },
    {
      "entropy": 6.708165502548217,
      "epoch": 0.376803231390652,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992246446480007,
      "loss": 6.5243,
      "mean_token_accuracy": 0.15462508499622346,
      "num_tokens": 8286355.0,
      "step": 3265
    },
    {
      "entropy": 6.684781646728515,
      "epoch": 0.3773802654356607,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992212181310235,
      "loss": 6.5255,
      "mean_token_accuracy": 0.15200337022542953,
      "num_tokens": 8298338.0,
      "step": 3270
    },
    {
      "entropy": 6.566628313064575,
      "epoch": 0.37795729948066936,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499217784072477,
      "loss": 6.487,
      "mean_token_accuracy": 0.1598159983754158,
      "num_tokens": 8312367.0,
      "step": 3275
    },
    {
      "entropy": 6.686645746231079,
      "epoch": 0.378534333525678,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004992143424724768,
      "loss": 6.5769,
      "mean_token_accuracy": 0.14797511398792268,
      "num_tokens": 8325577.0,
      "step": 3280
    },
    {
      "entropy": 6.64054536819458,
      "epoch": 0.3791113675706867,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992108933311385,
      "loss": 6.5993,
      "mean_token_accuracy": 0.15186274349689483,
      "num_tokens": 8338965.0,
      "step": 3285
    },
    {
      "entropy": 6.75788688659668,
      "epoch": 0.37968840161569534,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004992074366485782,
      "loss": 6.5477,
      "mean_token_accuracy": 0.14462305903434752,
      "num_tokens": 8351669.0,
      "step": 3290
    },
    {
      "entropy": 6.611213302612304,
      "epoch": 0.380265435660704,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499203972424912,
      "loss": 6.5972,
      "mean_token_accuracy": 0.1517086461186409,
      "num_tokens": 8364194.0,
      "step": 3295
    },
    {
      "entropy": 6.657028436660767,
      "epoch": 0.3808424697057126,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992005006602567,
      "loss": 6.6099,
      "mean_token_accuracy": 0.15367592871189117,
      "num_tokens": 8377001.0,
      "step": 3300
    },
    {
      "entropy": 6.6885782241821286,
      "epoch": 0.3814195037507213,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991970213547289,
      "loss": 6.5934,
      "mean_token_accuracy": 0.14891462922096252,
      "num_tokens": 8389247.0,
      "step": 3305
    },
    {
      "entropy": 6.680207204818726,
      "epoch": 0.38199653779572995,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991935345084457,
      "loss": 6.4901,
      "mean_token_accuracy": 0.15384454876184464,
      "num_tokens": 8402915.0,
      "step": 3310
    },
    {
      "entropy": 6.681994199752808,
      "epoch": 0.3825735718407386,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991900401215243,
      "loss": 6.6021,
      "mean_token_accuracy": 0.1523881047964096,
      "num_tokens": 8414590.0,
      "step": 3315
    },
    {
      "entropy": 6.729384279251098,
      "epoch": 0.3831506058857473,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004991865381940822,
      "loss": 6.5673,
      "mean_token_accuracy": 0.15600976794958116,
      "num_tokens": 8427337.0,
      "step": 3320
    },
    {
      "entropy": 6.55053391456604,
      "epoch": 0.3837276399307559,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991830287262374,
      "loss": 6.4612,
      "mean_token_accuracy": 0.16062667965888977,
      "num_tokens": 8440095.0,
      "step": 3325
    },
    {
      "entropy": 6.660853481292724,
      "epoch": 0.38430467397576457,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991795117181077,
      "loss": 6.4751,
      "mean_token_accuracy": 0.15257254242897034,
      "num_tokens": 8452448.0,
      "step": 3330
    },
    {
      "entropy": 6.684061670303345,
      "epoch": 0.3848817080207732,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991759871698113,
      "loss": 6.5163,
      "mean_token_accuracy": 0.15853206515312196,
      "num_tokens": 8465289.0,
      "step": 3335
    },
    {
      "entropy": 6.611589431762695,
      "epoch": 0.3854587420657819,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991724550814671,
      "loss": 6.5292,
      "mean_token_accuracy": 0.15679139643907547,
      "num_tokens": 8477365.0,
      "step": 3340
    },
    {
      "entropy": 6.678548574447632,
      "epoch": 0.38603577611079054,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991689154531937,
      "loss": 6.5883,
      "mean_token_accuracy": 0.15190599262714385,
      "num_tokens": 8489104.0,
      "step": 3345
    },
    {
      "entropy": 6.645474195480347,
      "epoch": 0.3866128101557992,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004991653682851102,
      "loss": 6.4714,
      "mean_token_accuracy": 0.15618738830089568,
      "num_tokens": 8501509.0,
      "step": 3350
    },
    {
      "entropy": 6.611126184463501,
      "epoch": 0.3871898442008078,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499161813577336,
      "loss": 6.5253,
      "mean_token_accuracy": 0.15038307160139083,
      "num_tokens": 8514224.0,
      "step": 3355
    },
    {
      "entropy": 6.688539123535156,
      "epoch": 0.3877668782458165,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991582513299903,
      "loss": 6.5916,
      "mean_token_accuracy": 0.14535096436738967,
      "num_tokens": 8526689.0,
      "step": 3360
    },
    {
      "entropy": 6.6264406681060795,
      "epoch": 0.38834391229082516,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991546815431932,
      "loss": 6.6281,
      "mean_token_accuracy": 0.14953829497098922,
      "num_tokens": 8539952.0,
      "step": 3365
    },
    {
      "entropy": 6.771113109588623,
      "epoch": 0.3889209463358338,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499151104217065,
      "loss": 6.6007,
      "mean_token_accuracy": 0.14805661290884017,
      "num_tokens": 8553428.0,
      "step": 3370
    },
    {
      "entropy": 6.625067234039307,
      "epoch": 0.3894979803808425,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991475193517254,
      "loss": 6.4898,
      "mean_token_accuracy": 0.1624234899878502,
      "num_tokens": 8566066.0,
      "step": 3375
    },
    {
      "entropy": 6.575003719329834,
      "epoch": 0.39007501442585113,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991439269472956,
      "loss": 6.4355,
      "mean_token_accuracy": 0.15627922117710114,
      "num_tokens": 8579211.0,
      "step": 3380
    },
    {
      "entropy": 6.679918241500855,
      "epoch": 0.39065204847085977,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991403270038961,
      "loss": 6.5416,
      "mean_token_accuracy": 0.15135526806116104,
      "num_tokens": 8591647.0,
      "step": 3385
    },
    {
      "entropy": 6.600178575515747,
      "epoch": 0.3912290825158684,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499136719521648,
      "loss": 6.5333,
      "mean_token_accuracy": 0.15030259266495705,
      "num_tokens": 8604279.0,
      "step": 3390
    },
    {
      "entropy": 6.637741947174073,
      "epoch": 0.3918061165608771,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004991331045006726,
      "loss": 6.4492,
      "mean_token_accuracy": 0.15820510387420655,
      "num_tokens": 8617722.0,
      "step": 3395
    },
    {
      "entropy": 6.528800201416016,
      "epoch": 0.39238315060588574,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991294819410916,
      "loss": 6.4793,
      "mean_token_accuracy": 0.16577797755599022,
      "num_tokens": 8630250.0,
      "step": 3400
    },
    {
      "entropy": 6.537851476669312,
      "epoch": 0.3929601846508944,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991258518430268,
      "loss": 6.3807,
      "mean_token_accuracy": 0.16118550151586533,
      "num_tokens": 8641900.0,
      "step": 3405
    },
    {
      "entropy": 6.631650495529175,
      "epoch": 0.3935372186959031,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004991222142066003,
      "loss": 6.4887,
      "mean_token_accuracy": 0.15559411495923997,
      "num_tokens": 8654588.0,
      "step": 3410
    },
    {
      "entropy": 6.676535940170288,
      "epoch": 0.3941142527409117,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991185690319345,
      "loss": 6.5208,
      "mean_token_accuracy": 0.14937663450837135,
      "num_tokens": 8668111.0,
      "step": 3415
    },
    {
      "entropy": 6.6537620544433596,
      "epoch": 0.39469128678592036,
      "grad_norm": 0.7578125,
      "learning_rate": 0.000499114916319152,
      "loss": 6.5939,
      "mean_token_accuracy": 0.15261775404214858,
      "num_tokens": 8681461.0,
      "step": 3420
    },
    {
      "entropy": 6.670454740524292,
      "epoch": 0.395268320830929,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991112560683755,
      "loss": 6.5282,
      "mean_token_accuracy": 0.1547023683786392,
      "num_tokens": 8692893.0,
      "step": 3425
    },
    {
      "entropy": 6.594304513931275,
      "epoch": 0.3958453548759377,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004991075882797283,
      "loss": 6.4067,
      "mean_token_accuracy": 0.16767280101776122,
      "num_tokens": 8704704.0,
      "step": 3430
    },
    {
      "entropy": 6.652268505096435,
      "epoch": 0.39642238892094633,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991039129533337,
      "loss": 6.5718,
      "mean_token_accuracy": 0.15412139147520065,
      "num_tokens": 8717595.0,
      "step": 3435
    },
    {
      "entropy": 6.561073684692383,
      "epoch": 0.396999422965955,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991002300893152,
      "loss": 6.523,
      "mean_token_accuracy": 0.15162818729877472,
      "num_tokens": 8729712.0,
      "step": 3440
    },
    {
      "entropy": 6.681739616394043,
      "epoch": 0.39757645701096367,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990965396877969,
      "loss": 6.4438,
      "mean_token_accuracy": 0.1530939370393753,
      "num_tokens": 8742668.0,
      "step": 3445
    },
    {
      "entropy": 6.681428575515747,
      "epoch": 0.3981534910559723,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990928417489027,
      "loss": 6.5241,
      "mean_token_accuracy": 0.15954759120941162,
      "num_tokens": 8754755.0,
      "step": 3450
    },
    {
      "entropy": 6.601272821426392,
      "epoch": 0.39873052510098095,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004990891362727572,
      "loss": 6.4595,
      "mean_token_accuracy": 0.16198282837867736,
      "num_tokens": 8766063.0,
      "step": 3455
    },
    {
      "entropy": 6.669272375106812,
      "epoch": 0.3993075591459896,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990854232594848,
      "loss": 6.5188,
      "mean_token_accuracy": 0.15411025285720825,
      "num_tokens": 8779226.0,
      "step": 3460
    },
    {
      "entropy": 6.599178600311279,
      "epoch": 0.3998845931909983,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004990817027092106,
      "loss": 6.5243,
      "mean_token_accuracy": 0.15882545411586763,
      "num_tokens": 8791404.0,
      "step": 3465
    },
    {
      "entropy": 6.6525890827178955,
      "epoch": 0.4004616272360069,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004990779746220595,
      "loss": 6.5689,
      "mean_token_accuracy": 0.152495901286602,
      "num_tokens": 8805306.0,
      "step": 3470
    },
    {
      "entropy": 6.589881801605225,
      "epoch": 0.40103866128101556,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990742389981572,
      "loss": 6.4515,
      "mean_token_accuracy": 0.16215916275978087,
      "num_tokens": 8817850.0,
      "step": 3475
    },
    {
      "entropy": 6.593970775604248,
      "epoch": 0.40161569532602426,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499070495837629,
      "loss": 6.4808,
      "mean_token_accuracy": 0.16321595013141632,
      "num_tokens": 8830765.0,
      "step": 3480
    },
    {
      "entropy": 6.676024723052978,
      "epoch": 0.4021927293710329,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990667451406012,
      "loss": 6.5189,
      "mean_token_accuracy": 0.15197994261980058,
      "num_tokens": 8843329.0,
      "step": 3485
    },
    {
      "entropy": 6.673951768875122,
      "epoch": 0.40276976341604154,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004990629869071995,
      "loss": 6.5201,
      "mean_token_accuracy": 0.15350718498229982,
      "num_tokens": 8856712.0,
      "step": 3490
    },
    {
      "entropy": 6.505955743789673,
      "epoch": 0.4033467974610502,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004990592211375506,
      "loss": 6.486,
      "mean_token_accuracy": 0.16068525463342667,
      "num_tokens": 8868364.0,
      "step": 3495
    },
    {
      "entropy": 6.594236516952515,
      "epoch": 0.4039238315060589,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990554478317811,
      "loss": 6.4549,
      "mean_token_accuracy": 0.15499066933989525,
      "num_tokens": 8881335.0,
      "step": 3500
    },
    {
      "entropy": 6.599120616912842,
      "epoch": 0.4045008655510675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990516669900179,
      "loss": 6.5337,
      "mean_token_accuracy": 0.15531266406178473,
      "num_tokens": 8894049.0,
      "step": 3505
    },
    {
      "entropy": 6.570689392089844,
      "epoch": 0.40507789959607615,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990478786123882,
      "loss": 6.4266,
      "mean_token_accuracy": 0.15987018644809722,
      "num_tokens": 8906056.0,
      "step": 3510
    },
    {
      "entropy": 6.697812080383301,
      "epoch": 0.40565493364108485,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990440826990193,
      "loss": 6.5057,
      "mean_token_accuracy": 0.15659655630588531,
      "num_tokens": 8919606.0,
      "step": 3515
    },
    {
      "entropy": 6.609763240814209,
      "epoch": 0.4062319676860935,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990402792500392,
      "loss": 6.4787,
      "mean_token_accuracy": 0.15734783709049224,
      "num_tokens": 8934733.0,
      "step": 3520
    },
    {
      "entropy": 6.5650359153747555,
      "epoch": 0.4068090017311021,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990364682655754,
      "loss": 6.4952,
      "mean_token_accuracy": 0.15562065839767455,
      "num_tokens": 8948331.0,
      "step": 3525
    },
    {
      "entropy": 6.646923923492432,
      "epoch": 0.40738603577611077,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004990326497457564,
      "loss": 6.4953,
      "mean_token_accuracy": 0.1590483382344246,
      "num_tokens": 8961171.0,
      "step": 3530
    },
    {
      "entropy": 6.594956350326538,
      "epoch": 0.40796306982111946,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004990288236907104,
      "loss": 6.5474,
      "mean_token_accuracy": 0.15281551331281662,
      "num_tokens": 8973164.0,
      "step": 3535
    },
    {
      "entropy": 6.689388084411621,
      "epoch": 0.4085401038661281,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990249901005661,
      "loss": 6.5875,
      "mean_token_accuracy": 0.1423983097076416,
      "num_tokens": 8985981.0,
      "step": 3540
    },
    {
      "entropy": 6.605531454086304,
      "epoch": 0.40911713791113674,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990211489754527,
      "loss": 6.4511,
      "mean_token_accuracy": 0.16109130084514617,
      "num_tokens": 8999473.0,
      "step": 3545
    },
    {
      "entropy": 6.491503286361694,
      "epoch": 0.40969417195614544,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004990173003154993,
      "loss": 6.4031,
      "mean_token_accuracy": 0.16445289999246598,
      "num_tokens": 9010876.0,
      "step": 3550
    },
    {
      "entropy": 6.700158166885376,
      "epoch": 0.4102712060011541,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004990134441208351,
      "loss": 6.5262,
      "mean_token_accuracy": 0.1531542234122753,
      "num_tokens": 9024704.0,
      "step": 3555
    },
    {
      "entropy": 6.5855193614959715,
      "epoch": 0.4108482400461627,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004990095803915901,
      "loss": 6.506,
      "mean_token_accuracy": 0.15667544901371003,
      "num_tokens": 9037006.0,
      "step": 3560
    },
    {
      "entropy": 6.70794711112976,
      "epoch": 0.41142527409117136,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004990057091278942,
      "loss": 6.5674,
      "mean_token_accuracy": 0.1453065738081932,
      "num_tokens": 9050205.0,
      "step": 3565
    },
    {
      "entropy": 6.664211702346802,
      "epoch": 0.41200230813618005,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004990018303298773,
      "loss": 6.4577,
      "mean_token_accuracy": 0.15301085487008095,
      "num_tokens": 9062900.0,
      "step": 3570
    },
    {
      "entropy": 6.509937000274658,
      "epoch": 0.4125793421811887,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989979439976703,
      "loss": 6.4133,
      "mean_token_accuracy": 0.15631442293524742,
      "num_tokens": 9074767.0,
      "step": 3575
    },
    {
      "entropy": 6.691759014129639,
      "epoch": 0.41315637622619733,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989940501314037,
      "loss": 6.5448,
      "mean_token_accuracy": 0.1505478873848915,
      "num_tokens": 9087314.0,
      "step": 3580
    },
    {
      "entropy": 6.565404176712036,
      "epoch": 0.413733410271206,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004989901487312084,
      "loss": 6.5062,
      "mean_token_accuracy": 0.15018775165081025,
      "num_tokens": 9099543.0,
      "step": 3585
    },
    {
      "entropy": 6.679724550247192,
      "epoch": 0.41431044431621467,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989862397972157,
      "loss": 6.4606,
      "mean_token_accuracy": 0.15562401264905928,
      "num_tokens": 9112262.0,
      "step": 3590
    },
    {
      "entropy": 6.4923858642578125,
      "epoch": 0.4148874783612233,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989823233295572,
      "loss": 6.361,
      "mean_token_accuracy": 0.16104505956172943,
      "num_tokens": 9124444.0,
      "step": 3595
    },
    {
      "entropy": 6.536341047286987,
      "epoch": 0.41546451240623195,
      "grad_norm": 1.375,
      "learning_rate": 0.0004989783993283644,
      "loss": 6.4165,
      "mean_token_accuracy": 0.15915344655513763,
      "num_tokens": 9137848.0,
      "step": 3600
    },
    {
      "entropy": 6.589638757705688,
      "epoch": 0.41604154645124064,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989744677937694,
      "loss": 6.4462,
      "mean_token_accuracy": 0.1550605461001396,
      "num_tokens": 9148985.0,
      "step": 3605
    },
    {
      "entropy": 6.614038467407227,
      "epoch": 0.4166185804962493,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989705287259046,
      "loss": 6.487,
      "mean_token_accuracy": 0.15665621310472488,
      "num_tokens": 9161313.0,
      "step": 3610
    },
    {
      "entropy": 6.613156127929687,
      "epoch": 0.4171956145412579,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004989665821249021,
      "loss": 6.473,
      "mean_token_accuracy": 0.1585191547870636,
      "num_tokens": 9173616.0,
      "step": 3615
    },
    {
      "entropy": 6.676098966598511,
      "epoch": 0.4177726485862666,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000498962627990895,
      "loss": 6.5422,
      "mean_token_accuracy": 0.1485801249742508,
      "num_tokens": 9185951.0,
      "step": 3620
    },
    {
      "entropy": 6.6056294441223145,
      "epoch": 0.41834968263127525,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004989586663240161,
      "loss": 6.5437,
      "mean_token_accuracy": 0.1514025017619133,
      "num_tokens": 9199844.0,
      "step": 3625
    },
    {
      "entropy": 6.510508060455322,
      "epoch": 0.4189267166762839,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004989546971243988,
      "loss": 6.439,
      "mean_token_accuracy": 0.16539452224969864,
      "num_tokens": 9213034.0,
      "step": 3630
    },
    {
      "entropy": 6.616938829421997,
      "epoch": 0.41950375072129253,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004989507203921763,
      "loss": 6.4364,
      "mean_token_accuracy": 0.15639522224664687,
      "num_tokens": 9224885.0,
      "step": 3635
    },
    {
      "entropy": 6.6597864627838135,
      "epoch": 0.42008078476630123,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004989467361274828,
      "loss": 6.5093,
      "mean_token_accuracy": 0.14930343627929688,
      "num_tokens": 9238960.0,
      "step": 3640
    },
    {
      "entropy": 6.61781587600708,
      "epoch": 0.42065781881130987,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004989427443304519,
      "loss": 6.5144,
      "mean_token_accuracy": 0.15002497881650925,
      "num_tokens": 9253075.0,
      "step": 3645
    },
    {
      "entropy": 6.559950542449951,
      "epoch": 0.4212348528563185,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498938745001218,
      "loss": 6.4494,
      "mean_token_accuracy": 0.15852921903133393,
      "num_tokens": 9266865.0,
      "step": 3650
    },
    {
      "entropy": 6.643388795852661,
      "epoch": 0.4218118869013272,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989347381399158,
      "loss": 6.4939,
      "mean_token_accuracy": 0.15345629155635834,
      "num_tokens": 9278046.0,
      "step": 3655
    },
    {
      "entropy": 6.561295509338379,
      "epoch": 0.42238892094633584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989307237466799,
      "loss": 6.3834,
      "mean_token_accuracy": 0.1620499387383461,
      "num_tokens": 9291998.0,
      "step": 3660
    },
    {
      "entropy": 6.547120189666748,
      "epoch": 0.4229659549913445,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989267018216453,
      "loss": 6.4648,
      "mean_token_accuracy": 0.16406473219394685,
      "num_tokens": 9305023.0,
      "step": 3665
    },
    {
      "entropy": 6.522882795333862,
      "epoch": 0.4235429890363531,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989226723649473,
      "loss": 6.4317,
      "mean_token_accuracy": 0.15442937165498732,
      "num_tokens": 9316635.0,
      "step": 3670
    },
    {
      "entropy": 6.663633203506469,
      "epoch": 0.4241200230813618,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004989186353767214,
      "loss": 6.4791,
      "mean_token_accuracy": 0.15076802521944047,
      "num_tokens": 9329671.0,
      "step": 3675
    },
    {
      "entropy": 6.582998037338257,
      "epoch": 0.42469705712637046,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004989145908571035,
      "loss": 6.5064,
      "mean_token_accuracy": 0.14730839282274247,
      "num_tokens": 9342634.0,
      "step": 3680
    },
    {
      "entropy": 6.696527910232544,
      "epoch": 0.4252740911713791,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989105388062295,
      "loss": 6.4849,
      "mean_token_accuracy": 0.158047653734684,
      "num_tokens": 9355096.0,
      "step": 3685
    },
    {
      "entropy": 6.54531717300415,
      "epoch": 0.4258511252163878,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989064792242358,
      "loss": 6.4678,
      "mean_token_accuracy": 0.15762947052717208,
      "num_tokens": 9367722.0,
      "step": 3690
    },
    {
      "entropy": 6.664209365844727,
      "epoch": 0.42642815926139643,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989024121112589,
      "loss": 6.5256,
      "mean_token_accuracy": 0.14759207218885423,
      "num_tokens": 9380247.0,
      "step": 3695
    },
    {
      "entropy": 6.544555616378784,
      "epoch": 0.4270051933064051,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988983374674356,
      "loss": 6.3786,
      "mean_token_accuracy": 0.1646985724568367,
      "num_tokens": 9392216.0,
      "step": 3700
    },
    {
      "entropy": 6.6155448913574215,
      "epoch": 0.4275822273514137,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004988942552929029,
      "loss": 6.5005,
      "mean_token_accuracy": 0.15172917097806932,
      "num_tokens": 9405215.0,
      "step": 3705
    },
    {
      "entropy": 6.557486343383789,
      "epoch": 0.4281592613964224,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988901655877981,
      "loss": 6.4623,
      "mean_token_accuracy": 0.15599302351474761,
      "num_tokens": 9417582.0,
      "step": 3710
    },
    {
      "entropy": 6.570103740692138,
      "epoch": 0.42873629544143105,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988860683522589,
      "loss": 6.4716,
      "mean_token_accuracy": 0.1522749125957489,
      "num_tokens": 9428680.0,
      "step": 3715
    },
    {
      "entropy": 6.646183061599731,
      "epoch": 0.4293133294864397,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000498881963586423,
      "loss": 6.363,
      "mean_token_accuracy": 0.1601622447371483,
      "num_tokens": 9441669.0,
      "step": 3720
    },
    {
      "entropy": 6.506447744369507,
      "epoch": 0.4298903635314484,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988778512904282,
      "loss": 6.521,
      "mean_token_accuracy": 0.15703264325857164,
      "num_tokens": 9455214.0,
      "step": 3725
    },
    {
      "entropy": 6.652627992630005,
      "epoch": 0.430467397576457,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988737314644134,
      "loss": 6.4786,
      "mean_token_accuracy": 0.1539377197623253,
      "num_tokens": 9468180.0,
      "step": 3730
    },
    {
      "entropy": 6.549683570861816,
      "epoch": 0.43104443162146566,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004988696041085168,
      "loss": 6.4634,
      "mean_token_accuracy": 0.16096458733081817,
      "num_tokens": 9481243.0,
      "step": 3735
    },
    {
      "entropy": 6.519383430480957,
      "epoch": 0.4316214656664743,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004988654692228772,
      "loss": 6.3802,
      "mean_token_accuracy": 0.17021922022104263,
      "num_tokens": 9493511.0,
      "step": 3740
    },
    {
      "entropy": 6.572857522964478,
      "epoch": 0.432198499711483,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004988613268076335,
      "loss": 6.4362,
      "mean_token_accuracy": 0.16488718390464782,
      "num_tokens": 9505450.0,
      "step": 3745
    },
    {
      "entropy": 6.567654037475586,
      "epoch": 0.43277553375649164,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988571768629257,
      "loss": 6.4744,
      "mean_token_accuracy": 0.15595853626728057,
      "num_tokens": 9518684.0,
      "step": 3750
    },
    {
      "entropy": 6.562350463867188,
      "epoch": 0.4333525678015003,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004988530193888927,
      "loss": 6.436,
      "mean_token_accuracy": 0.15653682202100755,
      "num_tokens": 9532064.0,
      "step": 3755
    },
    {
      "entropy": 6.541110992431641,
      "epoch": 0.4339296018465089,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988488543856747,
      "loss": 6.3928,
      "mean_token_accuracy": 0.16690051108598708,
      "num_tokens": 9543810.0,
      "step": 3760
    },
    {
      "entropy": 6.663044881820679,
      "epoch": 0.4345066358915176,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004988446818534115,
      "loss": 6.5377,
      "mean_token_accuracy": 0.14751381427049637,
      "num_tokens": 9556261.0,
      "step": 3765
    },
    {
      "entropy": 6.553533840179443,
      "epoch": 0.43508366993652625,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004988405017922438,
      "loss": 6.3675,
      "mean_token_accuracy": 0.16123850494623185,
      "num_tokens": 9569627.0,
      "step": 3770
    },
    {
      "entropy": 6.597948503494263,
      "epoch": 0.4356607039815349,
      "grad_norm": 0.8125,
      "learning_rate": 0.000498836314202312,
      "loss": 6.4731,
      "mean_token_accuracy": 0.15545672327280044,
      "num_tokens": 9582263.0,
      "step": 3775
    },
    {
      "entropy": 6.544510555267334,
      "epoch": 0.4362377380265436,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988321190837568,
      "loss": 6.3567,
      "mean_token_accuracy": 0.16753018498420716,
      "num_tokens": 9594282.0,
      "step": 3780
    },
    {
      "entropy": 6.534181356430054,
      "epoch": 0.4368147720715522,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988279164367196,
      "loss": 6.3893,
      "mean_token_accuracy": 0.16337036341428757,
      "num_tokens": 9607159.0,
      "step": 3785
    },
    {
      "entropy": 6.573097372055054,
      "epoch": 0.43739180611656087,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004988237062613415,
      "loss": 6.4135,
      "mean_token_accuracy": 0.16553914546966553,
      "num_tokens": 9619221.0,
      "step": 3790
    },
    {
      "entropy": 6.5621894836425785,
      "epoch": 0.4379688401615695,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988194885577644,
      "loss": 6.4533,
      "mean_token_accuracy": 0.15650678128004075,
      "num_tokens": 9631956.0,
      "step": 3795
    },
    {
      "entropy": 6.571819639205932,
      "epoch": 0.4385458742065782,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004988152633261299,
      "loss": 6.4729,
      "mean_token_accuracy": 0.1528559982776642,
      "num_tokens": 9643557.0,
      "step": 3800
    },
    {
      "entropy": 6.580244016647339,
      "epoch": 0.43912290825158684,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049881103056658,
      "loss": 6.4304,
      "mean_token_accuracy": 0.15968161821365356,
      "num_tokens": 9655422.0,
      "step": 3805
    },
    {
      "entropy": 6.623444700241089,
      "epoch": 0.4396999422965955,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004988067902792575,
      "loss": 6.4955,
      "mean_token_accuracy": 0.15388246327638627,
      "num_tokens": 9669522.0,
      "step": 3810
    },
    {
      "entropy": 6.619115352630615,
      "epoch": 0.4402769763416042,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988025424643047,
      "loss": 6.5132,
      "mean_token_accuracy": 0.1553269311785698,
      "num_tokens": 9681301.0,
      "step": 3815
    },
    {
      "entropy": 6.542429399490357,
      "epoch": 0.4408540103866128,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004987982871218645,
      "loss": 6.4245,
      "mean_token_accuracy": 0.15602936893701552,
      "num_tokens": 9693894.0,
      "step": 3820
    },
    {
      "entropy": 6.647011184692383,
      "epoch": 0.44143104443162146,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004987940242520802,
      "loss": 6.5104,
      "mean_token_accuracy": 0.15009001046419143,
      "num_tokens": 9706499.0,
      "step": 3825
    },
    {
      "entropy": 6.515692329406738,
      "epoch": 0.4420080784766301,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000498789753855095,
      "loss": 6.3681,
      "mean_token_accuracy": 0.16146169751882553,
      "num_tokens": 9718632.0,
      "step": 3830
    },
    {
      "entropy": 6.594277954101562,
      "epoch": 0.4425851125216388,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987854759310526,
      "loss": 6.4487,
      "mean_token_accuracy": 0.15747116953134538,
      "num_tokens": 9730451.0,
      "step": 3835
    },
    {
      "entropy": 6.623141956329346,
      "epoch": 0.44316214656664743,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987811904800968,
      "loss": 6.483,
      "mean_token_accuracy": 0.1548147827386856,
      "num_tokens": 9742693.0,
      "step": 3840
    },
    {
      "entropy": 6.539883422851562,
      "epoch": 0.44373918061165607,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987768975023719,
      "loss": 6.457,
      "mean_token_accuracy": 0.1584097445011139,
      "num_tokens": 9755642.0,
      "step": 3845
    },
    {
      "entropy": 6.5988892078399655,
      "epoch": 0.44431621465666477,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004987725969980222,
      "loss": 6.4743,
      "mean_token_accuracy": 0.15320970863103867,
      "num_tokens": 9767674.0,
      "step": 3850
    },
    {
      "entropy": 6.566948366165161,
      "epoch": 0.4448932487016734,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004987682889671923,
      "loss": 6.3987,
      "mean_token_accuracy": 0.1640625014901161,
      "num_tokens": 9781018.0,
      "step": 3855
    },
    {
      "entropy": 6.564487266540527,
      "epoch": 0.44547028274668204,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004987639734100272,
      "loss": 6.4639,
      "mean_token_accuracy": 0.1597461685538292,
      "num_tokens": 9792521.0,
      "step": 3860
    },
    {
      "entropy": 6.566139602661133,
      "epoch": 0.4460473167916907,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004987596503266721,
      "loss": 6.4513,
      "mean_token_accuracy": 0.16076726466417313,
      "num_tokens": 9805356.0,
      "step": 3865
    },
    {
      "entropy": 6.484199953079224,
      "epoch": 0.4466243508366994,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987553197172722,
      "loss": 6.4047,
      "mean_token_accuracy": 0.16520888954401017,
      "num_tokens": 9818779.0,
      "step": 3870
    },
    {
      "entropy": 6.664399433135986,
      "epoch": 0.447201384881708,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004987509815819732,
      "loss": 6.4761,
      "mean_token_accuracy": 0.15092621743679047,
      "num_tokens": 9832328.0,
      "step": 3875
    },
    {
      "entropy": 6.58227424621582,
      "epoch": 0.44777841892671666,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004987466359209213,
      "loss": 6.4288,
      "mean_token_accuracy": 0.15341632515192033,
      "num_tokens": 9846281.0,
      "step": 3880
    },
    {
      "entropy": 6.529612970352173,
      "epoch": 0.44835545297172535,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987422827342622,
      "loss": 6.4904,
      "mean_token_accuracy": 0.15297739431262017,
      "num_tokens": 9858737.0,
      "step": 3885
    },
    {
      "entropy": 6.580500078201294,
      "epoch": 0.448932487016734,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987379220221426,
      "loss": 6.5037,
      "mean_token_accuracy": 0.15690465569496154,
      "num_tokens": 9872852.0,
      "step": 3890
    },
    {
      "entropy": 6.58209867477417,
      "epoch": 0.44950952106174263,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004987335537847092,
      "loss": 6.4174,
      "mean_token_accuracy": 0.16698380708694457,
      "num_tokens": 9886339.0,
      "step": 3895
    },
    {
      "entropy": 6.635822200775147,
      "epoch": 0.4500865551067513,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987291780221088,
      "loss": 6.4112,
      "mean_token_accuracy": 0.1551224783062935,
      "num_tokens": 9897525.0,
      "step": 3900
    },
    {
      "entropy": 6.516907024383545,
      "epoch": 0.45066358915175997,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004987247947344887,
      "loss": 6.4289,
      "mean_token_accuracy": 0.16607835292816162,
      "num_tokens": 9909733.0,
      "step": 3905
    },
    {
      "entropy": 6.550909996032715,
      "epoch": 0.4512406231967686,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987204039219962,
      "loss": 6.4258,
      "mean_token_accuracy": 0.1511649563908577,
      "num_tokens": 9922503.0,
      "step": 3910
    },
    {
      "entropy": 6.588738107681275,
      "epoch": 0.45181765724177725,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987160055847791,
      "loss": 6.4409,
      "mean_token_accuracy": 0.15798533260822295,
      "num_tokens": 9935517.0,
      "step": 3915
    },
    {
      "entropy": 6.61040678024292,
      "epoch": 0.45239469128678594,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004987115997229852,
      "loss": 6.4937,
      "mean_token_accuracy": 0.15557873547077178,
      "num_tokens": 9947663.0,
      "step": 3920
    },
    {
      "entropy": 6.48343858718872,
      "epoch": 0.4529717253317946,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004987071863367629,
      "loss": 6.2877,
      "mean_token_accuracy": 0.1631803885102272,
      "num_tokens": 9960924.0,
      "step": 3925
    },
    {
      "entropy": 6.531168794631958,
      "epoch": 0.4535487593768032,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987027654262604,
      "loss": 6.4252,
      "mean_token_accuracy": 0.15896384865045549,
      "num_tokens": 9973506.0,
      "step": 3930
    },
    {
      "entropy": 6.592694234848023,
      "epoch": 0.45412579342181186,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004986983369916264,
      "loss": 6.408,
      "mean_token_accuracy": 0.16736576110124587,
      "num_tokens": 9988334.0,
      "step": 3935
    },
    {
      "entropy": 6.620164299011231,
      "epoch": 0.45470282746682056,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986939010330102,
      "loss": 6.495,
      "mean_token_accuracy": 0.15416028946638108,
      "num_tokens": 10001295.0,
      "step": 3940
    },
    {
      "entropy": 6.561420059204101,
      "epoch": 0.4552798615118292,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986894575505606,
      "loss": 6.3938,
      "mean_token_accuracy": 0.15716030150651933,
      "num_tokens": 10012396.0,
      "step": 3945
    },
    {
      "entropy": 6.5451727390289305,
      "epoch": 0.45585689555683784,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986850065444272,
      "loss": 6.4529,
      "mean_token_accuracy": 0.15944151133298873,
      "num_tokens": 10025918.0,
      "step": 3950
    },
    {
      "entropy": 6.643007612228393,
      "epoch": 0.45643392960184653,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004986805480147598,
      "loss": 6.4658,
      "mean_token_accuracy": 0.15336925834417342,
      "num_tokens": 10038674.0,
      "step": 3955
    },
    {
      "entropy": 6.53163480758667,
      "epoch": 0.4570109636468552,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004986760819617082,
      "loss": 6.4256,
      "mean_token_accuracy": 0.16304580122232437,
      "num_tokens": 10051623.0,
      "step": 3960
    },
    {
      "entropy": 6.5593581199646,
      "epoch": 0.4575879976918638,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004986716083854228,
      "loss": 6.4961,
      "mean_token_accuracy": 0.1553208589553833,
      "num_tokens": 10063731.0,
      "step": 3965
    },
    {
      "entropy": 6.648737621307373,
      "epoch": 0.45816503173687245,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986671272860538,
      "loss": 6.4949,
      "mean_token_accuracy": 0.15985720306634904,
      "num_tokens": 10075897.0,
      "step": 3970
    },
    {
      "entropy": 6.561565828323364,
      "epoch": 0.45874206578188115,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986626386637521,
      "loss": 6.4551,
      "mean_token_accuracy": 0.152682101726532,
      "num_tokens": 10088605.0,
      "step": 3975
    },
    {
      "entropy": 6.624797248840332,
      "epoch": 0.4593190998268898,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004986581425186688,
      "loss": 6.4371,
      "mean_token_accuracy": 0.1650243565440178,
      "num_tokens": 10101758.0,
      "step": 3980
    },
    {
      "entropy": 6.5097900390625,
      "epoch": 0.4598961338718984,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986536388509548,
      "loss": 6.4741,
      "mean_token_accuracy": 0.15137582868337632,
      "num_tokens": 10113838.0,
      "step": 3985
    },
    {
      "entropy": 6.527582025527954,
      "epoch": 0.4604731679169071,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004986491276607618,
      "loss": 6.412,
      "mean_token_accuracy": 0.15793143063783646,
      "num_tokens": 10127676.0,
      "step": 3990
    },
    {
      "entropy": 6.568745374679565,
      "epoch": 0.46105020196191576,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004986446089482416,
      "loss": 6.3727,
      "mean_token_accuracy": 0.15627624988555908,
      "num_tokens": 10139734.0,
      "step": 3995
    },
    {
      "entropy": 6.538666677474976,
      "epoch": 0.4616272360069244,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986400827135459,
      "loss": 6.4699,
      "mean_token_accuracy": 0.15631870180368423,
      "num_tokens": 10151842.0,
      "step": 4000
    },
    {
      "entropy": 6.649822473526001,
      "epoch": 0.46220427005193304,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986355489568272,
      "loss": 6.5041,
      "mean_token_accuracy": 0.151052263379097,
      "num_tokens": 10163539.0,
      "step": 4005
    },
    {
      "entropy": 6.57418909072876,
      "epoch": 0.46278130409694174,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986310076782379,
      "loss": 6.4568,
      "mean_token_accuracy": 0.1587604507803917,
      "num_tokens": 10174760.0,
      "step": 4010
    },
    {
      "entropy": 6.541180610656738,
      "epoch": 0.4633583381419504,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986264588779307,
      "loss": 6.4936,
      "mean_token_accuracy": 0.15445222556591034,
      "num_tokens": 10186627.0,
      "step": 4015
    },
    {
      "entropy": 6.666124773025513,
      "epoch": 0.463935372186959,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004986219025560586,
      "loss": 6.3784,
      "mean_token_accuracy": 0.168064945936203,
      "num_tokens": 10200084.0,
      "step": 4020
    },
    {
      "entropy": 6.546374320983887,
      "epoch": 0.4645124062319677,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498617338712775,
      "loss": 6.4466,
      "mean_token_accuracy": 0.1597646102309227,
      "num_tokens": 10212121.0,
      "step": 4025
    },
    {
      "entropy": 6.572313737869263,
      "epoch": 0.46508944027697635,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986127673482332,
      "loss": 6.4305,
      "mean_token_accuracy": 0.1606583520770073,
      "num_tokens": 10225138.0,
      "step": 4030
    },
    {
      "entropy": 6.667597007751465,
      "epoch": 0.465666474321985,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004986081884625871,
      "loss": 6.4703,
      "mean_token_accuracy": 0.15737486779689788,
      "num_tokens": 10238309.0,
      "step": 4035
    },
    {
      "entropy": 6.580912351608276,
      "epoch": 0.46624350836699363,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986036020559906,
      "loss": 6.4289,
      "mean_token_accuracy": 0.15446672439575196,
      "num_tokens": 10251405.0,
      "step": 4040
    },
    {
      "entropy": 6.587841939926148,
      "epoch": 0.4668205424120023,
      "grad_norm": 0.8125,
      "learning_rate": 0.000498599008128598,
      "loss": 6.382,
      "mean_token_accuracy": 0.16414790004491805,
      "num_tokens": 10262870.0,
      "step": 4045
    },
    {
      "entropy": 6.475424766540527,
      "epoch": 0.46739757645701097,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985944066805639,
      "loss": 6.4407,
      "mean_token_accuracy": 0.1515449434518814,
      "num_tokens": 10274283.0,
      "step": 4050
    },
    {
      "entropy": 6.612028789520264,
      "epoch": 0.4679746105020196,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498589797712043,
      "loss": 6.4517,
      "mean_token_accuracy": 0.15694061815738677,
      "num_tokens": 10287581.0,
      "step": 4055
    },
    {
      "entropy": 6.604260540008545,
      "epoch": 0.4685516445470283,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004985851812231903,
      "loss": 6.3818,
      "mean_token_accuracy": 0.1622477129101753,
      "num_tokens": 10299654.0,
      "step": 4060
    },
    {
      "entropy": 6.522708511352539,
      "epoch": 0.46912867859203694,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004985805572141611,
      "loss": 6.4622,
      "mean_token_accuracy": 0.15169159919023514,
      "num_tokens": 10312323.0,
      "step": 4065
    },
    {
      "entropy": 6.6179437160491945,
      "epoch": 0.4697057126370456,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985759256851108,
      "loss": 6.3254,
      "mean_token_accuracy": 0.16399953067302703,
      "num_tokens": 10325078.0,
      "step": 4070
    },
    {
      "entropy": 6.48213152885437,
      "epoch": 0.4702827466820542,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985712866361954,
      "loss": 6.4523,
      "mean_token_accuracy": 0.1553603947162628,
      "num_tokens": 10339938.0,
      "step": 4075
    },
    {
      "entropy": 6.573529958724976,
      "epoch": 0.4708597807270629,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985666400675709,
      "loss": 6.4761,
      "mean_token_accuracy": 0.1563069850206375,
      "num_tokens": 10352024.0,
      "step": 4080
    },
    {
      "entropy": 6.54999623298645,
      "epoch": 0.47143681477207156,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985619859793934,
      "loss": 6.2783,
      "mean_token_accuracy": 0.17008548229932785,
      "num_tokens": 10365919.0,
      "step": 4085
    },
    {
      "entropy": 6.421245479583741,
      "epoch": 0.4720138488170802,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985573243718195,
      "loss": 6.383,
      "mean_token_accuracy": 0.16240842491388321,
      "num_tokens": 10377488.0,
      "step": 4090
    },
    {
      "entropy": 6.564831829071045,
      "epoch": 0.4725908828620889,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004985526552450061,
      "loss": 6.3718,
      "mean_token_accuracy": 0.16302193850278854,
      "num_tokens": 10391187.0,
      "step": 4095
    },
    {
      "entropy": 6.48703351020813,
      "epoch": 0.47316791690709753,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985479785991103,
      "loss": 6.3607,
      "mean_token_accuracy": 0.16460198909044266,
      "num_tokens": 10403458.0,
      "step": 4100
    },
    {
      "entropy": 6.61144905090332,
      "epoch": 0.47374495095210617,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985432944342892,
      "loss": 6.5351,
      "mean_token_accuracy": 0.14650237709283828,
      "num_tokens": 10416533.0,
      "step": 4105
    },
    {
      "entropy": 6.613126182556153,
      "epoch": 0.4743219849971148,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985386027507003,
      "loss": 6.4002,
      "mean_token_accuracy": 0.15792774260044098,
      "num_tokens": 10429593.0,
      "step": 4110
    },
    {
      "entropy": 6.524380207061768,
      "epoch": 0.4748990190421235,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985339035485018,
      "loss": 6.4195,
      "mean_token_accuracy": 0.1630406856536865,
      "num_tokens": 10440520.0,
      "step": 4115
    },
    {
      "entropy": 6.5063190937042235,
      "epoch": 0.47547605308713214,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985291968278513,
      "loss": 6.4042,
      "mean_token_accuracy": 0.1612715020775795,
      "num_tokens": 10453971.0,
      "step": 4120
    },
    {
      "entropy": 6.541058540344238,
      "epoch": 0.4760530871321408,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004985244825889074,
      "loss": 6.4259,
      "mean_token_accuracy": 0.1584889307618141,
      "num_tokens": 10466033.0,
      "step": 4125
    },
    {
      "entropy": 6.533453750610351,
      "epoch": 0.4766301211771495,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004985197608318284,
      "loss": 6.3901,
      "mean_token_accuracy": 0.16374807804822922,
      "num_tokens": 10478608.0,
      "step": 4130
    },
    {
      "entropy": 6.620793581008911,
      "epoch": 0.4772071552221581,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985150315567735,
      "loss": 6.486,
      "mean_token_accuracy": 0.15880320221185684,
      "num_tokens": 10491223.0,
      "step": 4135
    },
    {
      "entropy": 6.608275985717773,
      "epoch": 0.47778418926716676,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004985102947639013,
      "loss": 6.5194,
      "mean_token_accuracy": 0.14948661625385284,
      "num_tokens": 10504398.0,
      "step": 4140
    },
    {
      "entropy": 6.520168113708496,
      "epoch": 0.4783612233121754,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004985055504533715,
      "loss": 6.3576,
      "mean_token_accuracy": 0.16035733073949815,
      "num_tokens": 10516998.0,
      "step": 4145
    },
    {
      "entropy": 6.559482908248901,
      "epoch": 0.4789382573571841,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004985007986253435,
      "loss": 6.3572,
      "mean_token_accuracy": 0.16101326942443847,
      "num_tokens": 10529632.0,
      "step": 4150
    },
    {
      "entropy": 6.449239063262939,
      "epoch": 0.47951529140219273,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984960392799772,
      "loss": 6.3698,
      "mean_token_accuracy": 0.161073037981987,
      "num_tokens": 10542479.0,
      "step": 4155
    },
    {
      "entropy": 6.47977499961853,
      "epoch": 0.4800923254472014,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984912724174326,
      "loss": 6.2916,
      "mean_token_accuracy": 0.1709906816482544,
      "num_tokens": 10555322.0,
      "step": 4160
    },
    {
      "entropy": 6.575711727142334,
      "epoch": 0.48066935949221,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004984864980378701,
      "loss": 6.437,
      "mean_token_accuracy": 0.15588219612836837,
      "num_tokens": 10567757.0,
      "step": 4165
    },
    {
      "entropy": 6.596047639846802,
      "epoch": 0.4812463935372187,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004984817161414502,
      "loss": 6.4697,
      "mean_token_accuracy": 0.1479870706796646,
      "num_tokens": 10580820.0,
      "step": 4170
    },
    {
      "entropy": 6.623038339614868,
      "epoch": 0.48182342758222735,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004984769267283338,
      "loss": 6.4656,
      "mean_token_accuracy": 0.1566685937345028,
      "num_tokens": 10595690.0,
      "step": 4175
    },
    {
      "entropy": 6.564116668701172,
      "epoch": 0.482400461627236,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984721297986819,
      "loss": 6.38,
      "mean_token_accuracy": 0.15687415599822999,
      "num_tokens": 10608393.0,
      "step": 4180
    },
    {
      "entropy": 6.582472229003907,
      "epoch": 0.4829774956722447,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984673253526561,
      "loss": 6.4586,
      "mean_token_accuracy": 0.15567171573638916,
      "num_tokens": 10621073.0,
      "step": 4185
    },
    {
      "entropy": 6.486496543884277,
      "epoch": 0.4835545297172533,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004984625133904176,
      "loss": 6.3335,
      "mean_token_accuracy": 0.16945450156927108,
      "num_tokens": 10633945.0,
      "step": 4190
    },
    {
      "entropy": 6.520366954803467,
      "epoch": 0.48413156376226196,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984576939121286,
      "loss": 6.4579,
      "mean_token_accuracy": 0.15404634773731232,
      "num_tokens": 10645727.0,
      "step": 4195
    },
    {
      "entropy": 6.560869789123535,
      "epoch": 0.4847085978072706,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984528669179511,
      "loss": 6.3169,
      "mean_token_accuracy": 0.15794729739427565,
      "num_tokens": 10658488.0,
      "step": 4200
    },
    {
      "entropy": 6.539811325073242,
      "epoch": 0.4852856318522793,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984480324080472,
      "loss": 6.3988,
      "mean_token_accuracy": 0.15879094898700713,
      "num_tokens": 10671361.0,
      "step": 4205
    },
    {
      "entropy": 6.540292549133301,
      "epoch": 0.48586266589728794,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049844319038258,
      "loss": 6.342,
      "mean_token_accuracy": 0.17195187956094743,
      "num_tokens": 10684825.0,
      "step": 4210
    },
    {
      "entropy": 6.4991264820098875,
      "epoch": 0.4864396999422966,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984383408417119,
      "loss": 6.4458,
      "mean_token_accuracy": 0.16079134345054627,
      "num_tokens": 10698627.0,
      "step": 4215
    },
    {
      "entropy": 6.572528266906739,
      "epoch": 0.4870167339873053,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984334837856064,
      "loss": 6.3773,
      "mean_token_accuracy": 0.1604588583111763,
      "num_tokens": 10712124.0,
      "step": 4220
    },
    {
      "entropy": 6.537895679473877,
      "epoch": 0.4875937680323139,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004984286192144263,
      "loss": 6.4168,
      "mean_token_accuracy": 0.15386243313550949,
      "num_tokens": 10723939.0,
      "step": 4225
    },
    {
      "entropy": 6.550821685791016,
      "epoch": 0.48817080207732255,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984237471283359,
      "loss": 6.4144,
      "mean_token_accuracy": 0.15769041180610657,
      "num_tokens": 10736795.0,
      "step": 4230
    },
    {
      "entropy": 6.475082588195801,
      "epoch": 0.4887478361223312,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984188675274986,
      "loss": 6.3047,
      "mean_token_accuracy": 0.17049374878406526,
      "num_tokens": 10749905.0,
      "step": 4235
    },
    {
      "entropy": 6.567593622207641,
      "epoch": 0.4893248701673399,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004984139804120787,
      "loss": 6.4184,
      "mean_token_accuracy": 0.1523547112941742,
      "num_tokens": 10761852.0,
      "step": 4240
    },
    {
      "entropy": 6.531088781356812,
      "epoch": 0.4899019042123485,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004984090857822406,
      "loss": 6.3953,
      "mean_token_accuracy": 0.16216247975826265,
      "num_tokens": 10774673.0,
      "step": 4245
    },
    {
      "entropy": 6.480283403396607,
      "epoch": 0.49047893825735717,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984041836381488,
      "loss": 6.3537,
      "mean_token_accuracy": 0.16093540489673613,
      "num_tokens": 10786892.0,
      "step": 4250
    },
    {
      "entropy": 6.518536567687988,
      "epoch": 0.49105597230236586,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983992739799682,
      "loss": 6.4035,
      "mean_token_accuracy": 0.16496101021766663,
      "num_tokens": 10798846.0,
      "step": 4255
    },
    {
      "entropy": 6.526611375808716,
      "epoch": 0.4916330063473745,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004983943568078642,
      "loss": 6.3143,
      "mean_token_accuracy": 0.16506008803844452,
      "num_tokens": 10811893.0,
      "step": 4260
    },
    {
      "entropy": 6.516178178787231,
      "epoch": 0.49221004039238314,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004983894321220017,
      "loss": 6.362,
      "mean_token_accuracy": 0.16890482902526854,
      "num_tokens": 10825595.0,
      "step": 4265
    },
    {
      "entropy": 6.485895586013794,
      "epoch": 0.4927870744373918,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983844999225468,
      "loss": 6.3435,
      "mean_token_accuracy": 0.16962877959012984,
      "num_tokens": 10837546.0,
      "step": 4270
    },
    {
      "entropy": 6.525856685638428,
      "epoch": 0.4933641084824005,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004983795602096651,
      "loss": 6.4474,
      "mean_token_accuracy": 0.15370720028877258,
      "num_tokens": 10851234.0,
      "step": 4275
    },
    {
      "entropy": 6.50496129989624,
      "epoch": 0.4939411425274091,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498374612983523,
      "loss": 6.3901,
      "mean_token_accuracy": 0.15387441217899323,
      "num_tokens": 10864397.0,
      "step": 4280
    },
    {
      "entropy": 6.561623764038086,
      "epoch": 0.49451817657241776,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983696582442866,
      "loss": 6.4006,
      "mean_token_accuracy": 0.15735192149877547,
      "num_tokens": 10877230.0,
      "step": 4285
    },
    {
      "entropy": 6.519133424758911,
      "epoch": 0.49509521061742645,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004983646959921227,
      "loss": 6.3556,
      "mean_token_accuracy": 0.1598821237683296,
      "num_tokens": 10890847.0,
      "step": 4290
    },
    {
      "entropy": 6.475875282287598,
      "epoch": 0.4956722446624351,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004983597262271984,
      "loss": 6.3347,
      "mean_token_accuracy": 0.16928167939186095,
      "num_tokens": 10903488.0,
      "step": 4295
    },
    {
      "entropy": 6.6245604991912845,
      "epoch": 0.49624927870744373,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983547489496804,
      "loss": 6.4087,
      "mean_token_accuracy": 0.15573253035545348,
      "num_tokens": 10914744.0,
      "step": 4300
    },
    {
      "entropy": 6.547078418731689,
      "epoch": 0.49682631275245237,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004983497641597365,
      "loss": 6.4114,
      "mean_token_accuracy": 0.1501732811331749,
      "num_tokens": 10928307.0,
      "step": 4305
    },
    {
      "entropy": 6.537446308135986,
      "epoch": 0.49740334679746107,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004983447718575342,
      "loss": 6.4439,
      "mean_token_accuracy": 0.15933653563261033,
      "num_tokens": 10939876.0,
      "step": 4310
    },
    {
      "entropy": 6.505248355865478,
      "epoch": 0.4979803808424697,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983397720432414,
      "loss": 6.3254,
      "mean_token_accuracy": 0.1652297183871269,
      "num_tokens": 10952123.0,
      "step": 4315
    },
    {
      "entropy": 6.540228748321534,
      "epoch": 0.49855741488747835,
      "grad_norm": 0.75,
      "learning_rate": 0.0004983347647170264,
      "loss": 6.4036,
      "mean_token_accuracy": 0.15447943806648254,
      "num_tokens": 10967374.0,
      "step": 4320
    },
    {
      "entropy": 6.534400272369385,
      "epoch": 0.49913444893248704,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004983297498790574,
      "loss": 6.4577,
      "mean_token_accuracy": 0.15325114727020264,
      "num_tokens": 10980761.0,
      "step": 4325
    },
    {
      "entropy": 6.575414419174194,
      "epoch": 0.4997114829774957,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983247275295034,
      "loss": 6.3945,
      "mean_token_accuracy": 0.15343135446310044,
      "num_tokens": 10994538.0,
      "step": 4330
    },
    {
      "entropy": 6.5833477020263675,
      "epoch": 0.5002885170225043,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004983196976685329,
      "loss": 6.4382,
      "mean_token_accuracy": 0.15443964004516603,
      "num_tokens": 11006956.0,
      "step": 4335
    },
    {
      "entropy": 6.53419942855835,
      "epoch": 0.500865551067513,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983146602963155,
      "loss": 6.4032,
      "mean_token_accuracy": 0.16294672191143036,
      "num_tokens": 11019202.0,
      "step": 4340
    },
    {
      "entropy": 6.613501739501953,
      "epoch": 0.5014425851125216,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004983096154130203,
      "loss": 6.3731,
      "mean_token_accuracy": 0.16164093166589738,
      "num_tokens": 11031560.0,
      "step": 4345
    },
    {
      "entropy": 6.558028364181519,
      "epoch": 0.5020196191575302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004983045630188171,
      "loss": 6.4849,
      "mean_token_accuracy": 0.15343987345695495,
      "num_tokens": 11044537.0,
      "step": 4350
    },
    {
      "entropy": 6.516075372695923,
      "epoch": 0.502596653202539,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982995031138759,
      "loss": 6.3252,
      "mean_token_accuracy": 0.16063188016414642,
      "num_tokens": 11056642.0,
      "step": 4355
    },
    {
      "entropy": 6.514281415939331,
      "epoch": 0.5031736872475476,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004982944356983666,
      "loss": 6.3048,
      "mean_token_accuracy": 0.1605512946844101,
      "num_tokens": 11068642.0,
      "step": 4360
    },
    {
      "entropy": 6.580352306365967,
      "epoch": 0.5037507212925563,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004982893607724602,
      "loss": 6.5844,
      "mean_token_accuracy": 0.15639905929565429,
      "num_tokens": 11082056.0,
      "step": 4365
    },
    {
      "entropy": 6.594468784332276,
      "epoch": 0.5043277553375649,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498284278336327,
      "loss": 6.4158,
      "mean_token_accuracy": 0.15388816744089126,
      "num_tokens": 11093722.0,
      "step": 4370
    },
    {
      "entropy": 6.547727251052857,
      "epoch": 0.5049047893825735,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982791883901379,
      "loss": 6.4151,
      "mean_token_accuracy": 0.15209265276789666,
      "num_tokens": 11105949.0,
      "step": 4375
    },
    {
      "entropy": 6.509112691879272,
      "epoch": 0.5054818234275822,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982740909340643,
      "loss": 6.299,
      "mean_token_accuracy": 0.16578411161899567,
      "num_tokens": 11118074.0,
      "step": 4380
    },
    {
      "entropy": 6.622136878967285,
      "epoch": 0.5060588574725908,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982689859682775,
      "loss": 6.3795,
      "mean_token_accuracy": 0.15766538828611373,
      "num_tokens": 11130917.0,
      "step": 4385
    },
    {
      "entropy": 6.49002857208252,
      "epoch": 0.5066358915175996,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004982638734929492,
      "loss": 6.4184,
      "mean_token_accuracy": 0.1600237637758255,
      "num_tokens": 11142829.0,
      "step": 4390
    },
    {
      "entropy": 6.528630256652832,
      "epoch": 0.5072129255626082,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982587535082515,
      "loss": 6.3764,
      "mean_token_accuracy": 0.16152664870023728,
      "num_tokens": 11154606.0,
      "step": 4395
    },
    {
      "entropy": 6.5705047130584715,
      "epoch": 0.5077899596076169,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982536260143565,
      "loss": 6.3694,
      "mean_token_accuracy": 0.1585418999195099,
      "num_tokens": 11167160.0,
      "step": 4400
    },
    {
      "entropy": 6.4587348937988285,
      "epoch": 0.5083669936526255,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982484910114367,
      "loss": 6.3573,
      "mean_token_accuracy": 0.16451280564069748,
      "num_tokens": 11178813.0,
      "step": 4405
    },
    {
      "entropy": 6.434975481033325,
      "epoch": 0.5089440276976341,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004982433484996648,
      "loss": 6.3155,
      "mean_token_accuracy": 0.16058938950300217,
      "num_tokens": 11192648.0,
      "step": 4410
    },
    {
      "entropy": 6.652609014511109,
      "epoch": 0.5095210617426428,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004982381984792138,
      "loss": 6.4321,
      "mean_token_accuracy": 0.1531258538365364,
      "num_tokens": 11205519.0,
      "step": 4415
    },
    {
      "entropy": 6.52750597000122,
      "epoch": 0.5100980957876514,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004982330409502568,
      "loss": 6.4065,
      "mean_token_accuracy": 0.15604918152093888,
      "num_tokens": 11218362.0,
      "step": 4420
    },
    {
      "entropy": 6.459440803527832,
      "epoch": 0.5106751298326602,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982278759129674,
      "loss": 6.3956,
      "mean_token_accuracy": 0.1558659553527832,
      "num_tokens": 11231821.0,
      "step": 4425
    },
    {
      "entropy": 6.593206834793091,
      "epoch": 0.5112521638776688,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004982227033675194,
      "loss": 6.4448,
      "mean_token_accuracy": 0.1499837964773178,
      "num_tokens": 11246300.0,
      "step": 4430
    },
    {
      "entropy": 6.525052499771118,
      "epoch": 0.5118291979226774,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004982175233140865,
      "loss": 6.3429,
      "mean_token_accuracy": 0.16502684354782104,
      "num_tokens": 11258561.0,
      "step": 4435
    },
    {
      "entropy": 6.5642249584198,
      "epoch": 0.5124062319676861,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982123357528431,
      "loss": 6.3303,
      "mean_token_accuracy": 0.16405102908611296,
      "num_tokens": 11271719.0,
      "step": 4440
    },
    {
      "entropy": 6.3442137241363525,
      "epoch": 0.5129832660126947,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982071406839636,
      "loss": 6.3408,
      "mean_token_accuracy": 0.16218033507466317,
      "num_tokens": 11284166.0,
      "step": 4445
    },
    {
      "entropy": 6.523007535934449,
      "epoch": 0.5135603000577034,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004982019381076229,
      "loss": 6.3807,
      "mean_token_accuracy": 0.1608651116490364,
      "num_tokens": 11296140.0,
      "step": 4450
    },
    {
      "entropy": 6.577465057373047,
      "epoch": 0.514137334102712,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004981967280239958,
      "loss": 6.4006,
      "mean_token_accuracy": 0.156875441968441,
      "num_tokens": 11308680.0,
      "step": 4455
    },
    {
      "entropy": 6.554678201675415,
      "epoch": 0.5147143681477208,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981915104332578,
      "loss": 6.3881,
      "mean_token_accuracy": 0.15510681122541428,
      "num_tokens": 11320887.0,
      "step": 4460
    },
    {
      "entropy": 6.561086893081665,
      "epoch": 0.5152914021927294,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498186285335584,
      "loss": 6.3939,
      "mean_token_accuracy": 0.15924655199050902,
      "num_tokens": 11333873.0,
      "step": 4465
    },
    {
      "entropy": 6.5453094959259035,
      "epoch": 0.515868436237738,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981810527311505,
      "loss": 6.4765,
      "mean_token_accuracy": 0.15514248758554458,
      "num_tokens": 11345526.0,
      "step": 4470
    },
    {
      "entropy": 6.589832925796509,
      "epoch": 0.5164454702827467,
      "grad_norm": 0.765625,
      "learning_rate": 0.000498175812620133,
      "loss": 6.3593,
      "mean_token_accuracy": 0.15877474546432496,
      "num_tokens": 11356840.0,
      "step": 4475
    },
    {
      "entropy": 6.511902713775635,
      "epoch": 0.5170225043277553,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981705650027081,
      "loss": 6.4396,
      "mean_token_accuracy": 0.15772426426410674,
      "num_tokens": 11369146.0,
      "step": 4480
    },
    {
      "entropy": 6.498725557327271,
      "epoch": 0.517599538372764,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498165309879052,
      "loss": 6.3728,
      "mean_token_accuracy": 0.1622074991464615,
      "num_tokens": 11382399.0,
      "step": 4485
    },
    {
      "entropy": 6.481703281402588,
      "epoch": 0.5181765724177726,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004981600472493415,
      "loss": 6.2713,
      "mean_token_accuracy": 0.16591350585222245,
      "num_tokens": 11395602.0,
      "step": 4490
    },
    {
      "entropy": 6.490931653976441,
      "epoch": 0.5187536064627813,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004981547771137537,
      "loss": 6.3686,
      "mean_token_accuracy": 0.15938565284013748,
      "num_tokens": 11408530.0,
      "step": 4495
    },
    {
      "entropy": 6.554465389251709,
      "epoch": 0.51933064050779,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981494994724658,
      "loss": 6.3772,
      "mean_token_accuracy": 0.16451527178287506,
      "num_tokens": 11420150.0,
      "step": 4500
    },
    {
      "entropy": 6.476536464691162,
      "epoch": 0.5199076745527986,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004981442143256554,
      "loss": 6.421,
      "mean_token_accuracy": 0.1505007266998291,
      "num_tokens": 11432318.0,
      "step": 4505
    },
    {
      "entropy": 6.543183517456055,
      "epoch": 0.5204847085978073,
      "grad_norm": 2.25,
      "learning_rate": 0.0004981389216735001,
      "loss": 6.4517,
      "mean_token_accuracy": 0.1579896852374077,
      "num_tokens": 11444143.0,
      "step": 4510
    },
    {
      "entropy": 6.475643825531006,
      "epoch": 0.5210617426428159,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498133621516178,
      "loss": 6.3557,
      "mean_token_accuracy": 0.16377224773168564,
      "num_tokens": 11455997.0,
      "step": 4515
    },
    {
      "entropy": 6.550042772293091,
      "epoch": 0.5216387766878245,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981283138538675,
      "loss": 6.4393,
      "mean_token_accuracy": 0.15353290140628814,
      "num_tokens": 11468271.0,
      "step": 4520
    },
    {
      "entropy": 6.545131015777588,
      "epoch": 0.5222158107328332,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498122998686747,
      "loss": 6.3608,
      "mean_token_accuracy": 0.1579316884279251,
      "num_tokens": 11479729.0,
      "step": 4525
    },
    {
      "entropy": 6.409595155715943,
      "epoch": 0.5227928447778419,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981176760149951,
      "loss": 6.2848,
      "mean_token_accuracy": 0.16170340776443481,
      "num_tokens": 11492261.0,
      "step": 4530
    },
    {
      "entropy": 6.509868383407593,
      "epoch": 0.5233698788228506,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004981123458387911,
      "loss": 6.3244,
      "mean_token_accuracy": 0.15999490320682525,
      "num_tokens": 11504802.0,
      "step": 4535
    },
    {
      "entropy": 6.5788795948028564,
      "epoch": 0.5239469128678592,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004981070081583142,
      "loss": 6.3737,
      "mean_token_accuracy": 0.1548797994852066,
      "num_tokens": 11518105.0,
      "step": 4540
    },
    {
      "entropy": 6.496658039093018,
      "epoch": 0.5245239469128679,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000498101662973744,
      "loss": 6.3673,
      "mean_token_accuracy": 0.15746258497238158,
      "num_tokens": 11532085.0,
      "step": 4545
    },
    {
      "entropy": 6.491401195526123,
      "epoch": 0.5251009809578765,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00049809631028526,
      "loss": 6.3211,
      "mean_token_accuracy": 0.16515219062566758,
      "num_tokens": 11544328.0,
      "step": 4550
    },
    {
      "entropy": 6.479079627990723,
      "epoch": 0.5256780150028851,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980909500930424,
      "loss": 6.4056,
      "mean_token_accuracy": 0.16223204433917998,
      "num_tokens": 11556496.0,
      "step": 4555
    },
    {
      "entropy": 6.618425369262695,
      "epoch": 0.5262550490478938,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004980855823972717,
      "loss": 6.4346,
      "mean_token_accuracy": 0.1574171245098114,
      "num_tokens": 11569330.0,
      "step": 4560
    },
    {
      "entropy": 6.5034425258636475,
      "epoch": 0.5268320830929025,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498080207198128,
      "loss": 6.3467,
      "mean_token_accuracy": 0.1647966295480728,
      "num_tokens": 11581335.0,
      "step": 4565
    },
    {
      "entropy": 6.504008865356445,
      "epoch": 0.5274091171379112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980748244957925,
      "loss": 6.3168,
      "mean_token_accuracy": 0.16169347763061523,
      "num_tokens": 11593130.0,
      "step": 4570
    },
    {
      "entropy": 6.4450465679168705,
      "epoch": 0.5279861511829198,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980694342904461,
      "loss": 6.3261,
      "mean_token_accuracy": 0.1652020573616028,
      "num_tokens": 11605140.0,
      "step": 4575
    },
    {
      "entropy": 6.508342981338501,
      "epoch": 0.5285631852279284,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004980640365822701,
      "loss": 6.3485,
      "mean_token_accuracy": 0.1583593711256981,
      "num_tokens": 11617361.0,
      "step": 4580
    },
    {
      "entropy": 6.490172624588013,
      "epoch": 0.5291402192729371,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498058631371446,
      "loss": 6.347,
      "mean_token_accuracy": 0.16494448632001876,
      "num_tokens": 11630780.0,
      "step": 4585
    },
    {
      "entropy": 6.413460969924927,
      "epoch": 0.5297172533179457,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980532186581555,
      "loss": 6.3278,
      "mean_token_accuracy": 0.1616680383682251,
      "num_tokens": 11643120.0,
      "step": 4590
    },
    {
      "entropy": 6.552847623825073,
      "epoch": 0.5302942873629544,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498047798442581,
      "loss": 6.3532,
      "mean_token_accuracy": 0.16128588616847991,
      "num_tokens": 11655849.0,
      "step": 4595
    },
    {
      "entropy": 6.4749092102050785,
      "epoch": 0.5308713214079631,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980423707249044,
      "loss": 6.3418,
      "mean_token_accuracy": 0.16079462617635726,
      "num_tokens": 11669072.0,
      "step": 4600
    },
    {
      "entropy": 6.412943744659424,
      "epoch": 0.5314483554529718,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980369355053086,
      "loss": 6.2983,
      "mean_token_accuracy": 0.16623102575540544,
      "num_tokens": 11680253.0,
      "step": 4605
    },
    {
      "entropy": 6.483307981491089,
      "epoch": 0.5320253894979804,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980314927839763,
      "loss": 6.2809,
      "mean_token_accuracy": 0.1597128376364708,
      "num_tokens": 11691465.0,
      "step": 4610
    },
    {
      "entropy": 6.494303131103516,
      "epoch": 0.532602423542989,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980260425610903,
      "loss": 6.4205,
      "mean_token_accuracy": 0.1597208321094513,
      "num_tokens": 11703679.0,
      "step": 4615
    },
    {
      "entropy": 6.54627251625061,
      "epoch": 0.5331794575879977,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004980205848368343,
      "loss": 6.3883,
      "mean_token_accuracy": 0.1554752230644226,
      "num_tokens": 11716204.0,
      "step": 4620
    },
    {
      "entropy": 6.548814105987549,
      "epoch": 0.5337564916330063,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004980151196113917,
      "loss": 6.3566,
      "mean_token_accuracy": 0.16065380424261094,
      "num_tokens": 11728404.0,
      "step": 4625
    },
    {
      "entropy": 6.456133556365967,
      "epoch": 0.534333525678015,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980096468849464,
      "loss": 6.3035,
      "mean_token_accuracy": 0.16273082345724105,
      "num_tokens": 11740170.0,
      "step": 4630
    },
    {
      "entropy": 6.4441171169281,
      "epoch": 0.5349105597230237,
      "grad_norm": 0.75,
      "learning_rate": 0.0004980041666576823,
      "loss": 6.2784,
      "mean_token_accuracy": 0.1635786160826683,
      "num_tokens": 11753062.0,
      "step": 4635
    },
    {
      "entropy": 6.488966655731201,
      "epoch": 0.5354875937680323,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004979986789297837,
      "loss": 6.2851,
      "mean_token_accuracy": 0.16575752943754196,
      "num_tokens": 11765588.0,
      "step": 4640
    },
    {
      "entropy": 6.490046072006225,
      "epoch": 0.536064627813041,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979931837014355,
      "loss": 6.4029,
      "mean_token_accuracy": 0.16148507297039033,
      "num_tokens": 11777358.0,
      "step": 4645
    },
    {
      "entropy": 6.561558675765991,
      "epoch": 0.5366416618580496,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979876809728223,
      "loss": 6.4261,
      "mean_token_accuracy": 0.1540203183889389,
      "num_tokens": 11791572.0,
      "step": 4650
    },
    {
      "entropy": 6.501104307174683,
      "epoch": 0.5372186959030583,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004979821707441292,
      "loss": 6.4531,
      "mean_token_accuracy": 0.15564129799604415,
      "num_tokens": 11805198.0,
      "step": 4655
    },
    {
      "entropy": 6.538952589035034,
      "epoch": 0.5377957299480669,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004979766530155416,
      "loss": 6.3667,
      "mean_token_accuracy": 0.1528536558151245,
      "num_tokens": 11817882.0,
      "step": 4660
    },
    {
      "entropy": 6.501694774627685,
      "epoch": 0.5383727639930755,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979711277872449,
      "loss": 6.2803,
      "mean_token_accuracy": 0.16876962184906005,
      "num_tokens": 11830334.0,
      "step": 4665
    },
    {
      "entropy": 6.409752893447876,
      "epoch": 0.5389497980380843,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979655950594252,
      "loss": 6.355,
      "mean_token_accuracy": 0.16761913448572158,
      "num_tokens": 11843057.0,
      "step": 4670
    },
    {
      "entropy": 6.5810261249542235,
      "epoch": 0.5395268320830929,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979600548322684,
      "loss": 6.4538,
      "mean_token_accuracy": 0.14982511550188066,
      "num_tokens": 11856086.0,
      "step": 4675
    },
    {
      "entropy": 6.520935726165772,
      "epoch": 0.5401038661281016,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000497954507105961,
      "loss": 6.3803,
      "mean_token_accuracy": 0.1535944387316704,
      "num_tokens": 11867799.0,
      "step": 4680
    },
    {
      "entropy": 6.440142202377319,
      "epoch": 0.5406809001731102,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004979489518806893,
      "loss": 6.2902,
      "mean_token_accuracy": 0.16890944987535478,
      "num_tokens": 11881488.0,
      "step": 4685
    },
    {
      "entropy": 6.544677066802978,
      "epoch": 0.5412579342181189,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004979433891566405,
      "loss": 6.3851,
      "mean_token_accuracy": 0.1536295682191849,
      "num_tokens": 11893495.0,
      "step": 4690
    },
    {
      "entropy": 6.500098705291748,
      "epoch": 0.5418349682631275,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004979378189340015,
      "loss": 6.3307,
      "mean_token_accuracy": 0.15971371084451674,
      "num_tokens": 11905456.0,
      "step": 4695
    },
    {
      "entropy": 6.484611320495605,
      "epoch": 0.5424120023081361,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004979322412129597,
      "loss": 6.3771,
      "mean_token_accuracy": 0.15654705613851547,
      "num_tokens": 11917987.0,
      "step": 4700
    },
    {
      "entropy": 6.506336069107055,
      "epoch": 0.5429890363531449,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979266559937028,
      "loss": 6.3483,
      "mean_token_accuracy": 0.16230462193489076,
      "num_tokens": 11930104.0,
      "step": 4705
    },
    {
      "entropy": 6.472575616836548,
      "epoch": 0.5435660703981535,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979210632764185,
      "loss": 6.3738,
      "mean_token_accuracy": 0.1527239166200161,
      "num_tokens": 11942841.0,
      "step": 4710
    },
    {
      "entropy": 6.521577930450439,
      "epoch": 0.5441431044431622,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004979154630612949,
      "loss": 6.3593,
      "mean_token_accuracy": 0.15818774700164795,
      "num_tokens": 11955509.0,
      "step": 4715
    },
    {
      "entropy": 6.502092933654785,
      "epoch": 0.5447201384881708,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004979098553485205,
      "loss": 6.3306,
      "mean_token_accuracy": 0.16365617513656616,
      "num_tokens": 11967389.0,
      "step": 4720
    },
    {
      "entropy": 6.537006950378418,
      "epoch": 0.5452971725331794,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004979042401382838,
      "loss": 6.3987,
      "mean_token_accuracy": 0.1641388550400734,
      "num_tokens": 11980173.0,
      "step": 4725
    },
    {
      "entropy": 6.552853393554687,
      "epoch": 0.5458742065781881,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004978986174307737,
      "loss": 6.3553,
      "mean_token_accuracy": 0.16312003284692764,
      "num_tokens": 11993039.0,
      "step": 4730
    },
    {
      "entropy": 6.406351327896118,
      "epoch": 0.5464512406231967,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004978929872261794,
      "loss": 6.2539,
      "mean_token_accuracy": 0.16683352440595628,
      "num_tokens": 12005861.0,
      "step": 4735
    },
    {
      "entropy": 6.361184930801391,
      "epoch": 0.5470282746682055,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004978873495246901,
      "loss": 6.2061,
      "mean_token_accuracy": 0.16693792939186097,
      "num_tokens": 12018137.0,
      "step": 4740
    },
    {
      "entropy": 6.528253793716431,
      "epoch": 0.5476053087132141,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978817043264955,
      "loss": 6.3639,
      "mean_token_accuracy": 0.15873690098524093,
      "num_tokens": 12030086.0,
      "step": 4745
    },
    {
      "entropy": 6.432341146469116,
      "epoch": 0.5481823427582228,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004978760516317856,
      "loss": 6.3288,
      "mean_token_accuracy": 0.16781895607709885,
      "num_tokens": 12043166.0,
      "step": 4750
    },
    {
      "entropy": 6.433397436141968,
      "epoch": 0.5487593768032314,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004978703914407503,
      "loss": 6.2825,
      "mean_token_accuracy": 0.16394588202238083,
      "num_tokens": 12055373.0,
      "step": 4755
    },
    {
      "entropy": 6.456381177902221,
      "epoch": 0.54933641084824,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978647237535802,
      "loss": 6.2905,
      "mean_token_accuracy": 0.16131811887025832,
      "num_tokens": 12066369.0,
      "step": 4760
    },
    {
      "entropy": 6.437442588806152,
      "epoch": 0.5499134448932487,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978590485704657,
      "loss": 6.2174,
      "mean_token_accuracy": 0.16695230156183244,
      "num_tokens": 12077939.0,
      "step": 4765
    },
    {
      "entropy": 6.559174108505249,
      "epoch": 0.5504904789382573,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978533658915979,
      "loss": 6.4118,
      "mean_token_accuracy": 0.15113805159926413,
      "num_tokens": 12091102.0,
      "step": 4770
    },
    {
      "entropy": 6.536936664581299,
      "epoch": 0.5510675129832661,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978476757171678,
      "loss": 6.3323,
      "mean_token_accuracy": 0.16064829379320145,
      "num_tokens": 12103144.0,
      "step": 4775
    },
    {
      "entropy": 6.460710620880127,
      "epoch": 0.5516445470282747,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978419780473668,
      "loss": 6.404,
      "mean_token_accuracy": 0.16159536987543105,
      "num_tokens": 12116664.0,
      "step": 4780
    },
    {
      "entropy": 6.526265573501587,
      "epoch": 0.5522215810732833,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004978362728823865,
      "loss": 6.3274,
      "mean_token_accuracy": 0.1602063536643982,
      "num_tokens": 12129656.0,
      "step": 4785
    },
    {
      "entropy": 6.4396782398223875,
      "epoch": 0.552798615118292,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978305602224189,
      "loss": 6.29,
      "mean_token_accuracy": 0.16184550821781157,
      "num_tokens": 12143221.0,
      "step": 4790
    },
    {
      "entropy": 6.490671825408936,
      "epoch": 0.5533756491633006,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004978248400676562,
      "loss": 6.2565,
      "mean_token_accuracy": 0.16548528671264648,
      "num_tokens": 12156409.0,
      "step": 4795
    },
    {
      "entropy": 6.3731786727905275,
      "epoch": 0.5539526832083093,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978191124182905,
      "loss": 6.1841,
      "mean_token_accuracy": 0.17455049157142638,
      "num_tokens": 12169589.0,
      "step": 4800
    },
    {
      "entropy": 6.4394989013671875,
      "epoch": 0.5545297172533179,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004978133772745149,
      "loss": 6.33,
      "mean_token_accuracy": 0.157976932823658,
      "num_tokens": 12182692.0,
      "step": 4805
    },
    {
      "entropy": 6.476097345352173,
      "epoch": 0.5551067512983267,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004978076346365218,
      "loss": 6.2889,
      "mean_token_accuracy": 0.16502978950738906,
      "num_tokens": 12195892.0,
      "step": 4810
    },
    {
      "entropy": 6.5237884521484375,
      "epoch": 0.5556837853433353,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004978018845045047,
      "loss": 6.3083,
      "mean_token_accuracy": 0.1582992434501648,
      "num_tokens": 12208974.0,
      "step": 4815
    },
    {
      "entropy": 6.4996720314025875,
      "epoch": 0.5562608193883439,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004977961268786568,
      "loss": 6.38,
      "mean_token_accuracy": 0.16242347955703734,
      "num_tokens": 12221342.0,
      "step": 4820
    },
    {
      "entropy": 6.446277618408203,
      "epoch": 0.5568378534333526,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977903617591719,
      "loss": 6.3611,
      "mean_token_accuracy": 0.15908932238817214,
      "num_tokens": 12233778.0,
      "step": 4825
    },
    {
      "entropy": 6.4188783168792725,
      "epoch": 0.5574148874783612,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977845891462439,
      "loss": 6.2051,
      "mean_token_accuracy": 0.16686934977769852,
      "num_tokens": 12246001.0,
      "step": 4830
    },
    {
      "entropy": 6.4489585876464846,
      "epoch": 0.5579919215233698,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977788090400668,
      "loss": 6.2398,
      "mean_token_accuracy": 0.16227732598781586,
      "num_tokens": 12258677.0,
      "step": 4835
    },
    {
      "entropy": 6.486279916763306,
      "epoch": 0.5585689555683785,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004977730214408352,
      "loss": 6.3491,
      "mean_token_accuracy": 0.15934401601552964,
      "num_tokens": 12270786.0,
      "step": 4840
    },
    {
      "entropy": 6.433901739120484,
      "epoch": 0.5591459896133872,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004977672263487435,
      "loss": 6.355,
      "mean_token_accuracy": 0.16431848853826522,
      "num_tokens": 12284439.0,
      "step": 4845
    },
    {
      "entropy": 6.502217817306518,
      "epoch": 0.5597230236583959,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000497761423763987,
      "loss": 6.3411,
      "mean_token_accuracy": 0.1564551204442978,
      "num_tokens": 12297005.0,
      "step": 4850
    },
    {
      "entropy": 6.517484569549561,
      "epoch": 0.5603000577034045,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004977556136867606,
      "loss": 6.352,
      "mean_token_accuracy": 0.1582645058631897,
      "num_tokens": 12308855.0,
      "step": 4855
    },
    {
      "entropy": 6.514178133010864,
      "epoch": 0.5608770917484132,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977497961172598,
      "loss": 6.3074,
      "mean_token_accuracy": 0.1609992191195488,
      "num_tokens": 12321120.0,
      "step": 4860
    },
    {
      "entropy": 6.476286888122559,
      "epoch": 0.5614541257934218,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004977439710556802,
      "loss": 6.3753,
      "mean_token_accuracy": 0.15675147473812104,
      "num_tokens": 12333793.0,
      "step": 4865
    },
    {
      "entropy": 6.429763603210449,
      "epoch": 0.5620311598384304,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004977381385022179,
      "loss": 6.2695,
      "mean_token_accuracy": 0.1623919278383255,
      "num_tokens": 12345998.0,
      "step": 4870
    },
    {
      "entropy": 6.461885595321656,
      "epoch": 0.5626081938834391,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004977322984570688,
      "loss": 6.3018,
      "mean_token_accuracy": 0.16285305768251418,
      "num_tokens": 12358580.0,
      "step": 4875
    },
    {
      "entropy": 6.526036548614502,
      "epoch": 0.5631852279284478,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004977264509204296,
      "loss": 6.3799,
      "mean_token_accuracy": 0.16008112132549285,
      "num_tokens": 12370867.0,
      "step": 4880
    },
    {
      "entropy": 6.5039674758911135,
      "epoch": 0.5637622619734565,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004977205958924967,
      "loss": 6.3593,
      "mean_token_accuracy": 0.15953048765659333,
      "num_tokens": 12384086.0,
      "step": 4885
    },
    {
      "entropy": 6.387411832809448,
      "epoch": 0.5643392960184651,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977147333734673,
      "loss": 6.2159,
      "mean_token_accuracy": 0.16773999333381653,
      "num_tokens": 12396852.0,
      "step": 4890
    },
    {
      "entropy": 6.491942930221557,
      "epoch": 0.5649163300634737,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004977088633635385,
      "loss": 6.3103,
      "mean_token_accuracy": 0.16319107711315156,
      "num_tokens": 12409924.0,
      "step": 4895
    },
    {
      "entropy": 6.351782751083374,
      "epoch": 0.5654933641084824,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977029858629076,
      "loss": 6.2794,
      "mean_token_accuracy": 0.16879811882972717,
      "num_tokens": 12422218.0,
      "step": 4900
    },
    {
      "entropy": 6.413300466537476,
      "epoch": 0.566070398153491,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004976971008717726,
      "loss": 6.3014,
      "mean_token_accuracy": 0.1605297327041626,
      "num_tokens": 12433522.0,
      "step": 4905
    },
    {
      "entropy": 6.64859299659729,
      "epoch": 0.5666474321984997,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004976912083903309,
      "loss": 6.3665,
      "mean_token_accuracy": 0.15510098934173583,
      "num_tokens": 12445526.0,
      "step": 4910
    },
    {
      "entropy": 6.437649726867676,
      "epoch": 0.5672244662435084,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004976853084187814,
      "loss": 6.3196,
      "mean_token_accuracy": 0.16614574939012527,
      "num_tokens": 12458403.0,
      "step": 4915
    },
    {
      "entropy": 6.522170686721802,
      "epoch": 0.5678015002885171,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004976794009573219,
      "loss": 6.3775,
      "mean_token_accuracy": 0.15785618424415587,
      "num_tokens": 12470437.0,
      "step": 4920
    },
    {
      "entropy": 6.512676572799682,
      "epoch": 0.5683785343335257,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004976734860061515,
      "loss": 6.3134,
      "mean_token_accuracy": 0.1652366876602173,
      "num_tokens": 12483345.0,
      "step": 4925
    },
    {
      "entropy": 6.425691747665406,
      "epoch": 0.5689555683785343,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004976675635654688,
      "loss": 6.3125,
      "mean_token_accuracy": 0.16011232286691665,
      "num_tokens": 12495675.0,
      "step": 4930
    },
    {
      "entropy": 6.464152193069458,
      "epoch": 0.569532602423543,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976616336354733,
      "loss": 6.327,
      "mean_token_accuracy": 0.16153873652219772,
      "num_tokens": 12507569.0,
      "step": 4935
    },
    {
      "entropy": 6.4666400909423825,
      "epoch": 0.5701096364685516,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004976556962163645,
      "loss": 6.3759,
      "mean_token_accuracy": 0.16025069802999498,
      "num_tokens": 12521198.0,
      "step": 4940
    },
    {
      "entropy": 6.412058877944946,
      "epoch": 0.5706866705135603,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976497513083419,
      "loss": 6.293,
      "mean_token_accuracy": 0.16026601642370225,
      "num_tokens": 12533242.0,
      "step": 4945
    },
    {
      "entropy": 6.520112228393555,
      "epoch": 0.571263704558569,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004976437989116055,
      "loss": 6.4046,
      "mean_token_accuracy": 0.16085704267024994,
      "num_tokens": 12545918.0,
      "step": 4950
    },
    {
      "entropy": 6.492485094070434,
      "epoch": 0.5718407386035776,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004976378390263554,
      "loss": 6.3433,
      "mean_token_accuracy": 0.1590990960597992,
      "num_tokens": 12558090.0,
      "step": 4955
    },
    {
      "entropy": 6.503842449188232,
      "epoch": 0.5724177726485863,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004976318716527924,
      "loss": 6.3572,
      "mean_token_accuracy": 0.16237506717443467,
      "num_tokens": 12571453.0,
      "step": 4960
    },
    {
      "entropy": 6.507378005981446,
      "epoch": 0.5729948066935949,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004976258967911168,
      "loss": 6.3148,
      "mean_token_accuracy": 0.15835612416267394,
      "num_tokens": 12584762.0,
      "step": 4965
    },
    {
      "entropy": 6.498177242279053,
      "epoch": 0.5735718407386036,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976199144415298,
      "loss": 6.304,
      "mean_token_accuracy": 0.15529713779687881,
      "num_tokens": 12597224.0,
      "step": 4970
    },
    {
      "entropy": 6.410339307785034,
      "epoch": 0.5741488747836122,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004976139246042325,
      "loss": 6.3678,
      "mean_token_accuracy": 0.1593444064259529,
      "num_tokens": 12609837.0,
      "step": 4975
    },
    {
      "entropy": 6.522058153152466,
      "epoch": 0.5747259088286208,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976079272794265,
      "loss": 6.402,
      "mean_token_accuracy": 0.14933762848377227,
      "num_tokens": 12622596.0,
      "step": 4980
    },
    {
      "entropy": 6.514160966873169,
      "epoch": 0.5753029428736296,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004976019224673134,
      "loss": 6.3048,
      "mean_token_accuracy": 0.1668033003807068,
      "num_tokens": 12634721.0,
      "step": 4985
    },
    {
      "entropy": 6.438699054718017,
      "epoch": 0.5758799769186382,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975959101680953,
      "loss": 6.3366,
      "mean_token_accuracy": 0.15978636145591735,
      "num_tokens": 12646389.0,
      "step": 4990
    },
    {
      "entropy": 6.509289073944092,
      "epoch": 0.5764570109636469,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004975898903819742,
      "loss": 6.4231,
      "mean_token_accuracy": 0.1489159494638443,
      "num_tokens": 12661081.0,
      "step": 4995
    },
    {
      "entropy": 6.628030252456665,
      "epoch": 0.5770340450086555,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004975838631091527,
      "loss": 6.367,
      "mean_token_accuracy": 0.15845488607883454,
      "num_tokens": 12672505.0,
      "step": 5000
    },
    {
      "entropy": 6.462829971313477,
      "epoch": 0.5776110790536642,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004975778283498336,
      "loss": 6.3748,
      "mean_token_accuracy": 0.15903144031763078,
      "num_tokens": 12684805.0,
      "step": 5005
    },
    {
      "entropy": 6.408445596694946,
      "epoch": 0.5781881130986728,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975717861042198,
      "loss": 6.3162,
      "mean_token_accuracy": 0.15930698812007904,
      "num_tokens": 12698557.0,
      "step": 5010
    },
    {
      "entropy": 6.5260742664337155,
      "epoch": 0.5787651471436814,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975657363725146,
      "loss": 6.3399,
      "mean_token_accuracy": 0.15470098406076432,
      "num_tokens": 12712472.0,
      "step": 5015
    },
    {
      "entropy": 6.462447595596314,
      "epoch": 0.5793421811886902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975596791549213,
      "loss": 6.211,
      "mean_token_accuracy": 0.16655874848365784,
      "num_tokens": 12725698.0,
      "step": 5020
    },
    {
      "entropy": 6.466801404953003,
      "epoch": 0.5799192152336988,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975536144516437,
      "loss": 6.3125,
      "mean_token_accuracy": 0.16142037212848664,
      "num_tokens": 12738126.0,
      "step": 5025
    },
    {
      "entropy": 6.51831431388855,
      "epoch": 0.5804962492787075,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497547542262886,
      "loss": 6.3282,
      "mean_token_accuracy": 0.17067889571189881,
      "num_tokens": 12751402.0,
      "step": 5030
    },
    {
      "entropy": 6.493655204772949,
      "epoch": 0.5810732833237161,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004975414625888521,
      "loss": 6.3458,
      "mean_token_accuracy": 0.1561669498682022,
      "num_tokens": 12763957.0,
      "step": 5035
    },
    {
      "entropy": 6.50370774269104,
      "epoch": 0.5816503173687247,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975353754297468,
      "loss": 6.2931,
      "mean_token_accuracy": 0.16369494050741196,
      "num_tokens": 12776498.0,
      "step": 5040
    },
    {
      "entropy": 6.447761821746826,
      "epoch": 0.5822273514137334,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004975292807857745,
      "loss": 6.3115,
      "mean_token_accuracy": 0.1632593885064125,
      "num_tokens": 12788498.0,
      "step": 5045
    },
    {
      "entropy": 6.385071325302124,
      "epoch": 0.582804385458742,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004975231786571406,
      "loss": 6.2291,
      "mean_token_accuracy": 0.16083939224481583,
      "num_tokens": 12799907.0,
      "step": 5050
    },
    {
      "entropy": 6.49673638343811,
      "epoch": 0.5833814195037508,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004975170690440499,
      "loss": 6.3411,
      "mean_token_accuracy": 0.16103529632091523,
      "num_tokens": 12812610.0,
      "step": 5055
    },
    {
      "entropy": 6.47235746383667,
      "epoch": 0.5839584535487594,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0004975109519467083,
      "loss": 6.3414,
      "mean_token_accuracy": 0.15436725318431854,
      "num_tokens": 12825082.0,
      "step": 5060
    },
    {
      "entropy": 6.404198455810547,
      "epoch": 0.584535487593768,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004975048273653212,
      "loss": 6.2609,
      "mean_token_accuracy": 0.16827418953180312,
      "num_tokens": 12837999.0,
      "step": 5065
    },
    {
      "entropy": 6.43675537109375,
      "epoch": 0.5851125216387767,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004974986953000948,
      "loss": 6.2143,
      "mean_token_accuracy": 0.17144626528024673,
      "num_tokens": 12850020.0,
      "step": 5070
    },
    {
      "entropy": 6.433012771606445,
      "epoch": 0.5856895556837853,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004974925557512354,
      "loss": 6.3105,
      "mean_token_accuracy": 0.164590884745121,
      "num_tokens": 12863818.0,
      "step": 5075
    },
    {
      "entropy": 6.489244079589843,
      "epoch": 0.586266589728794,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004974864087189493,
      "loss": 6.3248,
      "mean_token_accuracy": 0.15803768932819368,
      "num_tokens": 12876959.0,
      "step": 5080
    },
    {
      "entropy": 6.465778827667236,
      "epoch": 0.5868436237738026,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004974802542034434,
      "loss": 6.2942,
      "mean_token_accuracy": 0.16098903566598893,
      "num_tokens": 12889817.0,
      "step": 5085
    },
    {
      "entropy": 6.398091459274292,
      "epoch": 0.5874206578188114,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004974740922049246,
      "loss": 6.2775,
      "mean_token_accuracy": 0.1631662592291832,
      "num_tokens": 12901438.0,
      "step": 5090
    },
    {
      "entropy": 6.457873916625976,
      "epoch": 0.58799769186382,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974679227236004,
      "loss": 6.2088,
      "mean_token_accuracy": 0.17070800065994263,
      "num_tokens": 12914115.0,
      "step": 5095
    },
    {
      "entropy": 6.494083499908447,
      "epoch": 0.5885747259088286,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974617457596779,
      "loss": 6.2744,
      "mean_token_accuracy": 0.16690096408128738,
      "num_tokens": 12926822.0,
      "step": 5100
    },
    {
      "entropy": 6.405472755432129,
      "epoch": 0.5891517599538373,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004974555613133652,
      "loss": 6.2965,
      "mean_token_accuracy": 0.16091533601284028,
      "num_tokens": 12939515.0,
      "step": 5105
    },
    {
      "entropy": 6.300894165039063,
      "epoch": 0.5897287939988459,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974493693848702,
      "loss": 6.2228,
      "mean_token_accuracy": 0.18208237290382384,
      "num_tokens": 12952195.0,
      "step": 5110
    },
    {
      "entropy": 6.482909345626831,
      "epoch": 0.5903058280438546,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004974431699744011,
      "loss": 6.2629,
      "mean_token_accuracy": 0.16298617869615556,
      "num_tokens": 12964563.0,
      "step": 5115
    },
    {
      "entropy": 6.413863277435302,
      "epoch": 0.5908828620888632,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004974369630821665,
      "loss": 6.1888,
      "mean_token_accuracy": 0.1695662796497345,
      "num_tokens": 12976564.0,
      "step": 5120
    },
    {
      "entropy": 6.247827005386353,
      "epoch": 0.5914598961338718,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004974307487083752,
      "loss": 6.1389,
      "mean_token_accuracy": 0.1661507621407509,
      "num_tokens": 12989757.0,
      "step": 5125
    },
    {
      "entropy": 6.4758203506469725,
      "epoch": 0.5920369301788806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004974245268532361,
      "loss": 6.3533,
      "mean_token_accuracy": 0.16023973524570465,
      "num_tokens": 13003846.0,
      "step": 5130
    },
    {
      "entropy": 6.550791883468628,
      "epoch": 0.5926139642238892,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004974182975169585,
      "loss": 6.3555,
      "mean_token_accuracy": 0.1563681572675705,
      "num_tokens": 13018389.0,
      "step": 5135
    },
    {
      "entropy": 6.484573125839233,
      "epoch": 0.5931909982688979,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000497412060699752,
      "loss": 6.3554,
      "mean_token_accuracy": 0.15696738362312318,
      "num_tokens": 13031893.0,
      "step": 5140
    },
    {
      "entropy": 6.476515531539917,
      "epoch": 0.5937680323139065,
      "grad_norm": 0.75,
      "learning_rate": 0.0004974058164018263,
      "loss": 6.2922,
      "mean_token_accuracy": 0.167695315182209,
      "num_tokens": 13045916.0,
      "step": 5145
    },
    {
      "entropy": 6.510441398620605,
      "epoch": 0.5943450663589152,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004973995646233914,
      "loss": 6.3366,
      "mean_token_accuracy": 0.15636033564805984,
      "num_tokens": 13058468.0,
      "step": 5150
    },
    {
      "entropy": 6.487969064712525,
      "epoch": 0.5949221004039238,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004973933053646576,
      "loss": 6.3532,
      "mean_token_accuracy": 0.15452814549207688,
      "num_tokens": 13070778.0,
      "step": 5155
    },
    {
      "entropy": 6.47402811050415,
      "epoch": 0.5954991344489324,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004973870386258355,
      "loss": 6.2298,
      "mean_token_accuracy": 0.16576715409755707,
      "num_tokens": 13082287.0,
      "step": 5160
    },
    {
      "entropy": 6.364145660400391,
      "epoch": 0.5960761684939412,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004973807644071357,
      "loss": 6.2377,
      "mean_token_accuracy": 0.17309042662382126,
      "num_tokens": 13095505.0,
      "step": 5165
    },
    {
      "entropy": 6.383938789367676,
      "epoch": 0.5966532025389498,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973744827087695,
      "loss": 6.2689,
      "mean_token_accuracy": 0.16178690791130065,
      "num_tokens": 13108351.0,
      "step": 5170
    },
    {
      "entropy": 6.492494821548462,
      "epoch": 0.5972302365839585,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497368193530948,
      "loss": 6.3229,
      "mean_token_accuracy": 0.16320233643054963,
      "num_tokens": 13120941.0,
      "step": 5175
    },
    {
      "entropy": 6.459748411178589,
      "epoch": 0.5978072706289671,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004973618968738828,
      "loss": 6.2566,
      "mean_token_accuracy": 0.16999219357967377,
      "num_tokens": 13134217.0,
      "step": 5180
    },
    {
      "entropy": 6.392143726348877,
      "epoch": 0.5983843046739757,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004973555927377856,
      "loss": 6.2801,
      "mean_token_accuracy": 0.16705387085676193,
      "num_tokens": 13146245.0,
      "step": 5185
    },
    {
      "entropy": 6.4566864490509035,
      "epoch": 0.5989613387189844,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973492811228685,
      "loss": 6.3066,
      "mean_token_accuracy": 0.16118625104427337,
      "num_tokens": 13158498.0,
      "step": 5190
    },
    {
      "entropy": 6.467262935638428,
      "epoch": 0.599538372763993,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004973429620293438,
      "loss": 6.3382,
      "mean_token_accuracy": 0.157435604929924,
      "num_tokens": 13170464.0,
      "step": 5195
    },
    {
      "entropy": 6.527691745758057,
      "epoch": 0.6001154068090018,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973366354574239,
      "loss": 6.3898,
      "mean_token_accuracy": 0.15946318507194518,
      "num_tokens": 13183315.0,
      "step": 5200
    },
    {
      "entropy": 6.526863527297974,
      "epoch": 0.6006924408540104,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004973303014073219,
      "loss": 6.4012,
      "mean_token_accuracy": 0.15497395694255828,
      "num_tokens": 13196536.0,
      "step": 5205
    },
    {
      "entropy": 6.493622207641602,
      "epoch": 0.601269474899019,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004973239598792504,
      "loss": 6.2671,
      "mean_token_accuracy": 0.16305325478315352,
      "num_tokens": 13209159.0,
      "step": 5210
    },
    {
      "entropy": 6.512037229537964,
      "epoch": 0.6018465089440277,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973176108734232,
      "loss": 6.3093,
      "mean_token_accuracy": 0.1653384745121002,
      "num_tokens": 13221563.0,
      "step": 5215
    },
    {
      "entropy": 6.491204500198364,
      "epoch": 0.6024235429890363,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973112543900535,
      "loss": 6.3794,
      "mean_token_accuracy": 0.15575674176216125,
      "num_tokens": 13234209.0,
      "step": 5220
    },
    {
      "entropy": 6.569083213806152,
      "epoch": 0.603000577034045,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973048904293551,
      "loss": 6.2786,
      "mean_token_accuracy": 0.15728528350591658,
      "num_tokens": 13246198.0,
      "step": 5225
    },
    {
      "entropy": 6.410029506683349,
      "epoch": 0.6035776110790536,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004972985189915422,
      "loss": 6.2899,
      "mean_token_accuracy": 0.16580265015363693,
      "num_tokens": 13259149.0,
      "step": 5230
    },
    {
      "entropy": 6.415045642852784,
      "epoch": 0.6041546451240624,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000497292140076829,
      "loss": 6.2889,
      "mean_token_accuracy": 0.15860007032752038,
      "num_tokens": 13271138.0,
      "step": 5235
    },
    {
      "entropy": 6.492413473129273,
      "epoch": 0.604731679169071,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972857536854301,
      "loss": 6.2941,
      "mean_token_accuracy": 0.15741852372884751,
      "num_tokens": 13283639.0,
      "step": 5240
    },
    {
      "entropy": 6.446307229995727,
      "epoch": 0.6053087132140796,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972793598175603,
      "loss": 6.3186,
      "mean_token_accuracy": 0.16257961988449096,
      "num_tokens": 13295322.0,
      "step": 5245
    },
    {
      "entropy": 6.433833169937134,
      "epoch": 0.6058857472590883,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004972729584734347,
      "loss": 6.2716,
      "mean_token_accuracy": 0.16428455114364623,
      "num_tokens": 13308431.0,
      "step": 5250
    },
    {
      "entropy": 6.442098903656006,
      "epoch": 0.6064627813040969,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972665496532685,
      "loss": 6.2113,
      "mean_token_accuracy": 0.16595556437969208,
      "num_tokens": 13321878.0,
      "step": 5255
    },
    {
      "entropy": 6.385565567016601,
      "epoch": 0.6070398153491056,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004972601333572774,
      "loss": 6.3042,
      "mean_token_accuracy": 0.1635686531662941,
      "num_tokens": 13334166.0,
      "step": 5260
    },
    {
      "entropy": 6.534063529968262,
      "epoch": 0.6076168493941142,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004972537095856769,
      "loss": 6.3566,
      "mean_token_accuracy": 0.16068174988031386,
      "num_tokens": 13347704.0,
      "step": 5265
    },
    {
      "entropy": 6.432081651687622,
      "epoch": 0.608193883439123,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972472783386834,
      "loss": 6.2357,
      "mean_token_accuracy": 0.16729470044374467,
      "num_tokens": 13359659.0,
      "step": 5270
    },
    {
      "entropy": 6.412068843841553,
      "epoch": 0.6087709174841316,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004972408396165132,
      "loss": 6.2659,
      "mean_token_accuracy": 0.16037606298923493,
      "num_tokens": 13371952.0,
      "step": 5275
    },
    {
      "entropy": 6.51385087966919,
      "epoch": 0.6093479515291402,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004972343934193826,
      "loss": 6.288,
      "mean_token_accuracy": 0.16473593562841415,
      "num_tokens": 13385757.0,
      "step": 5280
    },
    {
      "entropy": 6.435386753082275,
      "epoch": 0.6099249855741489,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004972279397475086,
      "loss": 6.263,
      "mean_token_accuracy": 0.16103050708770753,
      "num_tokens": 13397657.0,
      "step": 5285
    },
    {
      "entropy": 6.36616096496582,
      "epoch": 0.6105020196191575,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004972214786011083,
      "loss": 6.272,
      "mean_token_accuracy": 0.16497953087091446,
      "num_tokens": 13410882.0,
      "step": 5290
    },
    {
      "entropy": 6.422442626953125,
      "epoch": 0.6110790536641661,
      "grad_norm": 0.796875,
      "learning_rate": 0.000497215009980399,
      "loss": 6.3103,
      "mean_token_accuracy": 0.15897688865661622,
      "num_tokens": 13423310.0,
      "step": 5295
    },
    {
      "entropy": 6.45792875289917,
      "epoch": 0.6116560877091748,
      "grad_norm": 0.73828125,
      "learning_rate": 0.000497208533885598,
      "loss": 6.2828,
      "mean_token_accuracy": 0.16484968066215516,
      "num_tokens": 13437837.0,
      "step": 5300
    },
    {
      "entropy": 6.504089260101319,
      "epoch": 0.6122331217541835,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972020503169235,
      "loss": 6.2873,
      "mean_token_accuracy": 0.16391800791025163,
      "num_tokens": 13450637.0,
      "step": 5305
    },
    {
      "entropy": 6.513200664520264,
      "epoch": 0.6128101557991922,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971955592745934,
      "loss": 6.3498,
      "mean_token_accuracy": 0.15506888329982757,
      "num_tokens": 13464429.0,
      "step": 5310
    },
    {
      "entropy": 6.454133605957031,
      "epoch": 0.6133871898442008,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497189060758826,
      "loss": 6.2231,
      "mean_token_accuracy": 0.16480964422225952,
      "num_tokens": 13476710.0,
      "step": 5315
    },
    {
      "entropy": 6.404278898239136,
      "epoch": 0.6139642238892095,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00049718255476984,
      "loss": 6.2663,
      "mean_token_accuracy": 0.16096706688404083,
      "num_tokens": 13489828.0,
      "step": 5320
    },
    {
      "entropy": 6.480580472946167,
      "epoch": 0.6145412579342181,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004971760413078539,
      "loss": 6.263,
      "mean_token_accuracy": 0.1644087016582489,
      "num_tokens": 13503584.0,
      "step": 5325
    },
    {
      "entropy": 6.404313373565674,
      "epoch": 0.6151182919792267,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971695203730872,
      "loss": 6.2618,
      "mean_token_accuracy": 0.1642356902360916,
      "num_tokens": 13514658.0,
      "step": 5330
    },
    {
      "entropy": 6.414643478393555,
      "epoch": 0.6156953260242354,
      "grad_norm": 0.828125,
      "learning_rate": 0.000497162991965759,
      "loss": 6.3014,
      "mean_token_accuracy": 0.17185865193605424,
      "num_tokens": 13526424.0,
      "step": 5335
    },
    {
      "entropy": 6.3885194778442385,
      "epoch": 0.6162723600692441,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004971564560860889,
      "loss": 6.2397,
      "mean_token_accuracy": 0.1621319070458412,
      "num_tokens": 13540515.0,
      "step": 5340
    },
    {
      "entropy": 6.484720373153687,
      "epoch": 0.6168493941142528,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004971499127342968,
      "loss": 6.3452,
      "mean_token_accuracy": 0.16039182096719742,
      "num_tokens": 13553634.0,
      "step": 5345
    },
    {
      "entropy": 6.4474649906158445,
      "epoch": 0.6174264281592614,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971433619106027,
      "loss": 6.3689,
      "mean_token_accuracy": 0.16041715145111085,
      "num_tokens": 13565168.0,
      "step": 5350
    },
    {
      "entropy": 6.404348373413086,
      "epoch": 0.61800346220427,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000497136803615227,
      "loss": 6.292,
      "mean_token_accuracy": 0.16248742789030074,
      "num_tokens": 13578511.0,
      "step": 5355
    },
    {
      "entropy": 6.452801609039307,
      "epoch": 0.6185804962492787,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004971302378483902,
      "loss": 6.2723,
      "mean_token_accuracy": 0.16374483555555344,
      "num_tokens": 13590535.0,
      "step": 5360
    },
    {
      "entropy": 6.416561508178711,
      "epoch": 0.6191575302942873,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004971236646103132,
      "loss": 6.2987,
      "mean_token_accuracy": 0.16325142830610276,
      "num_tokens": 13604549.0,
      "step": 5365
    },
    {
      "entropy": 6.45912880897522,
      "epoch": 0.619734564339296,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004971170839012171,
      "loss": 6.287,
      "mean_token_accuracy": 0.1573787122964859,
      "num_tokens": 13617658.0,
      "step": 5370
    },
    {
      "entropy": 6.402299642562866,
      "epoch": 0.6203115983843047,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971104957213233,
      "loss": 6.1915,
      "mean_token_accuracy": 0.1690565750002861,
      "num_tokens": 13630559.0,
      "step": 5375
    },
    {
      "entropy": 6.39977617263794,
      "epoch": 0.6208886324293134,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004971039000708531,
      "loss": 6.2731,
      "mean_token_accuracy": 0.1692923977971077,
      "num_tokens": 13643018.0,
      "step": 5380
    },
    {
      "entropy": 6.422565841674805,
      "epoch": 0.621465666474322,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004970972969500286,
      "loss": 6.2571,
      "mean_token_accuracy": 0.1639006808400154,
      "num_tokens": 13655053.0,
      "step": 5385
    },
    {
      "entropy": 6.4414605617523195,
      "epoch": 0.6220427005193306,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497090686359072,
      "loss": 6.2197,
      "mean_token_accuracy": 0.1663372591137886,
      "num_tokens": 13667420.0,
      "step": 5390
    },
    {
      "entropy": 6.375636005401612,
      "epoch": 0.6226197345643393,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004970840682982054,
      "loss": 6.2031,
      "mean_token_accuracy": 0.16604962944984436,
      "num_tokens": 13680662.0,
      "step": 5395
    },
    {
      "entropy": 6.403697204589844,
      "epoch": 0.6231967686093479,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004970774427676514,
      "loss": 6.3004,
      "mean_token_accuracy": 0.15936201810836792,
      "num_tokens": 13694720.0,
      "step": 5400
    },
    {
      "entropy": 6.441939163208008,
      "epoch": 0.6237738026543566,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497070809767633,
      "loss": 6.2243,
      "mean_token_accuracy": 0.16182428002357482,
      "num_tokens": 13706826.0,
      "step": 5405
    },
    {
      "entropy": 6.371305274963379,
      "epoch": 0.6243508366993653,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004970641692983731,
      "loss": 6.2852,
      "mean_token_accuracy": 0.16314123570919037,
      "num_tokens": 13719526.0,
      "step": 5410
    },
    {
      "entropy": 6.452666711807251,
      "epoch": 0.624927870744374,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970575213600954,
      "loss": 6.1842,
      "mean_token_accuracy": 0.17304476648569106,
      "num_tokens": 13731583.0,
      "step": 5415
    },
    {
      "entropy": 6.440716409683228,
      "epoch": 0.6255049047893826,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004970508659530231,
      "loss": 6.2966,
      "mean_token_accuracy": 0.1663545310497284,
      "num_tokens": 13743690.0,
      "step": 5420
    },
    {
      "entropy": 6.430726671218872,
      "epoch": 0.6260819388343912,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970442030773802,
      "loss": 6.2445,
      "mean_token_accuracy": 0.1653284952044487,
      "num_tokens": 13756264.0,
      "step": 5425
    },
    {
      "entropy": 6.448899984359741,
      "epoch": 0.6266589728793999,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970375327333909,
      "loss": 6.2768,
      "mean_token_accuracy": 0.1557541534304619,
      "num_tokens": 13768089.0,
      "step": 5430
    },
    {
      "entropy": 6.458224678039551,
      "epoch": 0.6272360069244085,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004970308549212796,
      "loss": 6.3077,
      "mean_token_accuracy": 0.15997258424758912,
      "num_tokens": 13780671.0,
      "step": 5435
    },
    {
      "entropy": 6.485933589935303,
      "epoch": 0.6278130409694171,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004970241696412705,
      "loss": 6.3001,
      "mean_token_accuracy": 0.15982202887535096,
      "num_tokens": 13793678.0,
      "step": 5440
    },
    {
      "entropy": 6.437276601791382,
      "epoch": 0.6283900750144259,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000497017476893589,
      "loss": 6.2395,
      "mean_token_accuracy": 0.16278810501098634,
      "num_tokens": 13807001.0,
      "step": 5445
    },
    {
      "entropy": 6.400468921661377,
      "epoch": 0.6289671090594345,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004970107766784598,
      "loss": 6.1994,
      "mean_token_accuracy": 0.166753850877285,
      "num_tokens": 13820227.0,
      "step": 5450
    },
    {
      "entropy": 6.477259588241577,
      "epoch": 0.6295441431044432,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004970040689961084,
      "loss": 6.3016,
      "mean_token_accuracy": 0.1641502395272255,
      "num_tokens": 13831663.0,
      "step": 5455
    },
    {
      "entropy": 6.331152725219726,
      "epoch": 0.6301211771494518,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004969973538467605,
      "loss": 6.1989,
      "mean_token_accuracy": 0.17121504694223405,
      "num_tokens": 13845173.0,
      "step": 5460
    },
    {
      "entropy": 6.439869022369384,
      "epoch": 0.6306982111944605,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969906312306419,
      "loss": 6.2332,
      "mean_token_accuracy": 0.1640544891357422,
      "num_tokens": 13857551.0,
      "step": 5465
    },
    {
      "entropy": 6.323156309127808,
      "epoch": 0.6312752452394691,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004969839011479786,
      "loss": 6.1566,
      "mean_token_accuracy": 0.1737811744213104,
      "num_tokens": 13870238.0,
      "step": 5470
    },
    {
      "entropy": 6.345542526245117,
      "epoch": 0.6318522792844777,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000496977163598997,
      "loss": 6.2456,
      "mean_token_accuracy": 0.16080596596002578,
      "num_tokens": 13883145.0,
      "step": 5475
    },
    {
      "entropy": 6.490610980987549,
      "epoch": 0.6324293133294865,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004969704185839239,
      "loss": 6.3086,
      "mean_token_accuracy": 0.15595276206731795,
      "num_tokens": 13895777.0,
      "step": 5480
    },
    {
      "entropy": 6.404451847076416,
      "epoch": 0.6330063473744951,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004969636661029859,
      "loss": 6.2006,
      "mean_token_accuracy": 0.17272072285413742,
      "num_tokens": 13907641.0,
      "step": 5485
    },
    {
      "entropy": 6.471062660217285,
      "epoch": 0.6335833814195038,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969569061564103,
      "loss": 6.2882,
      "mean_token_accuracy": 0.15671513378620147,
      "num_tokens": 13921576.0,
      "step": 5490
    },
    {
      "entropy": 6.4730736255645756,
      "epoch": 0.6341604154645124,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004969501387444245,
      "loss": 6.2504,
      "mean_token_accuracy": 0.16378487050533294,
      "num_tokens": 13935158.0,
      "step": 5495
    },
    {
      "entropy": 6.369989347457886,
      "epoch": 0.634737449509521,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004969433638672559,
      "loss": 6.1925,
      "mean_token_accuracy": 0.16805582195520402,
      "num_tokens": 13947803.0,
      "step": 5500
    },
    {
      "entropy": 6.393064069747925,
      "epoch": 0.6353144835545297,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004969365815251325,
      "loss": 6.2968,
      "mean_token_accuracy": 0.16199722290039062,
      "num_tokens": 13960573.0,
      "step": 5505
    },
    {
      "entropy": 6.419277477264404,
      "epoch": 0.6358915175995383,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969297917182825,
      "loss": 6.1607,
      "mean_token_accuracy": 0.1702695369720459,
      "num_tokens": 13970994.0,
      "step": 5510
    },
    {
      "entropy": 6.43389368057251,
      "epoch": 0.6364685516445471,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969229944469342,
      "loss": 6.3775,
      "mean_token_accuracy": 0.15624190866947174,
      "num_tokens": 13983617.0,
      "step": 5515
    },
    {
      "entropy": 6.450831413269043,
      "epoch": 0.6370455856895557,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004969161897113162,
      "loss": 6.2614,
      "mean_token_accuracy": 0.16926882266998292,
      "num_tokens": 13995779.0,
      "step": 5520
    },
    {
      "entropy": 6.405157804489136,
      "epoch": 0.6376226197345644,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004969093775116574,
      "loss": 6.2964,
      "mean_token_accuracy": 0.16423781961202621,
      "num_tokens": 14008650.0,
      "step": 5525
    },
    {
      "entropy": 6.346559667587281,
      "epoch": 0.638199653779573,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004969025578481869,
      "loss": 6.1963,
      "mean_token_accuracy": 0.16618811190128327,
      "num_tokens": 14021787.0,
      "step": 5530
    },
    {
      "entropy": 6.531115770339966,
      "epoch": 0.6387766878245816,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496895730721134,
      "loss": 6.3577,
      "mean_token_accuracy": 0.15821406841278077,
      "num_tokens": 14035036.0,
      "step": 5535
    },
    {
      "entropy": 6.45391526222229,
      "epoch": 0.6393537218695903,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004968888961307286,
      "loss": 6.299,
      "mean_token_accuracy": 0.16154887080192565,
      "num_tokens": 14048589.0,
      "step": 5540
    },
    {
      "entropy": 6.368120288848877,
      "epoch": 0.6399307559145989,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968820540772003,
      "loss": 6.1675,
      "mean_token_accuracy": 0.18676092177629472,
      "num_tokens": 14061007.0,
      "step": 5545
    },
    {
      "entropy": 6.459141635894776,
      "epoch": 0.6405077899596077,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004968752045607794,
      "loss": 6.3311,
      "mean_token_accuracy": 0.1602676048874855,
      "num_tokens": 14074165.0,
      "step": 5550
    },
    {
      "entropy": 6.524306535720825,
      "epoch": 0.6410848240046163,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004968683475816961,
      "loss": 6.3765,
      "mean_token_accuracy": 0.14993079304695128,
      "num_tokens": 14086261.0,
      "step": 5555
    },
    {
      "entropy": 6.448948669433594,
      "epoch": 0.641661858049625,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004968614831401811,
      "loss": 6.1853,
      "mean_token_accuracy": 0.17813284546136857,
      "num_tokens": 14098458.0,
      "step": 5560
    },
    {
      "entropy": 6.393662834167481,
      "epoch": 0.6422388920946336,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004968546112364654,
      "loss": 6.3727,
      "mean_token_accuracy": 0.15736806392669678,
      "num_tokens": 14110754.0,
      "step": 5565
    },
    {
      "entropy": 6.4795019149780275,
      "epoch": 0.6428159261396422,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049684773187078,
      "loss": 6.2741,
      "mean_token_accuracy": 0.15847027599811553,
      "num_tokens": 14123361.0,
      "step": 5570
    },
    {
      "entropy": 6.431928300857544,
      "epoch": 0.6433929601846509,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004968408450433565,
      "loss": 6.2101,
      "mean_token_accuracy": 0.17053861171007156,
      "num_tokens": 14135576.0,
      "step": 5575
    },
    {
      "entropy": 6.384128427505493,
      "epoch": 0.6439699942296595,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004968339507544263,
      "loss": 6.2859,
      "mean_token_accuracy": 0.16223161965608596,
      "num_tokens": 14147879.0,
      "step": 5580
    },
    {
      "entropy": 6.490727329254151,
      "epoch": 0.6445470282746683,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968270490042212,
      "loss": 6.3116,
      "mean_token_accuracy": 0.16167767196893693,
      "num_tokens": 14159449.0,
      "step": 5585
    },
    {
      "entropy": 6.390603494644165,
      "epoch": 0.6451240623196769,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968201397929737,
      "loss": 6.2011,
      "mean_token_accuracy": 0.16788838654756547,
      "num_tokens": 14172547.0,
      "step": 5590
    },
    {
      "entropy": 6.427271127700806,
      "epoch": 0.6457010963646855,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004968132231209158,
      "loss": 6.2684,
      "mean_token_accuracy": 0.16361706405878068,
      "num_tokens": 14184670.0,
      "step": 5595
    },
    {
      "entropy": 6.466847944259643,
      "epoch": 0.6462781304096942,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004968062989882803,
      "loss": 6.2391,
      "mean_token_accuracy": 0.1650494620203972,
      "num_tokens": 14198048.0,
      "step": 5600
    },
    {
      "entropy": 6.376773309707642,
      "epoch": 0.6468551644547028,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967993673953003,
      "loss": 6.1819,
      "mean_token_accuracy": 0.16672603487968446,
      "num_tokens": 14209470.0,
      "step": 5605
    },
    {
      "entropy": 6.39730978012085,
      "epoch": 0.6474321984997115,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967924283422087,
      "loss": 6.2128,
      "mean_token_accuracy": 0.16682232320308685,
      "num_tokens": 14221731.0,
      "step": 5610
    },
    {
      "entropy": 6.4741298198699955,
      "epoch": 0.6480092325447201,
      "grad_norm": 0.828125,
      "learning_rate": 0.000496785481829239,
      "loss": 6.2703,
      "mean_token_accuracy": 0.15549876540899277,
      "num_tokens": 14235125.0,
      "step": 5615
    },
    {
      "entropy": 6.406168842315674,
      "epoch": 0.6485862665897288,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004967785278566245,
      "loss": 6.2409,
      "mean_token_accuracy": 0.1698000580072403,
      "num_tokens": 14248941.0,
      "step": 5620
    },
    {
      "entropy": 6.398039102554321,
      "epoch": 0.6491633006347375,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967715664245997,
      "loss": 6.1679,
      "mean_token_accuracy": 0.16912340372800827,
      "num_tokens": 14260955.0,
      "step": 5625
    },
    {
      "entropy": 6.357013130187989,
      "epoch": 0.6497403346797461,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967645975333983,
      "loss": 6.1771,
      "mean_token_accuracy": 0.17226281762123108,
      "num_tokens": 14274123.0,
      "step": 5630
    },
    {
      "entropy": 6.461044502258301,
      "epoch": 0.6503173687247548,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967576211832548,
      "loss": 6.2356,
      "mean_token_accuracy": 0.17155105769634246,
      "num_tokens": 14286920.0,
      "step": 5635
    },
    {
      "entropy": 6.39765772819519,
      "epoch": 0.6508944027697634,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004967506373744039,
      "loss": 6.227,
      "mean_token_accuracy": 0.16639230251312256,
      "num_tokens": 14299297.0,
      "step": 5640
    },
    {
      "entropy": 6.431455039978028,
      "epoch": 0.651471436814772,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004967436461070805,
      "loss": 6.2354,
      "mean_token_accuracy": 0.16841167211532593,
      "num_tokens": 14312107.0,
      "step": 5645
    },
    {
      "entropy": 6.393067216873169,
      "epoch": 0.6520484708597807,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0004967366473815196,
      "loss": 6.2321,
      "mean_token_accuracy": 0.16371000111103057,
      "num_tokens": 14325777.0,
      "step": 5650
    },
    {
      "entropy": 6.42800498008728,
      "epoch": 0.6526255049047894,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004967296411979568,
      "loss": 6.2442,
      "mean_token_accuracy": 0.1633562907576561,
      "num_tokens": 14337617.0,
      "step": 5655
    },
    {
      "entropy": 6.373917818069458,
      "epoch": 0.6532025389497981,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967226275566275,
      "loss": 6.1783,
      "mean_token_accuracy": 0.1725485160946846,
      "num_tokens": 14349756.0,
      "step": 5660
    },
    {
      "entropy": 6.426472759246826,
      "epoch": 0.6537795729948067,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000496715606457768,
      "loss": 6.2842,
      "mean_token_accuracy": 0.16463049799203872,
      "num_tokens": 14363358.0,
      "step": 5665
    },
    {
      "entropy": 6.432434940338135,
      "epoch": 0.6543566070398154,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004967085779016142,
      "loss": 6.2488,
      "mean_token_accuracy": 0.16682310700416564,
      "num_tokens": 14375510.0,
      "step": 5670
    },
    {
      "entropy": 6.424771928787232,
      "epoch": 0.654933641084824,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967015418884022,
      "loss": 6.2429,
      "mean_token_accuracy": 0.1676553323864937,
      "num_tokens": 14389523.0,
      "step": 5675
    },
    {
      "entropy": 6.351522779464721,
      "epoch": 0.6555106751298326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004966944984183692,
      "loss": 6.2353,
      "mean_token_accuracy": 0.1677194431424141,
      "num_tokens": 14400457.0,
      "step": 5680
    },
    {
      "entropy": 6.313013029098511,
      "epoch": 0.6560877091748413,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004966874474917518,
      "loss": 6.1924,
      "mean_token_accuracy": 0.17274997234344483,
      "num_tokens": 14413088.0,
      "step": 5685
    },
    {
      "entropy": 6.482390022277832,
      "epoch": 0.65666474321985,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004966803891087873,
      "loss": 6.2338,
      "mean_token_accuracy": 0.1635465532541275,
      "num_tokens": 14425816.0,
      "step": 5690
    },
    {
      "entropy": 6.335096120834351,
      "epoch": 0.6572417772648587,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000496673323269713,
      "loss": 6.1253,
      "mean_token_accuracy": 0.17425674349069595,
      "num_tokens": 14438739.0,
      "step": 5695
    },
    {
      "entropy": 6.366029453277588,
      "epoch": 0.6578188113098673,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004966662499747666,
      "loss": 6.2736,
      "mean_token_accuracy": 0.16277661472558974,
      "num_tokens": 14451543.0,
      "step": 5700
    },
    {
      "entropy": 6.524085235595703,
      "epoch": 0.6583958453548759,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004966591692241859,
      "loss": 6.2894,
      "mean_token_accuracy": 0.1611143946647644,
      "num_tokens": 14464969.0,
      "step": 5705
    },
    {
      "entropy": 6.419612073898316,
      "epoch": 0.6589728793998846,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004966520810182092,
      "loss": 6.259,
      "mean_token_accuracy": 0.16098872125148772,
      "num_tokens": 14477622.0,
      "step": 5710
    },
    {
      "entropy": 6.460262775421143,
      "epoch": 0.6595499134448932,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966449853570749,
      "loss": 6.261,
      "mean_token_accuracy": 0.16220455318689347,
      "num_tokens": 14490306.0,
      "step": 5715
    },
    {
      "entropy": 6.342335033416748,
      "epoch": 0.6601269474899019,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966378822410215,
      "loss": 6.1179,
      "mean_token_accuracy": 0.18148637264966966,
      "num_tokens": 14501701.0,
      "step": 5720
    },
    {
      "entropy": 6.344096946716308,
      "epoch": 0.6607039815349106,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004966307716702881,
      "loss": 6.2304,
      "mean_token_accuracy": 0.16898464113473893,
      "num_tokens": 14515700.0,
      "step": 5725
    },
    {
      "entropy": 6.436907052993774,
      "epoch": 0.6612810155799193,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004966236536451138,
      "loss": 6.2337,
      "mean_token_accuracy": 0.1619272217154503,
      "num_tokens": 14528583.0,
      "step": 5730
    },
    {
      "entropy": 6.333746767044067,
      "epoch": 0.6618580496249279,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000496616528165738,
      "loss": 6.1764,
      "mean_token_accuracy": 0.17340188473463058,
      "num_tokens": 14541896.0,
      "step": 5735
    },
    {
      "entropy": 6.324051856994629,
      "epoch": 0.6624350836699365,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004966093952324003,
      "loss": 6.0955,
      "mean_token_accuracy": 0.1745638370513916,
      "num_tokens": 14556948.0,
      "step": 5740
    },
    {
      "entropy": 6.455702114105224,
      "epoch": 0.6630121177149452,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004966022548453406,
      "loss": 6.2724,
      "mean_token_accuracy": 0.16308293491601944,
      "num_tokens": 14568802.0,
      "step": 5745
    },
    {
      "entropy": 6.459200239181518,
      "epoch": 0.6635891517599538,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004965951070047993,
      "loss": 6.2145,
      "mean_token_accuracy": 0.16676018685102462,
      "num_tokens": 14582664.0,
      "step": 5750
    },
    {
      "entropy": 6.34871129989624,
      "epoch": 0.6641661858049625,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004965879517110166,
      "loss": 6.336,
      "mean_token_accuracy": 0.15799727141857148,
      "num_tokens": 14596805.0,
      "step": 5755
    },
    {
      "entropy": 6.385591745376587,
      "epoch": 0.6647432198499712,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965807889642333,
      "loss": 6.2485,
      "mean_token_accuracy": 0.16366442143917084,
      "num_tokens": 14610532.0,
      "step": 5760
    },
    {
      "entropy": 6.416805696487427,
      "epoch": 0.6653202538949798,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00049657361876469,
      "loss": 6.2606,
      "mean_token_accuracy": 0.1632024109363556,
      "num_tokens": 14622928.0,
      "step": 5765
    },
    {
      "entropy": 6.340898132324218,
      "epoch": 0.6658972879399885,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004965664411126282,
      "loss": 6.1348,
      "mean_token_accuracy": 0.17054639756679535,
      "num_tokens": 14634410.0,
      "step": 5770
    },
    {
      "entropy": 6.3054626941680905,
      "epoch": 0.6664743219849971,
      "grad_norm": 0.75,
      "learning_rate": 0.0004965592560082894,
      "loss": 6.1654,
      "mean_token_accuracy": 0.1657789871096611,
      "num_tokens": 14647571.0,
      "step": 5775
    },
    {
      "entropy": 6.328816318511963,
      "epoch": 0.6670513560300058,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004965520634519149,
      "loss": 6.0639,
      "mean_token_accuracy": 0.1815602406859398,
      "num_tokens": 14660656.0,
      "step": 5780
    },
    {
      "entropy": 6.43849925994873,
      "epoch": 0.6676283900750144,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004965448634437468,
      "loss": 6.2675,
      "mean_token_accuracy": 0.1647663116455078,
      "num_tokens": 14672267.0,
      "step": 5785
    },
    {
      "entropy": 6.400232458114624,
      "epoch": 0.668205424120023,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004965376559840272,
      "loss": 6.261,
      "mean_token_accuracy": 0.15839430093765258,
      "num_tokens": 14684590.0,
      "step": 5790
    },
    {
      "entropy": 6.415697956085205,
      "epoch": 0.6687824581650318,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004965304410729986,
      "loss": 6.2558,
      "mean_token_accuracy": 0.16338470578193665,
      "num_tokens": 14697299.0,
      "step": 5795
    },
    {
      "entropy": 6.384687328338623,
      "epoch": 0.6693594922100404,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004965232187109037,
      "loss": 6.237,
      "mean_token_accuracy": 0.16450579911470414,
      "num_tokens": 14709328.0,
      "step": 5800
    },
    {
      "entropy": 6.378979587554932,
      "epoch": 0.6699365262550491,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004965159888979854,
      "loss": 6.1939,
      "mean_token_accuracy": 0.16781411916017533,
      "num_tokens": 14721786.0,
      "step": 5805
    },
    {
      "entropy": 6.449314022064209,
      "epoch": 0.6705135603000577,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965087516344869,
      "loss": 6.2787,
      "mean_token_accuracy": 0.16612562984228135,
      "num_tokens": 14734120.0,
      "step": 5810
    },
    {
      "entropy": 6.398446130752563,
      "epoch": 0.6710905943450663,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004965015069206515,
      "loss": 6.2835,
      "mean_token_accuracy": 0.16125185340642928,
      "num_tokens": 14746159.0,
      "step": 5815
    },
    {
      "entropy": 6.455897188186645,
      "epoch": 0.671667628390075,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000496494254756723,
      "loss": 6.2094,
      "mean_token_accuracy": 0.1677599936723709,
      "num_tokens": 14758132.0,
      "step": 5820
    },
    {
      "entropy": 6.287702941894532,
      "epoch": 0.6722446624350836,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004964869951429451,
      "loss": 6.1087,
      "mean_token_accuracy": 0.17165531367063522,
      "num_tokens": 14770934.0,
      "step": 5825
    },
    {
      "entropy": 6.377002620697022,
      "epoch": 0.6728216964800924,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004964797280795622,
      "loss": 6.2579,
      "mean_token_accuracy": 0.16170469373464585,
      "num_tokens": 14785084.0,
      "step": 5830
    },
    {
      "entropy": 6.413485860824585,
      "epoch": 0.673398730525101,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004964724535668188,
      "loss": 6.2932,
      "mean_token_accuracy": 0.1723214104771614,
      "num_tokens": 14798553.0,
      "step": 5835
    },
    {
      "entropy": 6.42900242805481,
      "epoch": 0.6739757645701097,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004964651716049593,
      "loss": 6.261,
      "mean_token_accuracy": 0.1611432671546936,
      "num_tokens": 14810239.0,
      "step": 5840
    },
    {
      "entropy": 6.436808919906616,
      "epoch": 0.6745527986151183,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004964578821942288,
      "loss": 6.242,
      "mean_token_accuracy": 0.16524893343448638,
      "num_tokens": 14822704.0,
      "step": 5845
    },
    {
      "entropy": 6.405510377883911,
      "epoch": 0.6751298326601269,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004964505853348724,
      "loss": 6.2703,
      "mean_token_accuracy": 0.16530483216047287,
      "num_tokens": 14836913.0,
      "step": 5850
    },
    {
      "entropy": 6.313684940338135,
      "epoch": 0.6757068667051356,
      "grad_norm": 0.75,
      "learning_rate": 0.0004964432810271356,
      "loss": 6.2204,
      "mean_token_accuracy": 0.17138755917549134,
      "num_tokens": 14849697.0,
      "step": 5855
    },
    {
      "entropy": 6.353374385833741,
      "epoch": 0.6762839007501442,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004964359692712641,
      "loss": 6.1923,
      "mean_token_accuracy": 0.16545136272907257,
      "num_tokens": 14861719.0,
      "step": 5860
    },
    {
      "entropy": 6.432092809677124,
      "epoch": 0.676860934795153,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004964286500675037,
      "loss": 6.2078,
      "mean_token_accuracy": 0.1683959648013115,
      "num_tokens": 14873786.0,
      "step": 5865
    },
    {
      "entropy": 6.355663204193116,
      "epoch": 0.6774379688401616,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004964213234161007,
      "loss": 6.131,
      "mean_token_accuracy": 0.16872219145298004,
      "num_tokens": 14888952.0,
      "step": 5870
    },
    {
      "entropy": 6.42615270614624,
      "epoch": 0.6780150028851702,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004964139893173014,
      "loss": 6.3167,
      "mean_token_accuracy": 0.16259051263332366,
      "num_tokens": 14903039.0,
      "step": 5875
    },
    {
      "entropy": 6.402793931961059,
      "epoch": 0.6785920369301789,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964066477713525,
      "loss": 6.2453,
      "mean_token_accuracy": 0.1611890286207199,
      "num_tokens": 14915949.0,
      "step": 5880
    },
    {
      "entropy": 6.409963083267212,
      "epoch": 0.6791690709751875,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004963992987785011,
      "loss": 6.311,
      "mean_token_accuracy": 0.15956235229969024,
      "num_tokens": 14928724.0,
      "step": 5885
    },
    {
      "entropy": 6.457792806625366,
      "epoch": 0.6797461050201962,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004963919423389942,
      "loss": 6.2087,
      "mean_token_accuracy": 0.1660263478755951,
      "num_tokens": 14941759.0,
      "step": 5890
    },
    {
      "entropy": 6.358177757263183,
      "epoch": 0.6803231390652048,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004963845784530794,
      "loss": 6.2706,
      "mean_token_accuracy": 0.16284285485744476,
      "num_tokens": 14954373.0,
      "step": 5895
    },
    {
      "entropy": 6.369445943832398,
      "epoch": 0.6809001731102136,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000496377207121004,
      "loss": 6.1491,
      "mean_token_accuracy": 0.17449084520339966,
      "num_tokens": 14966195.0,
      "step": 5900
    },
    {
      "entropy": 6.380675411224365,
      "epoch": 0.6814772071552222,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004963698283430164,
      "loss": 6.2375,
      "mean_token_accuracy": 0.16495191305875778,
      "num_tokens": 14979594.0,
      "step": 5905
    },
    {
      "entropy": 6.4008690357208256,
      "epoch": 0.6820542412002308,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004963624421193646,
      "loss": 6.3011,
      "mean_token_accuracy": 0.16285253763198854,
      "num_tokens": 14993215.0,
      "step": 5910
    },
    {
      "entropy": 6.439036655426025,
      "epoch": 0.6826312752452395,
      "grad_norm": 0.796875,
      "learning_rate": 0.000496355048450297,
      "loss": 6.2691,
      "mean_token_accuracy": 0.1646467924118042,
      "num_tokens": 15006241.0,
      "step": 5915
    },
    {
      "entropy": 6.3777916431427,
      "epoch": 0.6832083092902481,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004963476473360623,
      "loss": 6.2391,
      "mean_token_accuracy": 0.16157302856445313,
      "num_tokens": 15018750.0,
      "step": 5920
    },
    {
      "entropy": 6.355571222305298,
      "epoch": 0.6837853433352568,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004963402387769094,
      "loss": 6.184,
      "mean_token_accuracy": 0.16624762564897538,
      "num_tokens": 15031065.0,
      "step": 5925
    },
    {
      "entropy": 6.42508020401001,
      "epoch": 0.6843623773802654,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004963328227730876,
      "loss": 6.1969,
      "mean_token_accuracy": 0.1655843064188957,
      "num_tokens": 15043129.0,
      "step": 5930
    },
    {
      "entropy": 6.383753156661987,
      "epoch": 0.684939411425274,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963253993248461,
      "loss": 6.2178,
      "mean_token_accuracy": 0.16666497588157653,
      "num_tokens": 15057531.0,
      "step": 5935
    },
    {
      "entropy": 6.392100429534912,
      "epoch": 0.6855164454702828,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004963179684324349,
      "loss": 6.265,
      "mean_token_accuracy": 0.16158232539892198,
      "num_tokens": 15070105.0,
      "step": 5940
    },
    {
      "entropy": 6.430133104324341,
      "epoch": 0.6860934795152914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963105300961036,
      "loss": 6.2466,
      "mean_token_accuracy": 0.15914711356163025,
      "num_tokens": 15082482.0,
      "step": 5945
    },
    {
      "entropy": 6.406900691986084,
      "epoch": 0.6866705135603001,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004963030843161026,
      "loss": 6.2706,
      "mean_token_accuracy": 0.16067442297935486,
      "num_tokens": 15095564.0,
      "step": 5950
    },
    {
      "entropy": 6.501013612747192,
      "epoch": 0.6872475476053087,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004962956310926823,
      "loss": 6.266,
      "mean_token_accuracy": 0.1601445868611336,
      "num_tokens": 15108948.0,
      "step": 5955
    },
    {
      "entropy": 6.394466733932495,
      "epoch": 0.6878245816503173,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962881704260934,
      "loss": 6.1838,
      "mean_token_accuracy": 0.16864433139562607,
      "num_tokens": 15121537.0,
      "step": 5960
    },
    {
      "entropy": 6.337121677398682,
      "epoch": 0.688401615695326,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004962807023165868,
      "loss": 6.1956,
      "mean_token_accuracy": 0.16982163339853287,
      "num_tokens": 15134657.0,
      "step": 5965
    },
    {
      "entropy": 6.385394525527954,
      "epoch": 0.6889786497403346,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004962732267644138,
      "loss": 6.1637,
      "mean_token_accuracy": 0.17017131447792053,
      "num_tokens": 15146301.0,
      "step": 5970
    },
    {
      "entropy": 6.342045068740845,
      "epoch": 0.6895556837853434,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004962657437698254,
      "loss": 6.2601,
      "mean_token_accuracy": 0.16192465722560884,
      "num_tokens": 15157886.0,
      "step": 5975
    },
    {
      "entropy": 6.442126131057739,
      "epoch": 0.690132717830352,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004962582533330739,
      "loss": 6.2173,
      "mean_token_accuracy": 0.16680624336004257,
      "num_tokens": 15170457.0,
      "step": 5980
    },
    {
      "entropy": 6.392907953262329,
      "epoch": 0.6907097518753607,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004962507554544109,
      "loss": 6.2573,
      "mean_token_accuracy": 0.16340704262256622,
      "num_tokens": 15182504.0,
      "step": 5985
    },
    {
      "entropy": 6.400666809082031,
      "epoch": 0.6912867859203693,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004962432501340886,
      "loss": 6.2494,
      "mean_token_accuracy": 0.16168738454580306,
      "num_tokens": 15194809.0,
      "step": 5990
    },
    {
      "entropy": 6.390184307098389,
      "epoch": 0.6918638199653779,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004962357373723596,
      "loss": 6.1953,
      "mean_token_accuracy": 0.16622450053691865,
      "num_tokens": 15207185.0,
      "step": 5995
    },
    {
      "entropy": 6.377176237106323,
      "epoch": 0.6924408540103866,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004962282171694763,
      "loss": 6.2028,
      "mean_token_accuracy": 0.16484107077121735,
      "num_tokens": 15220406.0,
      "step": 6000
    },
    {
      "epoch": 0.6924408540103866,
      "eval_entropy": 6.195665060948471,
      "eval_loss": 6.275278091430664,
      "eval_mean_token_accuracy": 0.1671429143809174,
      "eval_num_tokens": 15220406.0,
      "eval_runtime": 17.5527,
      "eval_samples_per_second": 1602.944,
      "eval_steps_per_second": 200.368,
      "step": 6000
    },
    {
      "entropy": 6.403058910369873,
      "epoch": 0.6930178880553952,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004962206895256919,
      "loss": 6.2785,
      "mean_token_accuracy": 0.1588844671845436,
      "num_tokens": 15231698.0,
      "step": 6005
    },
    {
      "entropy": 6.430747413635254,
      "epoch": 0.693594922100404,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004962131544412595,
      "loss": 6.2968,
      "mean_token_accuracy": 0.15798638761043549,
      "num_tokens": 15243462.0,
      "step": 6010
    },
    {
      "entropy": 6.525234031677246,
      "epoch": 0.6941719561454126,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004962056119164325,
      "loss": 6.303,
      "mean_token_accuracy": 0.15794008672237397,
      "num_tokens": 15257311.0,
      "step": 6015
    },
    {
      "entropy": 6.403410196304321,
      "epoch": 0.6947489901904212,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004961980619514646,
      "loss": 6.272,
      "mean_token_accuracy": 0.1627096489071846,
      "num_tokens": 15270186.0,
      "step": 6020
    },
    {
      "entropy": 6.362415742874146,
      "epoch": 0.6953260242354299,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004961905045466098,
      "loss": 6.2115,
      "mean_token_accuracy": 0.16145994067192077,
      "num_tokens": 15282573.0,
      "step": 6025
    },
    {
      "entropy": 6.415639114379883,
      "epoch": 0.6959030582804385,
      "grad_norm": 0.6875,
      "learning_rate": 0.0004961829397021222,
      "loss": 6.2569,
      "mean_token_accuracy": 0.16155208200216292,
      "num_tokens": 15296721.0,
      "step": 6030
    },
    {
      "entropy": 6.458365440368652,
      "epoch": 0.6964800923254472,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004961753674182564,
      "loss": 6.2384,
      "mean_token_accuracy": 0.16829407066106797,
      "num_tokens": 15308458.0,
      "step": 6035
    },
    {
      "entropy": 6.422654867172241,
      "epoch": 0.6970571263704558,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004961677876952669,
      "loss": 6.2437,
      "mean_token_accuracy": 0.1629546955227852,
      "num_tokens": 15320402.0,
      "step": 6040
    },
    {
      "entropy": 6.471780776977539,
      "epoch": 0.6976341604154646,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004961602005334087,
      "loss": 6.2758,
      "mean_token_accuracy": 0.16486609280109404,
      "num_tokens": 15333196.0,
      "step": 6045
    },
    {
      "entropy": 6.377042675018311,
      "epoch": 0.6982111944604732,
      "grad_norm": 0.796875,
      "learning_rate": 0.000496152605932937,
      "loss": 6.2178,
      "mean_token_accuracy": 0.16063844710588454,
      "num_tokens": 15345216.0,
      "step": 6050
    },
    {
      "entropy": 6.431419324874878,
      "epoch": 0.6987882285054818,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004961450038941074,
      "loss": 6.2718,
      "mean_token_accuracy": 0.16400076597929,
      "num_tokens": 15357772.0,
      "step": 6055
    },
    {
      "entropy": 6.3903600692749025,
      "epoch": 0.6993652625504905,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961373944171754,
      "loss": 6.1732,
      "mean_token_accuracy": 0.1675198808312416,
      "num_tokens": 15371110.0,
      "step": 6060
    },
    {
      "entropy": 6.323823022842407,
      "epoch": 0.6999422965954991,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004961297775023968,
      "loss": 6.1723,
      "mean_token_accuracy": 0.1711607024073601,
      "num_tokens": 15383731.0,
      "step": 6065
    },
    {
      "entropy": 6.457089138031006,
      "epoch": 0.7005193306405078,
      "grad_norm": 0.875,
      "learning_rate": 0.000496122153150028,
      "loss": 6.2283,
      "mean_token_accuracy": 0.16561086028814315,
      "num_tokens": 15395849.0,
      "step": 6070
    },
    {
      "entropy": 6.3939621925354,
      "epoch": 0.7010963646855164,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004961145213603255,
      "loss": 6.1742,
      "mean_token_accuracy": 0.1702093780040741,
      "num_tokens": 15408552.0,
      "step": 6075
    },
    {
      "entropy": 6.392090368270874,
      "epoch": 0.7016733987305251,
      "grad_norm": 1.34375,
      "learning_rate": 0.000496106882133546,
      "loss": 6.2467,
      "mean_token_accuracy": 0.16560420244932175,
      "num_tokens": 15421940.0,
      "step": 6080
    },
    {
      "entropy": 6.380345726013184,
      "epoch": 0.7022504327755338,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004960992354699463,
      "loss": 6.2168,
      "mean_token_accuracy": 0.1649238422513008,
      "num_tokens": 15434232.0,
      "step": 6085
    },
    {
      "entropy": 6.421299934387207,
      "epoch": 0.7028274668205424,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004960915813697836,
      "loss": 6.2208,
      "mean_token_accuracy": 0.1599582090973854,
      "num_tokens": 15448052.0,
      "step": 6090
    },
    {
      "entropy": 6.418045997619629,
      "epoch": 0.7034045008655511,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004960839198333154,
      "loss": 6.126,
      "mean_token_accuracy": 0.17226516604423522,
      "num_tokens": 15459141.0,
      "step": 6095
    },
    {
      "entropy": 6.295596790313721,
      "epoch": 0.7039815349105597,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004960762508607994,
      "loss": 6.0936,
      "mean_token_accuracy": 0.17302310317754746,
      "num_tokens": 15471705.0,
      "step": 6100
    },
    {
      "entropy": 6.355627346038818,
      "epoch": 0.7045585689555683,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004960685744524934,
      "loss": 6.1793,
      "mean_token_accuracy": 0.1686767742037773,
      "num_tokens": 15484821.0,
      "step": 6105
    },
    {
      "entropy": 6.456986331939698,
      "epoch": 0.705135603000577,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004960608906086558,
      "loss": 6.2245,
      "mean_token_accuracy": 0.1597846880555153,
      "num_tokens": 15497157.0,
      "step": 6110
    },
    {
      "entropy": 6.3890242099761965,
      "epoch": 0.7057126370455857,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496053199329545,
      "loss": 6.2268,
      "mean_token_accuracy": 0.16634002327919006,
      "num_tokens": 15508919.0,
      "step": 6115
    },
    {
      "entropy": 6.434191656112671,
      "epoch": 0.7062896710905944,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004960455006154196,
      "loss": 6.2611,
      "mean_token_accuracy": 0.1605444520711899,
      "num_tokens": 15521235.0,
      "step": 6120
    },
    {
      "entropy": 6.415744113922119,
      "epoch": 0.706866705135603,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004960377944665386,
      "loss": 6.2376,
      "mean_token_accuracy": 0.1591494560241699,
      "num_tokens": 15534278.0,
      "step": 6125
    },
    {
      "entropy": 6.323842239379883,
      "epoch": 0.7074437391806117,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004960300808831611,
      "loss": 6.1969,
      "mean_token_accuracy": 0.1674134060740471,
      "num_tokens": 15548034.0,
      "step": 6130
    },
    {
      "entropy": 6.413788986206055,
      "epoch": 0.7080207732256203,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004960223598655467,
      "loss": 6.173,
      "mean_token_accuracy": 0.1706375688314438,
      "num_tokens": 15559425.0,
      "step": 6135
    },
    {
      "entropy": 6.446317958831787,
      "epoch": 0.7085978072706289,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496014631413955,
      "loss": 6.2679,
      "mean_token_accuracy": 0.1664192110300064,
      "num_tokens": 15570634.0,
      "step": 6140
    },
    {
      "entropy": 6.40083646774292,
      "epoch": 0.7091748413156376,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004960068955286459,
      "loss": 6.297,
      "mean_token_accuracy": 0.16676503270864487,
      "num_tokens": 15582710.0,
      "step": 6145
    },
    {
      "entropy": 6.391730165481567,
      "epoch": 0.7097518753606463,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004959991522098797,
      "loss": 6.2297,
      "mean_token_accuracy": 0.1579965278506279,
      "num_tokens": 15594798.0,
      "step": 6150
    },
    {
      "entropy": 6.4401898860931395,
      "epoch": 0.710328909405655,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004959914014579168,
      "loss": 6.1574,
      "mean_token_accuracy": 0.16800634562969208,
      "num_tokens": 15606482.0,
      "step": 6155
    },
    {
      "entropy": 6.375606536865234,
      "epoch": 0.7109059434506636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004959836432730178,
      "loss": 6.1879,
      "mean_token_accuracy": 0.17139801681041716,
      "num_tokens": 15620178.0,
      "step": 6160
    },
    {
      "entropy": 6.335100984573364,
      "epoch": 0.7114829774956722,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959758776554438,
      "loss": 6.2204,
      "mean_token_accuracy": 0.16347949504852294,
      "num_tokens": 15632469.0,
      "step": 6165
    },
    {
      "entropy": 6.434891414642334,
      "epoch": 0.7120600115406809,
      "grad_norm": 0.74609375,
      "learning_rate": 0.000495968104605456,
      "loss": 6.2435,
      "mean_token_accuracy": 0.16189115941524507,
      "num_tokens": 15645610.0,
      "step": 6170
    },
    {
      "entropy": 6.436146354675293,
      "epoch": 0.7126370455856895,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004959603241233157,
      "loss": 6.2346,
      "mean_token_accuracy": 0.16635668277740479,
      "num_tokens": 15657716.0,
      "step": 6175
    },
    {
      "entropy": 6.370654678344726,
      "epoch": 0.7132140796306982,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004959525362092846,
      "loss": 6.2034,
      "mean_token_accuracy": 0.16088638752698897,
      "num_tokens": 15670258.0,
      "step": 6180
    },
    {
      "entropy": 6.361396217346192,
      "epoch": 0.7137911136757069,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959447408636247,
      "loss": 6.1679,
      "mean_token_accuracy": 0.16626855731010437,
      "num_tokens": 15681196.0,
      "step": 6185
    },
    {
      "entropy": 6.44526743888855,
      "epoch": 0.7143681477207156,
      "grad_norm": 0.75,
      "learning_rate": 0.0004959369380865983,
      "loss": 6.291,
      "mean_token_accuracy": 0.16206416636705398,
      "num_tokens": 15694591.0,
      "step": 6190
    },
    {
      "entropy": 6.351928186416626,
      "epoch": 0.7149451817657242,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004959291278784676,
      "loss": 6.1464,
      "mean_token_accuracy": 0.17528007328510284,
      "num_tokens": 15708272.0,
      "step": 6195
    },
    {
      "entropy": 6.40349383354187,
      "epoch": 0.7155222158107328,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004959213102394954,
      "loss": 6.1859,
      "mean_token_accuracy": 0.17357796430587769,
      "num_tokens": 15720136.0,
      "step": 6200
    },
    {
      "entropy": 6.444948053359985,
      "epoch": 0.7160992498557415,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004959134851699449,
      "loss": 6.1821,
      "mean_token_accuracy": 0.16729691177606582,
      "num_tokens": 15732043.0,
      "step": 6205
    },
    {
      "entropy": 6.395161962509155,
      "epoch": 0.7166762839007501,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004959056526700788,
      "loss": 6.2254,
      "mean_token_accuracy": 0.16458455175161363,
      "num_tokens": 15743471.0,
      "step": 6210
    },
    {
      "entropy": 6.377345561981201,
      "epoch": 0.7172533179457588,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004958978127401609,
      "loss": 6.2185,
      "mean_token_accuracy": 0.16235917508602143,
      "num_tokens": 15756846.0,
      "step": 6215
    },
    {
      "entropy": 6.357729434967041,
      "epoch": 0.7178303519907675,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004958899653804548,
      "loss": 6.256,
      "mean_token_accuracy": 0.16342784017324447,
      "num_tokens": 15769993.0,
      "step": 6220
    },
    {
      "entropy": 6.359043741226197,
      "epoch": 0.7184073860357761,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004958821105912246,
      "loss": 6.2114,
      "mean_token_accuracy": 0.1620841830968857,
      "num_tokens": 15782081.0,
      "step": 6225
    },
    {
      "entropy": 6.449085283279419,
      "epoch": 0.7189844200807848,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000495874248372734,
      "loss": 6.2394,
      "mean_token_accuracy": 0.16539080888032914,
      "num_tokens": 15793987.0,
      "step": 6230
    },
    {
      "entropy": 6.390167617797852,
      "epoch": 0.7195614541257934,
      "grad_norm": 0.84375,
      "learning_rate": 0.000495866378725248,
      "loss": 6.1704,
      "mean_token_accuracy": 0.17326804846525193,
      "num_tokens": 15805749.0,
      "step": 6235
    },
    {
      "entropy": 6.339812612533569,
      "epoch": 0.7201384881708021,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000495858501649031,
      "loss": 6.1432,
      "mean_token_accuracy": 0.1683582231402397,
      "num_tokens": 15818402.0,
      "step": 6240
    },
    {
      "entropy": 6.363752317428589,
      "epoch": 0.7207155222158107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495850617144348,
      "loss": 6.1067,
      "mean_token_accuracy": 0.17522320300340652,
      "num_tokens": 15830783.0,
      "step": 6245
    },
    {
      "entropy": 6.358516502380371,
      "epoch": 0.7212925562608193,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004958427252114643,
      "loss": 6.1179,
      "mean_token_accuracy": 0.1718319371342659,
      "num_tokens": 15843231.0,
      "step": 6250
    },
    {
      "entropy": 6.291909122467041,
      "epoch": 0.7218695903058281,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000495834825850645,
      "loss": 6.1503,
      "mean_token_accuracy": 0.17905625849962234,
      "num_tokens": 15855770.0,
      "step": 6255
    },
    {
      "entropy": 6.403748464584351,
      "epoch": 0.7224466243508367,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004958269190621562,
      "loss": 6.228,
      "mean_token_accuracy": 0.1606803447008133,
      "num_tokens": 15867979.0,
      "step": 6260
    },
    {
      "entropy": 6.387489128112793,
      "epoch": 0.7230236583958454,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958190048462637,
      "loss": 6.255,
      "mean_token_accuracy": 0.16454707086086273,
      "num_tokens": 15880878.0,
      "step": 6265
    },
    {
      "entropy": 6.3733234882354735,
      "epoch": 0.723600692440854,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004958110832032336,
      "loss": 6.1918,
      "mean_token_accuracy": 0.16965128034353255,
      "num_tokens": 15893599.0,
      "step": 6270
    },
    {
      "entropy": 6.30468053817749,
      "epoch": 0.7241777264858626,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004958031541333322,
      "loss": 6.1573,
      "mean_token_accuracy": 0.17099848687648772,
      "num_tokens": 15906566.0,
      "step": 6275
    },
    {
      "entropy": 6.423482370376587,
      "epoch": 0.7247547605308713,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004957952176368267,
      "loss": 6.1701,
      "mean_token_accuracy": 0.16515764892101287,
      "num_tokens": 15920269.0,
      "step": 6280
    },
    {
      "entropy": 6.375245380401611,
      "epoch": 0.7253317945758799,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004957872737139836,
      "loss": 6.2409,
      "mean_token_accuracy": 0.1627770632505417,
      "num_tokens": 15932559.0,
      "step": 6285
    },
    {
      "entropy": 6.362644052505493,
      "epoch": 0.7259088286208887,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004957793223650702,
      "loss": 6.1699,
      "mean_token_accuracy": 0.16183642596006392,
      "num_tokens": 15945554.0,
      "step": 6290
    },
    {
      "entropy": 6.330348300933838,
      "epoch": 0.7264858626658973,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000495771363590354,
      "loss": 6.0912,
      "mean_token_accuracy": 0.17449524998664856,
      "num_tokens": 15958012.0,
      "step": 6295
    },
    {
      "entropy": 6.395671749114991,
      "epoch": 0.727062896710906,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004957633973901027,
      "loss": 6.1958,
      "mean_token_accuracy": 0.164496810734272,
      "num_tokens": 15971508.0,
      "step": 6300
    },
    {
      "entropy": 6.3852269649505615,
      "epoch": 0.7276399307559146,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004957554237645841,
      "loss": 6.1695,
      "mean_token_accuracy": 0.16968413591384887,
      "num_tokens": 15983445.0,
      "step": 6305
    },
    {
      "entropy": 6.378427267074585,
      "epoch": 0.7282169648009232,
      "grad_norm": 0.875,
      "learning_rate": 0.0004957474427140665,
      "loss": 6.1899,
      "mean_token_accuracy": 0.16364562660455703,
      "num_tokens": 15995936.0,
      "step": 6310
    },
    {
      "entropy": 6.291472339630127,
      "epoch": 0.7287939988459319,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004957394542388182,
      "loss": 6.1542,
      "mean_token_accuracy": 0.1701158657670021,
      "num_tokens": 16008192.0,
      "step": 6315
    },
    {
      "entropy": 6.434455060958863,
      "epoch": 0.7293710328909405,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957314583391082,
      "loss": 6.2157,
      "mean_token_accuracy": 0.1633056089282036,
      "num_tokens": 16020585.0,
      "step": 6320
    },
    {
      "entropy": 6.461294221878052,
      "epoch": 0.7299480669359493,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004957234550152052,
      "loss": 6.2685,
      "mean_token_accuracy": 0.15828421115875244,
      "num_tokens": 16032627.0,
      "step": 6325
    },
    {
      "entropy": 6.370680522918701,
      "epoch": 0.7305251009809579,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0004957154442673784,
      "loss": 6.2104,
      "mean_token_accuracy": 0.17008405327796935,
      "num_tokens": 16047196.0,
      "step": 6330
    },
    {
      "entropy": 6.332737064361572,
      "epoch": 0.7311021350259665,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004957074260958972,
      "loss": 6.1513,
      "mean_token_accuracy": 0.16978222131729126,
      "num_tokens": 16059875.0,
      "step": 6335
    },
    {
      "entropy": 6.4214802265167235,
      "epoch": 0.7316791690709752,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004956994005010315,
      "loss": 6.2151,
      "mean_token_accuracy": 0.1705339267849922,
      "num_tokens": 16071991.0,
      "step": 6340
    },
    {
      "entropy": 6.416674041748047,
      "epoch": 0.7322562031159838,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000495691367483051,
      "loss": 6.2618,
      "mean_token_accuracy": 0.16010730117559432,
      "num_tokens": 16085628.0,
      "step": 6345
    },
    {
      "entropy": 6.388631391525268,
      "epoch": 0.7328332371609925,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004956833270422259,
      "loss": 6.2194,
      "mean_token_accuracy": 0.1635245516896248,
      "num_tokens": 16098020.0,
      "step": 6350
    },
    {
      "entropy": 6.381300592422486,
      "epoch": 0.7334102712060011,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004956752791788269,
      "loss": 6.1553,
      "mean_token_accuracy": 0.16594408452510834,
      "num_tokens": 16109965.0,
      "step": 6355
    },
    {
      "entropy": 6.398028421401977,
      "epoch": 0.7339873052510099,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0004956672238931244,
      "loss": 6.2429,
      "mean_token_accuracy": 0.1609364815056324,
      "num_tokens": 16121891.0,
      "step": 6360
    },
    {
      "entropy": 6.432336807250977,
      "epoch": 0.7345643392960185,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004956591611853894,
      "loss": 6.1661,
      "mean_token_accuracy": 0.16386929750442505,
      "num_tokens": 16134181.0,
      "step": 6365
    },
    {
      "entropy": 6.378114891052246,
      "epoch": 0.7351413733410271,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004956510910558931,
      "loss": 6.2315,
      "mean_token_accuracy": 0.16469560116529464,
      "num_tokens": 16146685.0,
      "step": 6370
    },
    {
      "entropy": 6.322031879425049,
      "epoch": 0.7357184073860358,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004956430135049071,
      "loss": 6.1949,
      "mean_token_accuracy": 0.16454170942306517,
      "num_tokens": 16159618.0,
      "step": 6375
    },
    {
      "entropy": 6.426507139205933,
      "epoch": 0.7362954414310444,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004956349285327028,
      "loss": 6.1158,
      "mean_token_accuracy": 0.17557175159454347,
      "num_tokens": 16172162.0,
      "step": 6380
    },
    {
      "entropy": 6.374025440216064,
      "epoch": 0.7368724754760531,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004956268361395523,
      "loss": 6.205,
      "mean_token_accuracy": 0.16355177909135818,
      "num_tokens": 16184663.0,
      "step": 6385
    },
    {
      "entropy": 6.3632265567779545,
      "epoch": 0.7374495095210617,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004956187363257278,
      "loss": 6.2099,
      "mean_token_accuracy": 0.1684990718960762,
      "num_tokens": 16196459.0,
      "step": 6390
    },
    {
      "entropy": 6.418331623077393,
      "epoch": 0.7380265435660704,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004956106290915017,
      "loss": 6.2045,
      "mean_token_accuracy": 0.16525492817163467,
      "num_tokens": 16208008.0,
      "step": 6395
    },
    {
      "entropy": 6.389336442947387,
      "epoch": 0.7386035776110791,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956025144371467,
      "loss": 6.2227,
      "mean_token_accuracy": 0.16817383021116256,
      "num_tokens": 16221844.0,
      "step": 6400
    },
    {
      "entropy": 6.368525552749634,
      "epoch": 0.7391806116560877,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004955943923629356,
      "loss": 6.2612,
      "mean_token_accuracy": 0.16519438475370407,
      "num_tokens": 16234773.0,
      "step": 6405
    },
    {
      "entropy": 6.312529706954956,
      "epoch": 0.7397576457010964,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004955862628691417,
      "loss": 6.1203,
      "mean_token_accuracy": 0.17223448902368546,
      "num_tokens": 16247464.0,
      "step": 6410
    },
    {
      "entropy": 6.459437131881714,
      "epoch": 0.740334679746105,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004955781259560386,
      "loss": 6.2155,
      "mean_token_accuracy": 0.156171815097332,
      "num_tokens": 16259947.0,
      "step": 6415
    },
    {
      "entropy": 6.429290008544922,
      "epoch": 0.7409117137911136,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004955699816238995,
      "loss": 6.3088,
      "mean_token_accuracy": 0.16252637803554534,
      "num_tokens": 16272280.0,
      "step": 6420
    },
    {
      "entropy": 6.419688081741333,
      "epoch": 0.7414887478361223,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004955618298729988,
      "loss": 6.2083,
      "mean_token_accuracy": 0.16577943414449692,
      "num_tokens": 16284476.0,
      "step": 6425
    },
    {
      "entropy": 6.339411067962646,
      "epoch": 0.742065781881131,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004955536707036105,
      "loss": 6.119,
      "mean_token_accuracy": 0.16848900616168977,
      "num_tokens": 16296877.0,
      "step": 6430
    },
    {
      "entropy": 6.258955383300782,
      "epoch": 0.7426428159261397,
      "grad_norm": 0.7421875,
      "learning_rate": 0.000495545504116009,
      "loss": 6.1236,
      "mean_token_accuracy": 0.17461392879486085,
      "num_tokens": 16310360.0,
      "step": 6435
    },
    {
      "entropy": 6.454283618927002,
      "epoch": 0.7432198499711483,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004955373301104691,
      "loss": 6.2774,
      "mean_token_accuracy": 0.15799676924943923,
      "num_tokens": 16322965.0,
      "step": 6440
    },
    {
      "entropy": 6.351188611984253,
      "epoch": 0.743796884016157,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0004955291486872657,
      "loss": 6.222,
      "mean_token_accuracy": 0.16237716376781464,
      "num_tokens": 16337756.0,
      "step": 6445
    },
    {
      "entropy": 6.378201866149903,
      "epoch": 0.7443739180611656,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004955209598466738,
      "loss": 6.1768,
      "mean_token_accuracy": 0.1628563940525055,
      "num_tokens": 16350400.0,
      "step": 6450
    },
    {
      "entropy": 6.4365991115570065,
      "epoch": 0.7449509521061742,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000495512763588969,
      "loss": 6.2824,
      "mean_token_accuracy": 0.159842848777771,
      "num_tokens": 16363655.0,
      "step": 6455
    },
    {
      "entropy": 6.422387981414795,
      "epoch": 0.7455279861511829,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004955045599144269,
      "loss": 6.1598,
      "mean_token_accuracy": 0.17184958904981612,
      "num_tokens": 16375580.0,
      "step": 6460
    },
    {
      "entropy": 6.367277050018311,
      "epoch": 0.7461050201961916,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004954963488233235,
      "loss": 6.2278,
      "mean_token_accuracy": 0.16555896550416946,
      "num_tokens": 16388447.0,
      "step": 6465
    },
    {
      "entropy": 6.379745817184448,
      "epoch": 0.7466820542412003,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000495488130315935,
      "loss": 6.1243,
      "mean_token_accuracy": 0.1724066451191902,
      "num_tokens": 16401082.0,
      "step": 6470
    },
    {
      "entropy": 6.290446996688843,
      "epoch": 0.7472590882862089,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004954799043925377,
      "loss": 6.1577,
      "mean_token_accuracy": 0.16472058445215226,
      "num_tokens": 16415108.0,
      "step": 6475
    },
    {
      "entropy": 6.3484659671783445,
      "epoch": 0.7478361223312175,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004954716710534082,
      "loss": 6.1003,
      "mean_token_accuracy": 0.1715170383453369,
      "num_tokens": 16427562.0,
      "step": 6480
    },
    {
      "entropy": 6.3726112842559814,
      "epoch": 0.7484131563762262,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004954634302988237,
      "loss": 6.1489,
      "mean_token_accuracy": 0.16994117796421052,
      "num_tokens": 16441282.0,
      "step": 6485
    },
    {
      "entropy": 6.354962253570557,
      "epoch": 0.7489901904212348,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004954551821290612,
      "loss": 6.1727,
      "mean_token_accuracy": 0.16542850881814958,
      "num_tokens": 16453433.0,
      "step": 6490
    },
    {
      "entropy": 6.381316232681274,
      "epoch": 0.7495672244662435,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004954469265443981,
      "loss": 6.202,
      "mean_token_accuracy": 0.16133319586515427,
      "num_tokens": 16466216.0,
      "step": 6495
    },
    {
      "entropy": 6.39223575592041,
      "epoch": 0.7501442585112522,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004954386635451123,
      "loss": 6.1951,
      "mean_token_accuracy": 0.16234314888715745,
      "num_tokens": 16479046.0,
      "step": 6500
    },
    {
      "entropy": 6.36065263748169,
      "epoch": 0.7507212925562609,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004954303931314814,
      "loss": 6.1467,
      "mean_token_accuracy": 0.16962222009897232,
      "num_tokens": 16491407.0,
      "step": 6505
    },
    {
      "entropy": 6.39806547164917,
      "epoch": 0.7512983266012695,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004954221153037837,
      "loss": 6.1613,
      "mean_token_accuracy": 0.16284280866384507,
      "num_tokens": 16504726.0,
      "step": 6510
    },
    {
      "entropy": 6.444783163070679,
      "epoch": 0.7518753606462781,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004954138300622978,
      "loss": 6.2385,
      "mean_token_accuracy": 0.1678844392299652,
      "num_tokens": 16517835.0,
      "step": 6515
    },
    {
      "entropy": 6.415285396575928,
      "epoch": 0.7524523946912868,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000495405537407302,
      "loss": 6.2218,
      "mean_token_accuracy": 0.16510699540376664,
      "num_tokens": 16532018.0,
      "step": 6520
    },
    {
      "entropy": 6.337543582916259,
      "epoch": 0.7530294287362954,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004953972373390755,
      "loss": 6.164,
      "mean_token_accuracy": 0.1710599184036255,
      "num_tokens": 16546211.0,
      "step": 6525
    },
    {
      "entropy": 6.411689567565918,
      "epoch": 0.753606462781304,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004953889298578975,
      "loss": 6.2248,
      "mean_token_accuracy": 0.1611586719751358,
      "num_tokens": 16559788.0,
      "step": 6530
    },
    {
      "entropy": 6.264769554138184,
      "epoch": 0.7541834968263128,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004953806149640473,
      "loss": 6.0833,
      "mean_token_accuracy": 0.17650727033615113,
      "num_tokens": 16573694.0,
      "step": 6535
    },
    {
      "entropy": 6.381378602981568,
      "epoch": 0.7547605308713214,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004953722926578045,
      "loss": 6.1806,
      "mean_token_accuracy": 0.16676169484853745,
      "num_tokens": 16585506.0,
      "step": 6540
    },
    {
      "entropy": 6.366420459747315,
      "epoch": 0.7553375649163301,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004953639629394492,
      "loss": 6.2281,
      "mean_token_accuracy": 0.16603047996759415,
      "num_tokens": 16598991.0,
      "step": 6545
    },
    {
      "entropy": 6.40582013130188,
      "epoch": 0.7559145989613387,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953556258092613,
      "loss": 6.2353,
      "mean_token_accuracy": 0.16572606861591338,
      "num_tokens": 16612025.0,
      "step": 6550
    },
    {
      "entropy": 6.349895334243774,
      "epoch": 0.7564916330063474,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953472812675216,
      "loss": 6.138,
      "mean_token_accuracy": 0.17004984319210054,
      "num_tokens": 16624509.0,
      "step": 6555
    },
    {
      "entropy": 6.345930051803589,
      "epoch": 0.757068667051356,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004953389293145104,
      "loss": 6.1894,
      "mean_token_accuracy": 0.1604577049612999,
      "num_tokens": 16636920.0,
      "step": 6560
    },
    {
      "entropy": 6.336128664016724,
      "epoch": 0.7576457010963646,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004953305699505088,
      "loss": 6.1022,
      "mean_token_accuracy": 0.1701395720243454,
      "num_tokens": 16649381.0,
      "step": 6565
    },
    {
      "entropy": 6.406035614013672,
      "epoch": 0.7582227351413734,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004953222031757979,
      "loss": 6.1962,
      "mean_token_accuracy": 0.16055445224046708,
      "num_tokens": 16662273.0,
      "step": 6570
    },
    {
      "entropy": 6.404620170593262,
      "epoch": 0.758799769186382,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004953138289906592,
      "loss": 6.2032,
      "mean_token_accuracy": 0.16861511170864105,
      "num_tokens": 16675184.0,
      "step": 6575
    },
    {
      "entropy": 6.359619283676148,
      "epoch": 0.7593768032313907,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004953054473953742,
      "loss": 6.1743,
      "mean_token_accuracy": 0.16798021495342255,
      "num_tokens": 16687658.0,
      "step": 6580
    },
    {
      "entropy": 6.389808940887451,
      "epoch": 0.7599538372763993,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004952970583902251,
      "loss": 6.1434,
      "mean_token_accuracy": 0.16392185240983964,
      "num_tokens": 16699806.0,
      "step": 6585
    },
    {
      "entropy": 6.365524530410767,
      "epoch": 0.760530871321408,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004952886619754937,
      "loss": 6.196,
      "mean_token_accuracy": 0.16961080133914946,
      "num_tokens": 16712647.0,
      "step": 6590
    },
    {
      "entropy": 6.395872735977173,
      "epoch": 0.7611079053664166,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952802581514625,
      "loss": 6.1513,
      "mean_token_accuracy": 0.1659959077835083,
      "num_tokens": 16725438.0,
      "step": 6595
    },
    {
      "entropy": 6.377916955947876,
      "epoch": 0.7616849394114252,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004952718469184144,
      "loss": 6.1703,
      "mean_token_accuracy": 0.17345594614744186,
      "num_tokens": 16737346.0,
      "step": 6600
    },
    {
      "entropy": 6.369835472106933,
      "epoch": 0.762261973456434,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004952634282766322,
      "loss": 6.1601,
      "mean_token_accuracy": 0.1693704679608345,
      "num_tokens": 16749327.0,
      "step": 6605
    },
    {
      "entropy": 6.362549734115601,
      "epoch": 0.7628390075014426,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004952550022263988,
      "loss": 6.1088,
      "mean_token_accuracy": 0.16926074177026748,
      "num_tokens": 16761323.0,
      "step": 6610
    },
    {
      "entropy": 6.36512598991394,
      "epoch": 0.7634160415464513,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004952465687679979,
      "loss": 6.2646,
      "mean_token_accuracy": 0.1608312338590622,
      "num_tokens": 16774688.0,
      "step": 6615
    },
    {
      "entropy": 6.478819036483765,
      "epoch": 0.7639930755914599,
      "grad_norm": 0.78125,
      "learning_rate": 0.000495238127901713,
      "loss": 6.2651,
      "mean_token_accuracy": 0.16290904581546783,
      "num_tokens": 16787538.0,
      "step": 6620
    },
    {
      "entropy": 6.294244432449341,
      "epoch": 0.7645701096364685,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004952296796278282,
      "loss": 6.158,
      "mean_token_accuracy": 0.16969640105962752,
      "num_tokens": 16798642.0,
      "step": 6625
    },
    {
      "entropy": 6.352237367630005,
      "epoch": 0.7651471436814772,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952212239466274,
      "loss": 6.2596,
      "mean_token_accuracy": 0.16831153631210327,
      "num_tokens": 16810785.0,
      "step": 6630
    },
    {
      "entropy": 6.454268121719361,
      "epoch": 0.7657241777264858,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004952127608583953,
      "loss": 6.2408,
      "mean_token_accuracy": 0.1620422899723053,
      "num_tokens": 16823978.0,
      "step": 6635
    },
    {
      "entropy": 6.3788317203521725,
      "epoch": 0.7663012117714946,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004952042903634162,
      "loss": 6.1759,
      "mean_token_accuracy": 0.174443618953228,
      "num_tokens": 16836878.0,
      "step": 6640
    },
    {
      "entropy": 6.378080177307129,
      "epoch": 0.7668782458165032,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004951958124619752,
      "loss": 6.2035,
      "mean_token_accuracy": 0.1723155453801155,
      "num_tokens": 16849849.0,
      "step": 6645
    },
    {
      "entropy": 6.396093320846558,
      "epoch": 0.7674552798615119,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004951873271543573,
      "loss": 6.2051,
      "mean_token_accuracy": 0.16407408565282822,
      "num_tokens": 16862022.0,
      "step": 6650
    },
    {
      "entropy": 6.466182136535645,
      "epoch": 0.7680323139065205,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004951788344408483,
      "loss": 6.2429,
      "mean_token_accuracy": 0.16660274267196656,
      "num_tokens": 16875421.0,
      "step": 6655
    },
    {
      "entropy": 6.3323774337768555,
      "epoch": 0.7686093479515291,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951703343217335,
      "loss": 6.1001,
      "mean_token_accuracy": 0.1834122806787491,
      "num_tokens": 16889520.0,
      "step": 6660
    },
    {
      "entropy": 6.2909516334533695,
      "epoch": 0.7691863819965378,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004951618267972987,
      "loss": 6.1519,
      "mean_token_accuracy": 0.16695373207330705,
      "num_tokens": 16901449.0,
      "step": 6665
    },
    {
      "entropy": 6.421184492111206,
      "epoch": 0.7697634160415464,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004951533118678304,
      "loss": 6.2154,
      "mean_token_accuracy": 0.1678440347313881,
      "num_tokens": 16914742.0,
      "step": 6670
    },
    {
      "entropy": 6.494894981384277,
      "epoch": 0.7703404500865552,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951447895336147,
      "loss": 6.1909,
      "mean_token_accuracy": 0.16298345923423768,
      "num_tokens": 16926645.0,
      "step": 6675
    },
    {
      "entropy": 6.336792993545532,
      "epoch": 0.7709174841315638,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004951362597949384,
      "loss": 6.1453,
      "mean_token_accuracy": 0.17599731534719468,
      "num_tokens": 16939499.0,
      "step": 6680
    },
    {
      "entropy": 6.3843560218811035,
      "epoch": 0.7714945181765724,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004951277226520883,
      "loss": 6.1272,
      "mean_token_accuracy": 0.17453108131885528,
      "num_tokens": 16951811.0,
      "step": 6685
    },
    {
      "entropy": 6.370804214477539,
      "epoch": 0.7720715522215811,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004951191781053517,
      "loss": 6.1671,
      "mean_token_accuracy": 0.1653602570295334,
      "num_tokens": 16963126.0,
      "step": 6690
    },
    {
      "entropy": 6.393895387649536,
      "epoch": 0.7726485862665897,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004951106261550157,
      "loss": 6.1778,
      "mean_token_accuracy": 0.16690278500318528,
      "num_tokens": 16976853.0,
      "step": 6695
    },
    {
      "entropy": 6.371247434616089,
      "epoch": 0.7732256203115984,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951020668013683,
      "loss": 6.2038,
      "mean_token_accuracy": 0.16455356776714325,
      "num_tokens": 16988720.0,
      "step": 6700
    },
    {
      "entropy": 6.360215902328491,
      "epoch": 0.773802654356607,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004950935000446972,
      "loss": 6.0983,
      "mean_token_accuracy": 0.18002919852733612,
      "num_tokens": 17000880.0,
      "step": 6705
    },
    {
      "entropy": 6.2722540378570555,
      "epoch": 0.7743796884016156,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950849258852905,
      "loss": 5.983,
      "mean_token_accuracy": 0.18147042840719224,
      "num_tokens": 17012540.0,
      "step": 6710
    },
    {
      "entropy": 6.3172304153442385,
      "epoch": 0.7749567224466244,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950763443234366,
      "loss": 6.1163,
      "mean_token_accuracy": 0.17636024802923203,
      "num_tokens": 17024163.0,
      "step": 6715
    },
    {
      "entropy": 6.399629688262939,
      "epoch": 0.775533756491633,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950677553594243,
      "loss": 6.1624,
      "mean_token_accuracy": 0.1638015702366829,
      "num_tokens": 17037020.0,
      "step": 6720
    },
    {
      "entropy": 6.411791276931763,
      "epoch": 0.7761107905366417,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004950591589935422,
      "loss": 6.2095,
      "mean_token_accuracy": 0.16194185465574265,
      "num_tokens": 17050132.0,
      "step": 6725
    },
    {
      "entropy": 6.335791063308716,
      "epoch": 0.7766878245816503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004950505552260798,
      "loss": 6.1433,
      "mean_token_accuracy": 0.167900712788105,
      "num_tokens": 17063299.0,
      "step": 6730
    },
    {
      "entropy": 6.321315670013428,
      "epoch": 0.777264858626659,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950419440573261,
      "loss": 6.1848,
      "mean_token_accuracy": 0.17039906084537507,
      "num_tokens": 17076312.0,
      "step": 6735
    },
    {
      "entropy": 6.3450603008270265,
      "epoch": 0.7778418926716676,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000495033325487571,
      "loss": 6.1671,
      "mean_token_accuracy": 0.16632827669382094,
      "num_tokens": 17089040.0,
      "step": 6740
    },
    {
      "entropy": 6.429263114929199,
      "epoch": 0.7784189267166762,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004950246995171042,
      "loss": 6.2492,
      "mean_token_accuracy": 0.16736022680997847,
      "num_tokens": 17102806.0,
      "step": 6745
    },
    {
      "entropy": 6.3069658279418945,
      "epoch": 0.778995960761685,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000495016066146216,
      "loss": 6.101,
      "mean_token_accuracy": 0.17719648629426957,
      "num_tokens": 17114782.0,
      "step": 6750
    },
    {
      "entropy": 6.3150811195373535,
      "epoch": 0.7795729948066936,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004950074253751968,
      "loss": 6.144,
      "mean_token_accuracy": 0.17379604130983353,
      "num_tokens": 17127386.0,
      "step": 6755
    },
    {
      "entropy": 6.338501644134522,
      "epoch": 0.7801500288517023,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494998777204337,
      "loss": 6.1319,
      "mean_token_accuracy": 0.17098195105791092,
      "num_tokens": 17139244.0,
      "step": 6760
    },
    {
      "entropy": 6.406604385375976,
      "epoch": 0.7807270628967109,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004949901216339276,
      "loss": 6.2446,
      "mean_token_accuracy": 0.16047115474939347,
      "num_tokens": 17153136.0,
      "step": 6765
    },
    {
      "entropy": 6.36369833946228,
      "epoch": 0.7813040969417195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949814586642598,
      "loss": 6.1184,
      "mean_token_accuracy": 0.16724678128957748,
      "num_tokens": 17166420.0,
      "step": 6770
    },
    {
      "entropy": 6.397933530807495,
      "epoch": 0.7818811309867282,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000494972788295625,
      "loss": 6.2494,
      "mean_token_accuracy": 0.15947156101465226,
      "num_tokens": 17180796.0,
      "step": 6775
    },
    {
      "entropy": 6.4296667098999025,
      "epoch": 0.7824581650317368,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004949641105283144,
      "loss": 6.2002,
      "mean_token_accuracy": 0.16173045486211776,
      "num_tokens": 17192275.0,
      "step": 6780
    },
    {
      "entropy": 6.336249876022339,
      "epoch": 0.7830351990767456,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004949554253626205,
      "loss": 6.1719,
      "mean_token_accuracy": 0.16753976345062255,
      "num_tokens": 17203651.0,
      "step": 6785
    },
    {
      "entropy": 6.441379642486572,
      "epoch": 0.7836122331217542,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004949467327988351,
      "loss": 6.2238,
      "mean_token_accuracy": 0.1699335977435112,
      "num_tokens": 17216521.0,
      "step": 6790
    },
    {
      "entropy": 6.350410509109497,
      "epoch": 0.7841892671667628,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004949380328372505,
      "loss": 6.1461,
      "mean_token_accuracy": 0.1731462612748146,
      "num_tokens": 17228435.0,
      "step": 6795
    },
    {
      "entropy": 6.339268684387207,
      "epoch": 0.7847663012117715,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949293254781595,
      "loss": 6.0854,
      "mean_token_accuracy": 0.17319487482309343,
      "num_tokens": 17240746.0,
      "step": 6800
    },
    {
      "entropy": 6.240341567993164,
      "epoch": 0.7853433352567801,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004949206107218547,
      "loss": 6.09,
      "mean_token_accuracy": 0.17679003179073333,
      "num_tokens": 17254089.0,
      "step": 6805
    },
    {
      "entropy": 6.3090087413787845,
      "epoch": 0.7859203693017888,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004949118885686296,
      "loss": 6.1233,
      "mean_token_accuracy": 0.17154750227928162,
      "num_tokens": 17266389.0,
      "step": 6810
    },
    {
      "entropy": 6.367681503295898,
      "epoch": 0.7864974033467974,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004949031590187774,
      "loss": 6.1424,
      "mean_token_accuracy": 0.16933028101921083,
      "num_tokens": 17279049.0,
      "step": 6815
    },
    {
      "entropy": 6.392622423171997,
      "epoch": 0.7870744373918062,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004948944220725915,
      "loss": 6.2099,
      "mean_token_accuracy": 0.1631815880537033,
      "num_tokens": 17291134.0,
      "step": 6820
    },
    {
      "entropy": 6.3588300228118895,
      "epoch": 0.7876514714368148,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000494885677730366,
      "loss": 6.0968,
      "mean_token_accuracy": 0.1776794210076332,
      "num_tokens": 17304728.0,
      "step": 6825
    },
    {
      "entropy": 6.338762283325195,
      "epoch": 0.7882285054818234,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494876925992395,
      "loss": 6.1723,
      "mean_token_accuracy": 0.16636593043804168,
      "num_tokens": 17317392.0,
      "step": 6830
    },
    {
      "entropy": 6.32216591835022,
      "epoch": 0.7888055395268321,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004948681668589728,
      "loss": 6.1356,
      "mean_token_accuracy": 0.17075138837099074,
      "num_tokens": 17329639.0,
      "step": 6835
    },
    {
      "entropy": 6.340901803970337,
      "epoch": 0.7893825735718407,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000494859400330394,
      "loss": 6.1639,
      "mean_token_accuracy": 0.16864123791456223,
      "num_tokens": 17341246.0,
      "step": 6840
    },
    {
      "entropy": 6.338127374649048,
      "epoch": 0.7899596076168494,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948506264069535,
      "loss": 6.164,
      "mean_token_accuracy": 0.17023713886737823,
      "num_tokens": 17353699.0,
      "step": 6845
    },
    {
      "entropy": 6.381475782394409,
      "epoch": 0.790536641661858,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004948418450889464,
      "loss": 6.1581,
      "mean_token_accuracy": 0.1735671043395996,
      "num_tokens": 17365950.0,
      "step": 6850
    },
    {
      "entropy": 6.287161064147949,
      "epoch": 0.7911136757068667,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000494833056376668,
      "loss": 6.086,
      "mean_token_accuracy": 0.16979992687702178,
      "num_tokens": 17378654.0,
      "step": 6855
    },
    {
      "entropy": 6.341888761520385,
      "epoch": 0.7916907097518754,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004948242602704139,
      "loss": 6.1709,
      "mean_token_accuracy": 0.1684945911169052,
      "num_tokens": 17391354.0,
      "step": 6860
    },
    {
      "entropy": 6.368467950820923,
      "epoch": 0.792267743796884,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004948154567704801,
      "loss": 6.165,
      "mean_token_accuracy": 0.17008308619260787,
      "num_tokens": 17403019.0,
      "step": 6865
    },
    {
      "entropy": 6.347121620178223,
      "epoch": 0.7928447778418927,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004948066458771625,
      "loss": 6.1565,
      "mean_token_accuracy": 0.1705833375453949,
      "num_tokens": 17415886.0,
      "step": 6870
    },
    {
      "entropy": 6.417398500442505,
      "epoch": 0.7934218118869013,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004947978275907577,
      "loss": 6.2731,
      "mean_token_accuracy": 0.1570700705051422,
      "num_tokens": 17429734.0,
      "step": 6875
    },
    {
      "entropy": 6.391656827926636,
      "epoch": 0.79399884593191,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000494789001911562,
      "loss": 6.0983,
      "mean_token_accuracy": 0.170063416659832,
      "num_tokens": 17441749.0,
      "step": 6880
    },
    {
      "entropy": 6.391465759277343,
      "epoch": 0.7945758799769186,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004947801688398725,
      "loss": 6.197,
      "mean_token_accuracy": 0.16510217785835266,
      "num_tokens": 17453626.0,
      "step": 6885
    },
    {
      "entropy": 6.341178512573242,
      "epoch": 0.7951529140219273,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004947713283759862,
      "loss": 6.1523,
      "mean_token_accuracy": 0.17330178022384643,
      "num_tokens": 17465038.0,
      "step": 6890
    },
    {
      "entropy": 6.325580787658692,
      "epoch": 0.795729948066936,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004947624805202003,
      "loss": 6.2253,
      "mean_token_accuracy": 0.15935006588697434,
      "num_tokens": 17478319.0,
      "step": 6895
    },
    {
      "entropy": 6.380958700180054,
      "epoch": 0.7963069821119446,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004947536252728126,
      "loss": 6.0879,
      "mean_token_accuracy": 0.1703493520617485,
      "num_tokens": 17491534.0,
      "step": 6900
    },
    {
      "entropy": 6.267689371109009,
      "epoch": 0.7968840161569533,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947447626341209,
      "loss": 6.1157,
      "mean_token_accuracy": 0.17074696868658065,
      "num_tokens": 17504957.0,
      "step": 6905
    },
    {
      "entropy": 6.384418296813965,
      "epoch": 0.7974610502019619,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004947358926044232,
      "loss": 6.1401,
      "mean_token_accuracy": 0.1719054564833641,
      "num_tokens": 17518265.0,
      "step": 6910
    },
    {
      "entropy": 6.324901294708252,
      "epoch": 0.7980380842469705,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004947270151840179,
      "loss": 6.108,
      "mean_token_accuracy": 0.17972162514925002,
      "num_tokens": 17530588.0,
      "step": 6915
    },
    {
      "entropy": 6.2338704586029055,
      "epoch": 0.7986151182919792,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004947181303732038,
      "loss": 6.0626,
      "mean_token_accuracy": 0.1753245547413826,
      "num_tokens": 17545429.0,
      "step": 6920
    },
    {
      "entropy": 6.397284126281738,
      "epoch": 0.7991921523369879,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004947092381722793,
      "loss": 6.1757,
      "mean_token_accuracy": 0.1684811607003212,
      "num_tokens": 17558437.0,
      "step": 6925
    },
    {
      "entropy": 6.346995830535889,
      "epoch": 0.7997691863819966,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004947003385815438,
      "loss": 6.1228,
      "mean_token_accuracy": 0.16606940180063248,
      "num_tokens": 17569863.0,
      "step": 6930
    },
    {
      "entropy": 6.3710277557373045,
      "epoch": 0.8003462204270052,
      "grad_norm": 0.875,
      "learning_rate": 0.0004946914316012964,
      "loss": 6.1889,
      "mean_token_accuracy": 0.16496853083372115,
      "num_tokens": 17581987.0,
      "step": 6935
    },
    {
      "entropy": 6.331367158889771,
      "epoch": 0.8009232544720138,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000494682517231837,
      "loss": 6.2048,
      "mean_token_accuracy": 0.16396823078393935,
      "num_tokens": 17594064.0,
      "step": 6940
    },
    {
      "entropy": 6.354979991912842,
      "epoch": 0.8015002885170225,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004946735954734652,
      "loss": 6.0929,
      "mean_token_accuracy": 0.1688402473926544,
      "num_tokens": 17606557.0,
      "step": 6945
    },
    {
      "entropy": 6.374159669876098,
      "epoch": 0.8020773225620311,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004946646663264811,
      "loss": 6.1963,
      "mean_token_accuracy": 0.1712319478392601,
      "num_tokens": 17620095.0,
      "step": 6950
    },
    {
      "entropy": 6.318175411224365,
      "epoch": 0.8026543566070398,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946557297911852,
      "loss": 6.0429,
      "mean_token_accuracy": 0.17838633954524993,
      "num_tokens": 17633006.0,
      "step": 6955
    },
    {
      "entropy": 6.3448278427124025,
      "epoch": 0.8032313906520485,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004946467858678777,
      "loss": 6.1556,
      "mean_token_accuracy": 0.1669870063662529,
      "num_tokens": 17646863.0,
      "step": 6960
    },
    {
      "entropy": 6.404659128189087,
      "epoch": 0.8038084246970572,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004946378345568597,
      "loss": 6.1515,
      "mean_token_accuracy": 0.16299126744270326,
      "num_tokens": 17659917.0,
      "step": 6965
    },
    {
      "entropy": 6.419924354553222,
      "epoch": 0.8043854587420658,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004946288758584324,
      "loss": 6.1741,
      "mean_token_accuracy": 0.16308792680501938,
      "num_tokens": 17672867.0,
      "step": 6970
    },
    {
      "entropy": 6.398059606552124,
      "epoch": 0.8049624927870744,
      "grad_norm": 0.875,
      "learning_rate": 0.0004946199097728968,
      "loss": 6.1772,
      "mean_token_accuracy": 0.1667613223195076,
      "num_tokens": 17684600.0,
      "step": 6975
    },
    {
      "entropy": 6.348398303985595,
      "epoch": 0.8055395268320831,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004946109363005548,
      "loss": 6.1216,
      "mean_token_accuracy": 0.1718461260199547,
      "num_tokens": 17697023.0,
      "step": 6980
    },
    {
      "entropy": 6.3436507225036625,
      "epoch": 0.8061165608770917,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000494601955441708,
      "loss": 6.1669,
      "mean_token_accuracy": 0.1690715506672859,
      "num_tokens": 17711592.0,
      "step": 6985
    },
    {
      "entropy": 6.451710510253906,
      "epoch": 0.8066935949221004,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945929671966585,
      "loss": 6.174,
      "mean_token_accuracy": 0.1678098663687706,
      "num_tokens": 17723039.0,
      "step": 6990
    },
    {
      "entropy": 6.315439033508301,
      "epoch": 0.8072706289671091,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004945839715657085,
      "loss": 6.1225,
      "mean_token_accuracy": 0.170714071393013,
      "num_tokens": 17736851.0,
      "step": 6995
    },
    {
      "entropy": 6.405031394958496,
      "epoch": 0.8078476630121177,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004945749685491607,
      "loss": 6.1644,
      "mean_token_accuracy": 0.16586832851171493,
      "num_tokens": 17749483.0,
      "step": 7000
    },
    {
      "entropy": 6.362979173660278,
      "epoch": 0.8084246970571264,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004945659581473181,
      "loss": 6.1892,
      "mean_token_accuracy": 0.16127808690071105,
      "num_tokens": 17760463.0,
      "step": 7005
    },
    {
      "entropy": 6.423250293731689,
      "epoch": 0.809001731102135,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004945569403604833,
      "loss": 6.246,
      "mean_token_accuracy": 0.16158504486083985,
      "num_tokens": 17772557.0,
      "step": 7010
    },
    {
      "entropy": 6.3385059356689455,
      "epoch": 0.8095787651471437,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004945479151889601,
      "loss": 6.1545,
      "mean_token_accuracy": 0.17329340428113937,
      "num_tokens": 17785399.0,
      "step": 7015
    },
    {
      "entropy": 6.414361000061035,
      "epoch": 0.8101557991921523,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004945388826330517,
      "loss": 6.1657,
      "mean_token_accuracy": 0.16833894997835158,
      "num_tokens": 17798766.0,
      "step": 7020
    },
    {
      "entropy": 6.381486511230468,
      "epoch": 0.8107328332371609,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004945298426930621,
      "loss": 6.1677,
      "mean_token_accuracy": 0.1663966253399849,
      "num_tokens": 17811726.0,
      "step": 7025
    },
    {
      "entropy": 6.321613836288452,
      "epoch": 0.8113098672821697,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004945207953692952,
      "loss": 6.1089,
      "mean_token_accuracy": 0.17208786457777023,
      "num_tokens": 17824142.0,
      "step": 7030
    },
    {
      "entropy": 6.33484959602356,
      "epoch": 0.8118869013271783,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004945117406620556,
      "loss": 6.1267,
      "mean_token_accuracy": 0.1715193048119545,
      "num_tokens": 17837497.0,
      "step": 7035
    },
    {
      "entropy": 6.343142461776734,
      "epoch": 0.812463935372187,
      "grad_norm": 0.875,
      "learning_rate": 0.0004945026785716474,
      "loss": 6.1536,
      "mean_token_accuracy": 0.1694921672344208,
      "num_tokens": 17849186.0,
      "step": 7040
    },
    {
      "entropy": 6.358126306533814,
      "epoch": 0.8130409694171956,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944936090983757,
      "loss": 6.1906,
      "mean_token_accuracy": 0.16527757346630095,
      "num_tokens": 17861460.0,
      "step": 7045
    },
    {
      "entropy": 6.373239183425904,
      "epoch": 0.8136180034622043,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944845322425455,
      "loss": 6.1034,
      "mean_token_accuracy": 0.17605217546224594,
      "num_tokens": 17874031.0,
      "step": 7050
    },
    {
      "entropy": 6.256667566299439,
      "epoch": 0.8141950375072129,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004944754480044621,
      "loss": 6.136,
      "mean_token_accuracy": 0.16608326733112336,
      "num_tokens": 17886623.0,
      "step": 7055
    },
    {
      "entropy": 6.373006582260132,
      "epoch": 0.8147720715522215,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944663563844311,
      "loss": 6.1237,
      "mean_token_accuracy": 0.16468188315629959,
      "num_tokens": 17899888.0,
      "step": 7060
    },
    {
      "entropy": 6.342667818069458,
      "epoch": 0.8153491055972303,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004944572573827581,
      "loss": 6.1519,
      "mean_token_accuracy": 0.170205919444561,
      "num_tokens": 17912089.0,
      "step": 7065
    },
    {
      "entropy": 6.369680547714234,
      "epoch": 0.8159261396422389,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004944481509997494,
      "loss": 6.2281,
      "mean_token_accuracy": 0.16549010276794435,
      "num_tokens": 17923790.0,
      "step": 7070
    },
    {
      "entropy": 6.407808256149292,
      "epoch": 0.8165031736872476,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000494439037235711,
      "loss": 6.1421,
      "mean_token_accuracy": 0.17216986715793609,
      "num_tokens": 17936537.0,
      "step": 7075
    },
    {
      "entropy": 6.317977619171143,
      "epoch": 0.8170802077322562,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004944299160909495,
      "loss": 6.1287,
      "mean_token_accuracy": 0.16496385782957076,
      "num_tokens": 17948230.0,
      "step": 7080
    },
    {
      "entropy": 6.360863304138183,
      "epoch": 0.8176572417772648,
      "grad_norm": 0.796875,
      "learning_rate": 0.000494420787565772,
      "loss": 6.115,
      "mean_token_accuracy": 0.17125698626041413,
      "num_tokens": 17960927.0,
      "step": 7085
    },
    {
      "entropy": 6.301209497451782,
      "epoch": 0.8182342758222735,
      "grad_norm": 0.703125,
      "learning_rate": 0.0004944116516604852,
      "loss": 6.0823,
      "mean_token_accuracy": 0.17663372606039046,
      "num_tokens": 17974392.0,
      "step": 7090
    },
    {
      "entropy": 6.268465423583985,
      "epoch": 0.8188113098672821,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004944025083753965,
      "loss": 6.0342,
      "mean_token_accuracy": 0.17979125529527665,
      "num_tokens": 17987845.0,
      "step": 7095
    },
    {
      "entropy": 6.347061634063721,
      "epoch": 0.8193883439122909,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004943933577108133,
      "loss": 6.1832,
      "mean_token_accuracy": 0.16562148183584213,
      "num_tokens": 18000610.0,
      "step": 7100
    },
    {
      "entropy": 6.299385595321655,
      "epoch": 0.8199653779572995,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004943841996670436,
      "loss": 6.0983,
      "mean_token_accuracy": 0.16856419742107392,
      "num_tokens": 18014322.0,
      "step": 7105
    },
    {
      "entropy": 6.354523181915283,
      "epoch": 0.8205424120023082,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004943750342443953,
      "loss": 6.1443,
      "mean_token_accuracy": 0.16706048399209977,
      "num_tokens": 18028072.0,
      "step": 7110
    },
    {
      "entropy": 6.375401163101197,
      "epoch": 0.8211194460473168,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943658614431767,
      "loss": 6.123,
      "mean_token_accuracy": 0.167181858420372,
      "num_tokens": 18039629.0,
      "step": 7115
    },
    {
      "entropy": 6.4030773639678955,
      "epoch": 0.8216964800923254,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004943566812636963,
      "loss": 6.1518,
      "mean_token_accuracy": 0.17692724764347076,
      "num_tokens": 18052879.0,
      "step": 7120
    },
    {
      "entropy": 6.340735054016113,
      "epoch": 0.8222735141373341,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494347493706263,
      "loss": 6.1698,
      "mean_token_accuracy": 0.17013720273971558,
      "num_tokens": 18065037.0,
      "step": 7125
    },
    {
      "entropy": 6.326661062240601,
      "epoch": 0.8228505481823427,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004943382987711856,
      "loss": 6.1085,
      "mean_token_accuracy": 0.17652219235897065,
      "num_tokens": 18078595.0,
      "step": 7130
    },
    {
      "entropy": 6.234781694412232,
      "epoch": 0.8234275822273515,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943290964587734,
      "loss": 5.9711,
      "mean_token_accuracy": 0.18064039647579194,
      "num_tokens": 18089790.0,
      "step": 7135
    },
    {
      "entropy": 6.222256422042847,
      "epoch": 0.8240046162723601,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004943198867693361,
      "loss": 6.0561,
      "mean_token_accuracy": 0.17768249213695525,
      "num_tokens": 18101343.0,
      "step": 7140
    },
    {
      "entropy": 6.3912993431091305,
      "epoch": 0.8245816503173687,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004943106697031833,
      "loss": 6.2047,
      "mean_token_accuracy": 0.16716157495975495,
      "num_tokens": 18113124.0,
      "step": 7145
    },
    {
      "entropy": 6.372874927520752,
      "epoch": 0.8251586843623774,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943014452606251,
      "loss": 6.1589,
      "mean_token_accuracy": 0.171916726231575,
      "num_tokens": 18125322.0,
      "step": 7150
    },
    {
      "entropy": 6.3357312202453615,
      "epoch": 0.825735718407386,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942922134419717,
      "loss": 6.2166,
      "mean_token_accuracy": 0.16244128495454788,
      "num_tokens": 18137975.0,
      "step": 7155
    },
    {
      "entropy": 6.388973569869995,
      "epoch": 0.8263127524523947,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004942829742475336,
      "loss": 6.1551,
      "mean_token_accuracy": 0.16712315678596495,
      "num_tokens": 18151149.0,
      "step": 7160
    },
    {
      "entropy": 6.392431020736694,
      "epoch": 0.8268897864974033,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004942737276776217,
      "loss": 6.1385,
      "mean_token_accuracy": 0.16350337564945222,
      "num_tokens": 18164451.0,
      "step": 7165
    },
    {
      "entropy": 6.3157727241516115,
      "epoch": 0.827466820542412,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004942644737325468,
      "loss": 6.109,
      "mean_token_accuracy": 0.16836577653884888,
      "num_tokens": 18177148.0,
      "step": 7170
    },
    {
      "entropy": 6.355120515823364,
      "epoch": 0.8280438545874207,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004942552124126202,
      "loss": 6.1359,
      "mean_token_accuracy": 0.16590700298547745,
      "num_tokens": 18189428.0,
      "step": 7175
    },
    {
      "entropy": 6.336666011810303,
      "epoch": 0.8286208886324293,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004942459437181535,
      "loss": 6.1086,
      "mean_token_accuracy": 0.16857766807079316,
      "num_tokens": 18201644.0,
      "step": 7180
    },
    {
      "entropy": 6.325605726242065,
      "epoch": 0.829197922677438,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942366676494584,
      "loss": 6.1404,
      "mean_token_accuracy": 0.16688464730978012,
      "num_tokens": 18213366.0,
      "step": 7185
    },
    {
      "entropy": 6.312024021148682,
      "epoch": 0.8297749567224466,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004942273842068469,
      "loss": 6.1466,
      "mean_token_accuracy": 0.17542206794023513,
      "num_tokens": 18226767.0,
      "step": 7190
    },
    {
      "entropy": 6.321409416198731,
      "epoch": 0.8303519907674553,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004942180933906311,
      "loss": 6.0851,
      "mean_token_accuracy": 0.17369858920574188,
      "num_tokens": 18239250.0,
      "step": 7195
    },
    {
      "entropy": 6.259433269500732,
      "epoch": 0.8309290248124639,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004942087952011237,
      "loss": 6.1092,
      "mean_token_accuracy": 0.17164007276296617,
      "num_tokens": 18252053.0,
      "step": 7200
    },
    {
      "entropy": 6.329160928726196,
      "epoch": 0.8315060588574726,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941994896386374,
      "loss": 6.1618,
      "mean_token_accuracy": 0.16128275990486146,
      "num_tokens": 18264769.0,
      "step": 7205
    },
    {
      "entropy": 6.380938577651977,
      "epoch": 0.8320830929024813,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004941901767034851,
      "loss": 6.1878,
      "mean_token_accuracy": 0.16496667116880417,
      "num_tokens": 18278185.0,
      "step": 7210
    },
    {
      "entropy": 6.306327247619629,
      "epoch": 0.8326601269474899,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004941808563959802,
      "loss": 6.0909,
      "mean_token_accuracy": 0.1723773032426834,
      "num_tokens": 18290997.0,
      "step": 7215
    },
    {
      "entropy": 6.271544933319092,
      "epoch": 0.8332371609924986,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004941715287164359,
      "loss": 6.0676,
      "mean_token_accuracy": 0.17277420908212662,
      "num_tokens": 18303735.0,
      "step": 7220
    },
    {
      "entropy": 6.245544052124023,
      "epoch": 0.8338141950375072,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004941621936651661,
      "loss": 6.0509,
      "mean_token_accuracy": 0.17647685110569,
      "num_tokens": 18316061.0,
      "step": 7225
    },
    {
      "entropy": 6.330385017395019,
      "epoch": 0.8343912290825158,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941528512424849,
      "loss": 6.0833,
      "mean_token_accuracy": 0.17309328615665437,
      "num_tokens": 18329478.0,
      "step": 7230
    },
    {
      "entropy": 6.333926105499268,
      "epoch": 0.8349682631275245,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004941435014487064,
      "loss": 6.0801,
      "mean_token_accuracy": 0.1704792007803917,
      "num_tokens": 18341790.0,
      "step": 7235
    },
    {
      "entropy": 6.29333872795105,
      "epoch": 0.8355452971725332,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494134144284145,
      "loss": 6.0634,
      "mean_token_accuracy": 0.17730342000722885,
      "num_tokens": 18353807.0,
      "step": 7240
    },
    {
      "entropy": 6.318775796890259,
      "epoch": 0.8361223312175419,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004941247797491156,
      "loss": 6.1353,
      "mean_token_accuracy": 0.16538234949111938,
      "num_tokens": 18366118.0,
      "step": 7245
    },
    {
      "entropy": 6.28317813873291,
      "epoch": 0.8366993652625505,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004941154078439329,
      "loss": 6.1052,
      "mean_token_accuracy": 0.17743841409683228,
      "num_tokens": 18380542.0,
      "step": 7250
    },
    {
      "entropy": 6.392764854431152,
      "epoch": 0.8372763993075591,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004941060285689126,
      "loss": 6.1693,
      "mean_token_accuracy": 0.1679488956928253,
      "num_tokens": 18393976.0,
      "step": 7255
    },
    {
      "entropy": 6.288981533050537,
      "epoch": 0.8378534333525678,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004940966419243695,
      "loss": 6.1261,
      "mean_token_accuracy": 0.16967616230249405,
      "num_tokens": 18405770.0,
      "step": 7260
    },
    {
      "entropy": 6.352445316314697,
      "epoch": 0.8384304673975764,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049408724791062,
      "loss": 6.188,
      "mean_token_accuracy": 0.1693796083331108,
      "num_tokens": 18417980.0,
      "step": 7265
    },
    {
      "entropy": 6.406365442276001,
      "epoch": 0.8390075014425851,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004940778465279797,
      "loss": 6.0911,
      "mean_token_accuracy": 0.17416509985923767,
      "num_tokens": 18429511.0,
      "step": 7270
    },
    {
      "entropy": 6.3113720417022705,
      "epoch": 0.8395845354875938,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940684377767647,
      "loss": 6.1427,
      "mean_token_accuracy": 0.16878841817378998,
      "num_tokens": 18443176.0,
      "step": 7275
    },
    {
      "entropy": 6.35668978691101,
      "epoch": 0.8401615695326025,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940590216572916,
      "loss": 6.1093,
      "mean_token_accuracy": 0.17457308620214462,
      "num_tokens": 18454664.0,
      "step": 7280
    },
    {
      "entropy": 6.311358499526977,
      "epoch": 0.8407386035776111,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004940495981698772,
      "loss": 6.0904,
      "mean_token_accuracy": 0.1729792758822441,
      "num_tokens": 18465919.0,
      "step": 7285
    },
    {
      "entropy": 6.289607524871826,
      "epoch": 0.8413156376226197,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940401673148384,
      "loss": 6.0265,
      "mean_token_accuracy": 0.17177464365959166,
      "num_tokens": 18478847.0,
      "step": 7290
    },
    {
      "entropy": 6.358646488189697,
      "epoch": 0.8418926716676284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940307290924923,
      "loss": 6.1507,
      "mean_token_accuracy": 0.16476942002773284,
      "num_tokens": 18490932.0,
      "step": 7295
    },
    {
      "entropy": 6.380516624450683,
      "epoch": 0.842469705712637,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004940212835031565,
      "loss": 6.148,
      "mean_token_accuracy": 0.16350125670433044,
      "num_tokens": 18504333.0,
      "step": 7300
    },
    {
      "entropy": 6.315944814682007,
      "epoch": 0.8430467397576457,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004940118305471486,
      "loss": 6.1177,
      "mean_token_accuracy": 0.1682500198483467,
      "num_tokens": 18516021.0,
      "step": 7305
    },
    {
      "entropy": 6.279258584976196,
      "epoch": 0.8436237738026544,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004940023702247866,
      "loss": 6.0389,
      "mean_token_accuracy": 0.17713080495595931,
      "num_tokens": 18529281.0,
      "step": 7310
    },
    {
      "entropy": 6.269589805603028,
      "epoch": 0.844200807847663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939929025363886,
      "loss": 6.078,
      "mean_token_accuracy": 0.17484681010246278,
      "num_tokens": 18541605.0,
      "step": 7315
    },
    {
      "entropy": 6.37536392211914,
      "epoch": 0.8447778418926717,
      "grad_norm": 0.875,
      "learning_rate": 0.0004939834274822731,
      "loss": 6.1313,
      "mean_token_accuracy": 0.17120876014232636,
      "num_tokens": 18553171.0,
      "step": 7320
    },
    {
      "entropy": 6.356645250320435,
      "epoch": 0.8453548759376803,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939739450627588,
      "loss": 5.9983,
      "mean_token_accuracy": 0.17383179813623428,
      "num_tokens": 18565347.0,
      "step": 7325
    },
    {
      "entropy": 6.309364986419678,
      "epoch": 0.845931909982689,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004939644552781647,
      "loss": 6.1597,
      "mean_token_accuracy": 0.16488830000162125,
      "num_tokens": 18578102.0,
      "step": 7330
    },
    {
      "entropy": 6.322349405288696,
      "epoch": 0.8465089440276976,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939549581288099,
      "loss": 6.126,
      "mean_token_accuracy": 0.16842911392450333,
      "num_tokens": 18590184.0,
      "step": 7335
    },
    {
      "entropy": 6.336501169204712,
      "epoch": 0.8470859780727062,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939454536150138,
      "loss": 6.1421,
      "mean_token_accuracy": 0.16784720569849015,
      "num_tokens": 18603026.0,
      "step": 7340
    },
    {
      "entropy": 6.312475967407226,
      "epoch": 0.847663012117715,
      "grad_norm": 0.78125,
      "learning_rate": 0.000493935941737096,
      "loss": 6.1272,
      "mean_token_accuracy": 0.1679028794169426,
      "num_tokens": 18616841.0,
      "step": 7345
    },
    {
      "entropy": 6.346449661254883,
      "epoch": 0.8482400461627236,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004939264224953768,
      "loss": 6.0874,
      "mean_token_accuracy": 0.17581291496753693,
      "num_tokens": 18628777.0,
      "step": 7350
    },
    {
      "entropy": 6.378287839889526,
      "epoch": 0.8488170802077323,
      "grad_norm": 0.8125,
      "learning_rate": 0.000493916895890176,
      "loss": 6.1389,
      "mean_token_accuracy": 0.16260108798742295,
      "num_tokens": 18641626.0,
      "step": 7355
    },
    {
      "entropy": 6.386941862106323,
      "epoch": 0.8493941142527409,
      "grad_norm": 0.84375,
      "learning_rate": 0.000493907361921814,
      "loss": 6.1056,
      "mean_token_accuracy": 0.1709580108523369,
      "num_tokens": 18653576.0,
      "step": 7360
    },
    {
      "entropy": 6.286987733840943,
      "epoch": 0.8499711482977496,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004938978205906118,
      "loss": 6.1237,
      "mean_token_accuracy": 0.1724120333790779,
      "num_tokens": 18666134.0,
      "step": 7365
    },
    {
      "entropy": 6.289409017562866,
      "epoch": 0.8505481823427582,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938882718968901,
      "loss": 6.0976,
      "mean_token_accuracy": 0.16496776938438415,
      "num_tokens": 18680029.0,
      "step": 7370
    },
    {
      "entropy": 6.40006833076477,
      "epoch": 0.8511252163877668,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004938787158409702,
      "loss": 6.117,
      "mean_token_accuracy": 0.17224864065647125,
      "num_tokens": 18693136.0,
      "step": 7375
    },
    {
      "entropy": 6.338391923904419,
      "epoch": 0.8517022504327756,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004938691524231733,
      "loss": 6.1709,
      "mean_token_accuracy": 0.16487063318490983,
      "num_tokens": 18706613.0,
      "step": 7380
    },
    {
      "entropy": 6.340526151657104,
      "epoch": 0.8522792844777842,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004938595816438212,
      "loss": 6.142,
      "mean_token_accuracy": 0.1600414291024208,
      "num_tokens": 18719541.0,
      "step": 7385
    },
    {
      "entropy": 6.367289924621582,
      "epoch": 0.8528563185227929,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004938500035032358,
      "loss": 6.1351,
      "mean_token_accuracy": 0.17412920743227006,
      "num_tokens": 18730992.0,
      "step": 7390
    },
    {
      "entropy": 6.312771415710449,
      "epoch": 0.8534333525678015,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004938404180017392,
      "loss": 6.0723,
      "mean_token_accuracy": 0.17232910990715028,
      "num_tokens": 18743945.0,
      "step": 7395
    },
    {
      "entropy": 6.260341644287109,
      "epoch": 0.8540103866128101,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004938308251396539,
      "loss": 6.1403,
      "mean_token_accuracy": 0.17283896505832672,
      "num_tokens": 18756656.0,
      "step": 7400
    },
    {
      "entropy": 6.34693398475647,
      "epoch": 0.8545874206578188,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938212249173024,
      "loss": 6.0862,
      "mean_token_accuracy": 0.17401126325130462,
      "num_tokens": 18769224.0,
      "step": 7405
    },
    {
      "entropy": 6.317062568664551,
      "epoch": 0.8551644547028274,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004938116173350078,
      "loss": 6.0992,
      "mean_token_accuracy": 0.17960949689149858,
      "num_tokens": 18782162.0,
      "step": 7410
    },
    {
      "entropy": 6.306476974487305,
      "epoch": 0.8557414887478362,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004938020023930932,
      "loss": 6.149,
      "mean_token_accuracy": 0.17139672636985778,
      "num_tokens": 18793014.0,
      "step": 7415
    },
    {
      "entropy": 6.415736961364746,
      "epoch": 0.8563185227928448,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004937923800918817,
      "loss": 6.2033,
      "mean_token_accuracy": 0.1621045172214508,
      "num_tokens": 18804879.0,
      "step": 7420
    },
    {
      "entropy": 6.363468360900879,
      "epoch": 0.8568955568378535,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004937827504316974,
      "loss": 6.0632,
      "mean_token_accuracy": 0.17355253249406816,
      "num_tokens": 18817301.0,
      "step": 7425
    },
    {
      "entropy": 6.247975254058838,
      "epoch": 0.8574725908828621,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937731134128639,
      "loss": 6.0907,
      "mean_token_accuracy": 0.17120175808668137,
      "num_tokens": 18828424.0,
      "step": 7430
    },
    {
      "entropy": 6.3082421779632565,
      "epoch": 0.8580496249278707,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004937634690357054,
      "loss": 6.1257,
      "mean_token_accuracy": 0.16765787899494172,
      "num_tokens": 18840526.0,
      "step": 7435
    },
    {
      "entropy": 6.408977603912353,
      "epoch": 0.8586266589728794,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004937538173005462,
      "loss": 6.2009,
      "mean_token_accuracy": 0.16263508349657058,
      "num_tokens": 18854557.0,
      "step": 7440
    },
    {
      "entropy": 6.323116827011108,
      "epoch": 0.859203693017888,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937441582077111,
      "loss": 6.1389,
      "mean_token_accuracy": 0.16918336153030394,
      "num_tokens": 18868469.0,
      "step": 7445
    },
    {
      "entropy": 6.341149616241455,
      "epoch": 0.8597807270628968,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004937344917575249,
      "loss": 6.1014,
      "mean_token_accuracy": 0.170886567234993,
      "num_tokens": 18880507.0,
      "step": 7450
    },
    {
      "entropy": 6.391666841506958,
      "epoch": 0.8603577611079054,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937248179503127,
      "loss": 6.1863,
      "mean_token_accuracy": 0.15916220992803573,
      "num_tokens": 18892666.0,
      "step": 7455
    },
    {
      "entropy": 6.386204576492309,
      "epoch": 0.860934795152914,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004937151367863998,
      "loss": 6.1659,
      "mean_token_accuracy": 0.16778618693351746,
      "num_tokens": 18905022.0,
      "step": 7460
    },
    {
      "entropy": 6.382054328918457,
      "epoch": 0.8615118291979227,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000493705448266112,
      "loss": 6.1788,
      "mean_token_accuracy": 0.16661759167909623,
      "num_tokens": 18918279.0,
      "step": 7465
    },
    {
      "entropy": 6.320206069946289,
      "epoch": 0.8620888632429313,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936957523897752,
      "loss": 6.0787,
      "mean_token_accuracy": 0.17139777094125747,
      "num_tokens": 18930494.0,
      "step": 7470
    },
    {
      "entropy": 6.276682758331299,
      "epoch": 0.86266589728794,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004936860491577153,
      "loss": 6.0333,
      "mean_token_accuracy": 0.17490018606185914,
      "num_tokens": 18943918.0,
      "step": 7475
    },
    {
      "entropy": 6.308687543869018,
      "epoch": 0.8632429313329486,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004936763385702588,
      "loss": 6.146,
      "mean_token_accuracy": 0.1680581673979759,
      "num_tokens": 18957170.0,
      "step": 7480
    },
    {
      "entropy": 6.293296480178833,
      "epoch": 0.8638199653779572,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004936666206277322,
      "loss": 6.0779,
      "mean_token_accuracy": 0.1573074623942375,
      "num_tokens": 18970106.0,
      "step": 7485
    },
    {
      "entropy": 6.374524068832398,
      "epoch": 0.864396999422966,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004936568953304624,
      "loss": 6.152,
      "mean_token_accuracy": 0.16933187693357468,
      "num_tokens": 18982797.0,
      "step": 7490
    },
    {
      "entropy": 6.353385543823242,
      "epoch": 0.8649740334679746,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004936471626787766,
      "loss": 6.2206,
      "mean_token_accuracy": 0.16988059729337693,
      "num_tokens": 18997122.0,
      "step": 7495
    },
    {
      "entropy": 6.360155010223389,
      "epoch": 0.8655510675129833,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004936374226730022,
      "loss": 6.1914,
      "mean_token_accuracy": 0.16151489317417145,
      "num_tokens": 19009345.0,
      "step": 7500
    },
    {
      "entropy": 6.4074300765991214,
      "epoch": 0.8661281015579919,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004936276753134666,
      "loss": 6.1776,
      "mean_token_accuracy": 0.16569527089595795,
      "num_tokens": 19021585.0,
      "step": 7505
    },
    {
      "entropy": 6.347685956954956,
      "epoch": 0.8667051356030006,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004936179206004976,
      "loss": 6.147,
      "mean_token_accuracy": 0.16409811973571778,
      "num_tokens": 19034701.0,
      "step": 7510
    },
    {
      "entropy": 6.303962707519531,
      "epoch": 0.8672821696480092,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936081585344236,
      "loss": 6.117,
      "mean_token_accuracy": 0.17443791031837463,
      "num_tokens": 19047810.0,
      "step": 7515
    },
    {
      "entropy": 6.332028913497925,
      "epoch": 0.8678592036930178,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004935983891155727,
      "loss": 6.1125,
      "mean_token_accuracy": 0.17651036232709885,
      "num_tokens": 19059546.0,
      "step": 7520
    },
    {
      "entropy": 6.37941575050354,
      "epoch": 0.8684362377380266,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004935886123442737,
      "loss": 6.1784,
      "mean_token_accuracy": 0.16661784052848816,
      "num_tokens": 19072302.0,
      "step": 7525
    },
    {
      "entropy": 6.346638488769531,
      "epoch": 0.8690132717830352,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004935788282208551,
      "loss": 6.1497,
      "mean_token_accuracy": 0.17058232575654983,
      "num_tokens": 19086283.0,
      "step": 7530
    },
    {
      "entropy": 6.313622093200683,
      "epoch": 0.8695903058280439,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004935690367456464,
      "loss": 6.0776,
      "mean_token_accuracy": 0.1749485731124878,
      "num_tokens": 19098586.0,
      "step": 7535
    },
    {
      "entropy": 6.319880771636963,
      "epoch": 0.8701673398730525,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004935592379189766,
      "loss": 6.0841,
      "mean_token_accuracy": 0.1700226277112961,
      "num_tokens": 19110795.0,
      "step": 7540
    },
    {
      "entropy": 6.340610694885254,
      "epoch": 0.8707443739180611,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004935494317411754,
      "loss": 6.0211,
      "mean_token_accuracy": 0.17584108859300612,
      "num_tokens": 19123391.0,
      "step": 7545
    },
    {
      "entropy": 6.319567918777466,
      "epoch": 0.8713214079630698,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004935396182125726,
      "loss": 6.1084,
      "mean_token_accuracy": 0.16964961737394332,
      "num_tokens": 19135436.0,
      "step": 7550
    },
    {
      "entropy": 6.301428127288818,
      "epoch": 0.8718984420080784,
      "grad_norm": 0.875,
      "learning_rate": 0.0004935297973334983,
      "loss": 6.0731,
      "mean_token_accuracy": 0.17406256645917892,
      "num_tokens": 19147210.0,
      "step": 7555
    },
    {
      "entropy": 6.340644502639771,
      "epoch": 0.8724754760530872,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004935199691042828,
      "loss": 6.2168,
      "mean_token_accuracy": 0.17108030915260314,
      "num_tokens": 19161309.0,
      "step": 7560
    },
    {
      "entropy": 6.327101135253907,
      "epoch": 0.8730525100980958,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004935101335252568,
      "loss": 6.1608,
      "mean_token_accuracy": 0.16548355668783188,
      "num_tokens": 19174073.0,
      "step": 7565
    },
    {
      "entropy": 6.321848392486572,
      "epoch": 0.8736295441431045,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004935002905967509,
      "loss": 6.0431,
      "mean_token_accuracy": 0.17418338358402252,
      "num_tokens": 19187120.0,
      "step": 7570
    },
    {
      "entropy": 6.35199728012085,
      "epoch": 0.8742065781881131,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004934904403190963,
      "loss": 6.0181,
      "mean_token_accuracy": 0.17857957929372786,
      "num_tokens": 19201195.0,
      "step": 7575
    },
    {
      "entropy": 6.296740293502808,
      "epoch": 0.8747836122331217,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004934805826926242,
      "loss": 6.1638,
      "mean_token_accuracy": 0.16306511014699937,
      "num_tokens": 19214050.0,
      "step": 7580
    },
    {
      "entropy": 6.384359073638916,
      "epoch": 0.8753606462781304,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004934707177176663,
      "loss": 6.1338,
      "mean_token_accuracy": 0.1667471244931221,
      "num_tokens": 19226678.0,
      "step": 7585
    },
    {
      "entropy": 6.370206451416015,
      "epoch": 0.875937680323139,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004934608453945543,
      "loss": 6.1668,
      "mean_token_accuracy": 0.1634979486465454,
      "num_tokens": 19239038.0,
      "step": 7590
    },
    {
      "entropy": 6.182282304763794,
      "epoch": 0.8765147143681478,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004934509657236203,
      "loss": 6.0475,
      "mean_token_accuracy": 0.17127805054187775,
      "num_tokens": 19252934.0,
      "step": 7595
    },
    {
      "entropy": 6.3947070121765135,
      "epoch": 0.8770917484131564,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004934410787051965,
      "loss": 6.1126,
      "mean_token_accuracy": 0.16948466897010803,
      "num_tokens": 19267646.0,
      "step": 7600
    },
    {
      "entropy": 6.35531792640686,
      "epoch": 0.877668782458165,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004934311843396157,
      "loss": 6.1142,
      "mean_token_accuracy": 0.16879450678825378,
      "num_tokens": 19279529.0,
      "step": 7605
    },
    {
      "entropy": 6.342835569381714,
      "epoch": 0.8782458165031737,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004934212826272104,
      "loss": 6.1782,
      "mean_token_accuracy": 0.17134388089179992,
      "num_tokens": 19291315.0,
      "step": 7610
    },
    {
      "entropy": 6.2911194324493405,
      "epoch": 0.8788228505481823,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004934113735683137,
      "loss": 6.0069,
      "mean_token_accuracy": 0.17789260894060135,
      "num_tokens": 19303763.0,
      "step": 7615
    },
    {
      "entropy": 6.340110778808594,
      "epoch": 0.879399884593191,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000493401457163259,
      "loss": 6.1417,
      "mean_token_accuracy": 0.16427685767412187,
      "num_tokens": 19316400.0,
      "step": 7620
    },
    {
      "entropy": 6.295943403244019,
      "epoch": 0.8799769186381996,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004933915334123798,
      "loss": 6.0343,
      "mean_token_accuracy": 0.1810584247112274,
      "num_tokens": 19328227.0,
      "step": 7625
    },
    {
      "entropy": 6.221355772018432,
      "epoch": 0.8805539526832084,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004933816023160099,
      "loss": 6.0023,
      "mean_token_accuracy": 0.1760854810476303,
      "num_tokens": 19341211.0,
      "step": 7630
    },
    {
      "entropy": 6.314968013763428,
      "epoch": 0.881130986728217,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004933716638744832,
      "loss": 6.1554,
      "mean_token_accuracy": 0.17098452597856523,
      "num_tokens": 19353165.0,
      "step": 7635
    },
    {
      "entropy": 6.37896466255188,
      "epoch": 0.8817080207732256,
      "grad_norm": 1.703125,
      "learning_rate": 0.000493361718088134,
      "loss": 6.1926,
      "mean_token_accuracy": 0.17017002701759337,
      "num_tokens": 19365601.0,
      "step": 7640
    },
    {
      "entropy": 6.392019462585449,
      "epoch": 0.8822850548182343,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000493351764957297,
      "loss": 6.1205,
      "mean_token_accuracy": 0.16742375642061233,
      "num_tokens": 19377828.0,
      "step": 7645
    },
    {
      "entropy": 6.3836784839630125,
      "epoch": 0.8828620888632429,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004933418044823068,
      "loss": 6.1313,
      "mean_token_accuracy": 0.16712106615304947,
      "num_tokens": 19390019.0,
      "step": 7650
    },
    {
      "entropy": 6.283777713775635,
      "epoch": 0.8834391229082516,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004933318366634984,
      "loss": 6.0898,
      "mean_token_accuracy": 0.17001585215330123,
      "num_tokens": 19402179.0,
      "step": 7655
    },
    {
      "entropy": 6.335080194473266,
      "epoch": 0.8840161569532602,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933218615012072,
      "loss": 6.0511,
      "mean_token_accuracy": 0.16598242968320848,
      "num_tokens": 19412794.0,
      "step": 7660
    },
    {
      "entropy": 6.311235332489014,
      "epoch": 0.8845931909982689,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004933118789957687,
      "loss": 6.1578,
      "mean_token_accuracy": 0.16024495661258698,
      "num_tokens": 19425932.0,
      "step": 7665
    },
    {
      "entropy": 6.370776271820068,
      "epoch": 0.8851702250432776,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004933018891475186,
      "loss": 6.086,
      "mean_token_accuracy": 0.1693466141819954,
      "num_tokens": 19437543.0,
      "step": 7670
    },
    {
      "entropy": 6.352818298339844,
      "epoch": 0.8857472590882862,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004932918919567927,
      "loss": 6.0564,
      "mean_token_accuracy": 0.1737107202410698,
      "num_tokens": 19450122.0,
      "step": 7675
    },
    {
      "entropy": 6.24740195274353,
      "epoch": 0.8863242931332949,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004932818874239275,
      "loss": 6.0626,
      "mean_token_accuracy": 0.1702693447470665,
      "num_tokens": 19464553.0,
      "step": 7680
    },
    {
      "entropy": 6.240301847457886,
      "epoch": 0.8869013271783035,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004932718755492595,
      "loss": 6.0379,
      "mean_token_accuracy": 0.17298130840063095,
      "num_tokens": 19477956.0,
      "step": 7685
    },
    {
      "entropy": 6.28462176322937,
      "epoch": 0.8874783612233121,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004932618563331254,
      "loss": 6.0811,
      "mean_token_accuracy": 0.16695266515016555,
      "num_tokens": 19491610.0,
      "step": 7690
    },
    {
      "entropy": 6.30009708404541,
      "epoch": 0.8880553952683208,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004932518297758622,
      "loss": 5.9959,
      "mean_token_accuracy": 0.17590903639793395,
      "num_tokens": 19503612.0,
      "step": 7695
    },
    {
      "entropy": 6.268136787414551,
      "epoch": 0.8886324293133295,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004932417958778071,
      "loss": 6.0582,
      "mean_token_accuracy": 0.17468879073858262,
      "num_tokens": 19516956.0,
      "step": 7700
    },
    {
      "entropy": 6.229768562316894,
      "epoch": 0.8892094633583382,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932317546392976,
      "loss": 6.0549,
      "mean_token_accuracy": 0.1741514101624489,
      "num_tokens": 19528545.0,
      "step": 7705
    },
    {
      "entropy": 6.329782152175904,
      "epoch": 0.8897864974033468,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932217060606714,
      "loss": 5.9936,
      "mean_token_accuracy": 0.178321772813797,
      "num_tokens": 19539920.0,
      "step": 7710
    },
    {
      "entropy": 6.283861589431763,
      "epoch": 0.8903635314483554,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004932116501422665,
      "loss": 6.1142,
      "mean_token_accuracy": 0.17093467265367507,
      "num_tokens": 19551975.0,
      "step": 7715
    },
    {
      "entropy": 6.328822898864746,
      "epoch": 0.8909405654933641,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004932015868844211,
      "loss": 6.1172,
      "mean_token_accuracy": 0.16731198877096176,
      "num_tokens": 19565952.0,
      "step": 7720
    },
    {
      "entropy": 6.282176876068116,
      "epoch": 0.8915175995383727,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004931915162874738,
      "loss": 6.0689,
      "mean_token_accuracy": 0.17317880243062972,
      "num_tokens": 19578400.0,
      "step": 7725
    },
    {
      "entropy": 6.242160987854004,
      "epoch": 0.8920946335833814,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004931814383517632,
      "loss": 6.0374,
      "mean_token_accuracy": 0.17461694926023483,
      "num_tokens": 19592237.0,
      "step": 7730
    },
    {
      "entropy": 6.386451339721679,
      "epoch": 0.8926716676283901,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004931713530776284,
      "loss": 6.2046,
      "mean_token_accuracy": 0.16910806745290757,
      "num_tokens": 19606172.0,
      "step": 7735
    },
    {
      "entropy": 6.321779870986939,
      "epoch": 0.8932487016733988,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004931612604654083,
      "loss": 6.1511,
      "mean_token_accuracy": 0.16871884763240813,
      "num_tokens": 19618245.0,
      "step": 7740
    },
    {
      "entropy": 6.36405234336853,
      "epoch": 0.8938257357184074,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004931511605154428,
      "loss": 6.1005,
      "mean_token_accuracy": 0.17241024523973464,
      "num_tokens": 19630921.0,
      "step": 7745
    },
    {
      "entropy": 6.329953145980835,
      "epoch": 0.894402769763416,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004931410532280711,
      "loss": 6.0928,
      "mean_token_accuracy": 0.16922309696674348,
      "num_tokens": 19643869.0,
      "step": 7750
    },
    {
      "entropy": 6.352374649047851,
      "epoch": 0.8949798038084247,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004931309386036337,
      "loss": 6.0813,
      "mean_token_accuracy": 0.17119230180978776,
      "num_tokens": 19657819.0,
      "step": 7755
    },
    {
      "entropy": 6.346218204498291,
      "epoch": 0.8955568378534333,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931208166424704,
      "loss": 6.1593,
      "mean_token_accuracy": 0.16632794886827468,
      "num_tokens": 19670847.0,
      "step": 7760
    },
    {
      "entropy": 6.324069356918335,
      "epoch": 0.896133871898442,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004931106873449219,
      "loss": 6.076,
      "mean_token_accuracy": 0.16754509657621383,
      "num_tokens": 19684055.0,
      "step": 7765
    },
    {
      "entropy": 6.251480340957642,
      "epoch": 0.8967109059434507,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931005507113285,
      "loss": 5.9831,
      "mean_token_accuracy": 0.1751714378595352,
      "num_tokens": 19696731.0,
      "step": 7770
    },
    {
      "entropy": 6.362442827224731,
      "epoch": 0.8972879399884593,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930904067420317,
      "loss": 6.1246,
      "mean_token_accuracy": 0.16654115468263625,
      "num_tokens": 19708787.0,
      "step": 7775
    },
    {
      "entropy": 6.344279527664185,
      "epoch": 0.897864974033468,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004930802554373723,
      "loss": 6.1366,
      "mean_token_accuracy": 0.16518602818250655,
      "num_tokens": 19722570.0,
      "step": 7780
    },
    {
      "entropy": 6.365567255020141,
      "epoch": 0.8984420080784766,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004930700967976918,
      "loss": 6.1262,
      "mean_token_accuracy": 0.17212459594011306,
      "num_tokens": 19734690.0,
      "step": 7785
    },
    {
      "entropy": 6.334938287734985,
      "epoch": 0.8990190421234853,
      "grad_norm": 0.828125,
      "learning_rate": 0.000493059930823332,
      "loss": 6.1356,
      "mean_token_accuracy": 0.16845792382955552,
      "num_tokens": 19747907.0,
      "step": 7790
    },
    {
      "entropy": 6.319498729705811,
      "epoch": 0.8995960761684939,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004930497575146346,
      "loss": 6.1109,
      "mean_token_accuracy": 0.16970169842243193,
      "num_tokens": 19760869.0,
      "step": 7795
    },
    {
      "entropy": 6.298462533950806,
      "epoch": 0.9001731102135025,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930395768719421,
      "loss": 6.0006,
      "mean_token_accuracy": 0.17433494627475737,
      "num_tokens": 19773097.0,
      "step": 7800
    },
    {
      "entropy": 6.287778615951538,
      "epoch": 0.9007501442585113,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004930293888955966,
      "loss": 6.1284,
      "mean_token_accuracy": 0.16367465257644653,
      "num_tokens": 19786070.0,
      "step": 7805
    },
    {
      "entropy": 6.361891651153565,
      "epoch": 0.9013271783035199,
      "grad_norm": 0.78125,
      "learning_rate": 0.000493019193585941,
      "loss": 6.1124,
      "mean_token_accuracy": 0.17385626435279847,
      "num_tokens": 19799244.0,
      "step": 7810
    },
    {
      "entropy": 6.301440572738647,
      "epoch": 0.9019042123485286,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000493008990943318,
      "loss": 6.144,
      "mean_token_accuracy": 0.16414995938539506,
      "num_tokens": 19811965.0,
      "step": 7815
    },
    {
      "entropy": 6.37474365234375,
      "epoch": 0.9024812463935372,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004929987809680709,
      "loss": 6.1108,
      "mean_token_accuracy": 0.16805853843688964,
      "num_tokens": 19825531.0,
      "step": 7820
    },
    {
      "entropy": 6.321238422393799,
      "epoch": 0.9030582804385459,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004929885636605432,
      "loss": 6.0547,
      "mean_token_accuracy": 0.17446014732122422,
      "num_tokens": 19839166.0,
      "step": 7825
    },
    {
      "entropy": 6.3322361469268795,
      "epoch": 0.9036353144835545,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004929783390210784,
      "loss": 6.0863,
      "mean_token_accuracy": 0.16568157374858855,
      "num_tokens": 19852187.0,
      "step": 7830
    },
    {
      "entropy": 6.315406608581543,
      "epoch": 0.9042123485285631,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004929681070500204,
      "loss": 6.1376,
      "mean_token_accuracy": 0.16454064548015596,
      "num_tokens": 19864309.0,
      "step": 7835
    },
    {
      "entropy": 6.316937685012817,
      "epoch": 0.9047893825735719,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004929578677477135,
      "loss": 6.1439,
      "mean_token_accuracy": 0.16100410372018814,
      "num_tokens": 19877475.0,
      "step": 7840
    },
    {
      "entropy": 6.378486156463623,
      "epoch": 0.9053664166185805,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929476211145019,
      "loss": 6.0833,
      "mean_token_accuracy": 0.17034784704446793,
      "num_tokens": 19889984.0,
      "step": 7845
    },
    {
      "entropy": 6.299851369857788,
      "epoch": 0.9059434506635892,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004929373671507303,
      "loss": 6.0405,
      "mean_token_accuracy": 0.1822006508708,
      "num_tokens": 19902135.0,
      "step": 7850
    },
    {
      "entropy": 6.269701147079468,
      "epoch": 0.9065204847085978,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004929271058567436,
      "loss": 6.0957,
      "mean_token_accuracy": 0.17251716405153275,
      "num_tokens": 19915214.0,
      "step": 7855
    },
    {
      "entropy": 6.321777105331421,
      "epoch": 0.9070975187536064,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492916837232887,
      "loss": 6.1169,
      "mean_token_accuracy": 0.1645059496164322,
      "num_tokens": 19929113.0,
      "step": 7860
    },
    {
      "entropy": 6.430020093917847,
      "epoch": 0.9076745527986151,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929065612795058,
      "loss": 6.142,
      "mean_token_accuracy": 0.16679659783840178,
      "num_tokens": 19942727.0,
      "step": 7865
    },
    {
      "entropy": 6.271768093109131,
      "epoch": 0.9082515868436237,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928962779969456,
      "loss": 6.0785,
      "mean_token_accuracy": 0.17226183861494065,
      "num_tokens": 19954863.0,
      "step": 7870
    },
    {
      "entropy": 6.243986463546753,
      "epoch": 0.9088286208886325,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928859873855524,
      "loss": 6.0212,
      "mean_token_accuracy": 0.179158616065979,
      "num_tokens": 19967555.0,
      "step": 7875
    },
    {
      "entropy": 6.396019315719604,
      "epoch": 0.9094056549336411,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004928756894456723,
      "loss": 6.1286,
      "mean_token_accuracy": 0.1713392600417137,
      "num_tokens": 19979992.0,
      "step": 7880
    },
    {
      "entropy": 6.345692825317383,
      "epoch": 0.9099826889786498,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004928653841776515,
      "loss": 6.1643,
      "mean_token_accuracy": 0.16938916742801666,
      "num_tokens": 19993231.0,
      "step": 7885
    },
    {
      "entropy": 6.371894836425781,
      "epoch": 0.9105597230236584,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004928550715818368,
      "loss": 6.1413,
      "mean_token_accuracy": 0.1652555137872696,
      "num_tokens": 20005839.0,
      "step": 7890
    },
    {
      "entropy": 6.319307708740235,
      "epoch": 0.911136757068667,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004928447516585749,
      "loss": 6.1078,
      "mean_token_accuracy": 0.16926864087581633,
      "num_tokens": 20017423.0,
      "step": 7895
    },
    {
      "entropy": 6.3037055969238285,
      "epoch": 0.9117137911136757,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492834424408213,
      "loss": 6.1399,
      "mean_token_accuracy": 0.1685601219534874,
      "num_tokens": 20030989.0,
      "step": 7900
    },
    {
      "entropy": 6.361561441421509,
      "epoch": 0.9122908251586843,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004928240898310984,
      "loss": 6.1912,
      "mean_token_accuracy": 0.16204766035079957,
      "num_tokens": 20043763.0,
      "step": 7905
    },
    {
      "entropy": 6.3040454387664795,
      "epoch": 0.9128678592036931,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004928137479275789,
      "loss": 6.0531,
      "mean_token_accuracy": 0.17904019355773926,
      "num_tokens": 20057012.0,
      "step": 7910
    },
    {
      "entropy": 6.308187437057495,
      "epoch": 0.9134448932487017,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000492803398698002,
      "loss": 6.0794,
      "mean_token_accuracy": 0.1680925503373146,
      "num_tokens": 20069093.0,
      "step": 7915
    },
    {
      "entropy": 6.265644502639771,
      "epoch": 0.9140219272937103,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004927930421427161,
      "loss": 6.0545,
      "mean_token_accuracy": 0.16955652683973313,
      "num_tokens": 20081392.0,
      "step": 7920
    },
    {
      "entropy": 6.296371650695801,
      "epoch": 0.914598961338719,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927826782620694,
      "loss": 6.0728,
      "mean_token_accuracy": 0.17164398729801178,
      "num_tokens": 20092621.0,
      "step": 7925
    },
    {
      "entropy": 6.3193436622619625,
      "epoch": 0.9151759953837276,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004927723070564104,
      "loss": 6.0926,
      "mean_token_accuracy": 0.16736245900392532,
      "num_tokens": 20105423.0,
      "step": 7930
    },
    {
      "entropy": 6.328817844390869,
      "epoch": 0.9157530294287363,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004927619285260881,
      "loss": 6.0829,
      "mean_token_accuracy": 0.1673406407237053,
      "num_tokens": 20117835.0,
      "step": 7935
    },
    {
      "entropy": 6.243988561630249,
      "epoch": 0.9163300634737449,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004927515426714515,
      "loss": 5.9606,
      "mean_token_accuracy": 0.1805577903985977,
      "num_tokens": 20130563.0,
      "step": 7940
    },
    {
      "entropy": 6.216890621185303,
      "epoch": 0.9169070975187537,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00049274114949285,
      "loss": 6.0384,
      "mean_token_accuracy": 0.1770210310816765,
      "num_tokens": 20144507.0,
      "step": 7945
    },
    {
      "entropy": 6.297721862792969,
      "epoch": 0.9174841315637623,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000492730748990633,
      "loss": 6.0587,
      "mean_token_accuracy": 0.17653549313545228,
      "num_tokens": 20156341.0,
      "step": 7950
    },
    {
      "entropy": 6.281712436676026,
      "epoch": 0.9180611656087709,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004927203411651504,
      "loss": 6.1138,
      "mean_token_accuracy": 0.17542605996131896,
      "num_tokens": 20169410.0,
      "step": 7955
    },
    {
      "entropy": 6.367187356948852,
      "epoch": 0.9186381996537796,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927099260167523,
      "loss": 6.1011,
      "mean_token_accuracy": 0.16979494839906692,
      "num_tokens": 20181233.0,
      "step": 7960
    },
    {
      "entropy": 6.352732086181641,
      "epoch": 0.9192152336987882,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004926995035457889,
      "loss": 6.1114,
      "mean_token_accuracy": 0.17020532637834548,
      "num_tokens": 20193234.0,
      "step": 7965
    },
    {
      "entropy": 6.235786628723145,
      "epoch": 0.9197922677437969,
      "grad_norm": 0.8125,
      "learning_rate": 0.000492689073752611,
      "loss": 6.0073,
      "mean_token_accuracy": 0.17449528872966766,
      "num_tokens": 20206784.0,
      "step": 7970
    },
    {
      "entropy": 6.272397851943969,
      "epoch": 0.9203693017888055,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004926786366375691,
      "loss": 6.0271,
      "mean_token_accuracy": 0.18076363652944566,
      "num_tokens": 20218597.0,
      "step": 7975
    },
    {
      "entropy": 6.223327016830444,
      "epoch": 0.9209463358338142,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004926681922010145,
      "loss": 6.0001,
      "mean_token_accuracy": 0.17741193175315856,
      "num_tokens": 20232703.0,
      "step": 7980
    },
    {
      "entropy": 6.312099599838257,
      "epoch": 0.9215233698788229,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004926577404432982,
      "loss": 6.08,
      "mean_token_accuracy": 0.1708337426185608,
      "num_tokens": 20245811.0,
      "step": 7985
    },
    {
      "entropy": 6.2235712051391605,
      "epoch": 0.9221004039238315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004926472813647721,
      "loss": 5.9707,
      "mean_token_accuracy": 0.17567360550165176,
      "num_tokens": 20258055.0,
      "step": 7990
    },
    {
      "entropy": 6.328189039230347,
      "epoch": 0.9226774379688402,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004926368149657876,
      "loss": 6.0672,
      "mean_token_accuracy": 0.1733057051897049,
      "num_tokens": 20271250.0,
      "step": 7995
    },
    {
      "entropy": 6.267770671844483,
      "epoch": 0.9232544720138488,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004926263412466972,
      "loss": 6.0083,
      "mean_token_accuracy": 0.17971379309892654,
      "num_tokens": 20284013.0,
      "step": 8000
    },
    {
      "entropy": 6.2823432922363285,
      "epoch": 0.9238315060588574,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004926158602078527,
      "loss": 6.0848,
      "mean_token_accuracy": 0.17510406225919722,
      "num_tokens": 20297413.0,
      "step": 8005
    },
    {
      "entropy": 6.262951850891113,
      "epoch": 0.9244085401038661,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004926053718496069,
      "loss": 6.0599,
      "mean_token_accuracy": 0.1757744312286377,
      "num_tokens": 20310026.0,
      "step": 8010
    },
    {
      "entropy": 6.244164228439331,
      "epoch": 0.9249855741488748,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004925948761723126,
      "loss": 6.0354,
      "mean_token_accuracy": 0.18111930638551713,
      "num_tokens": 20322118.0,
      "step": 8015
    },
    {
      "entropy": 6.328146457672119,
      "epoch": 0.9255626081938835,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004925843731763226,
      "loss": 6.1125,
      "mean_token_accuracy": 0.16659914404153825,
      "num_tokens": 20335252.0,
      "step": 8020
    },
    {
      "entropy": 6.4609192371368405,
      "epoch": 0.9261396422388921,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004925738628619904,
      "loss": 6.143,
      "mean_token_accuracy": 0.16860777735710145,
      "num_tokens": 20348201.0,
      "step": 8025
    },
    {
      "entropy": 6.339458322525024,
      "epoch": 0.9267166762839008,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004925633452296693,
      "loss": 6.0803,
      "mean_token_accuracy": 0.17183510214090347,
      "num_tokens": 20360250.0,
      "step": 8030
    },
    {
      "entropy": 6.266357278823852,
      "epoch": 0.9272937103289094,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004925528202797131,
      "loss": 6.0533,
      "mean_token_accuracy": 0.17167010456323623,
      "num_tokens": 20372376.0,
      "step": 8035
    },
    {
      "entropy": 6.336796998977661,
      "epoch": 0.927870744373918,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004925422880124761,
      "loss": 6.0416,
      "mean_token_accuracy": 0.16531915813684464,
      "num_tokens": 20384484.0,
      "step": 8040
    },
    {
      "entropy": 6.330718564987182,
      "epoch": 0.9284477784189267,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004925317484283121,
      "loss": 6.1023,
      "mean_token_accuracy": 0.1714852049946785,
      "num_tokens": 20396519.0,
      "step": 8045
    },
    {
      "entropy": 6.303335094451905,
      "epoch": 0.9290248124639354,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004925212015275758,
      "loss": 6.0762,
      "mean_token_accuracy": 0.1695145472884178,
      "num_tokens": 20409108.0,
      "step": 8050
    },
    {
      "entropy": 6.31664810180664,
      "epoch": 0.9296018465089441,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000492510647310622,
      "loss": 6.0988,
      "mean_token_accuracy": 0.17210331559181213,
      "num_tokens": 20421929.0,
      "step": 8055
    },
    {
      "entropy": 6.286113023757935,
      "epoch": 0.9301788805539527,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004925000857778055,
      "loss": 6.1092,
      "mean_token_accuracy": 0.17088423371315004,
      "num_tokens": 20435453.0,
      "step": 8060
    },
    {
      "entropy": 6.2916816711425785,
      "epoch": 0.9307559145989613,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004924895169294818,
      "loss": 6.0715,
      "mean_token_accuracy": 0.17014946788549423,
      "num_tokens": 20447867.0,
      "step": 8065
    },
    {
      "entropy": 6.322897338867188,
      "epoch": 0.93133294864397,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004924789407660062,
      "loss": 6.1249,
      "mean_token_accuracy": 0.17596817016601562,
      "num_tokens": 20461150.0,
      "step": 8070
    },
    {
      "entropy": 6.285948038101196,
      "epoch": 0.9319099826889786,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004924683572877345,
      "loss": 5.9925,
      "mean_token_accuracy": 0.1782120630145073,
      "num_tokens": 20474012.0,
      "step": 8075
    },
    {
      "entropy": 6.347293901443481,
      "epoch": 0.9324870167339873,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004924577664950225,
      "loss": 6.0593,
      "mean_token_accuracy": 0.16766516119241714,
      "num_tokens": 20485833.0,
      "step": 8080
    },
    {
      "entropy": 6.313635921478271,
      "epoch": 0.933064050778996,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004924471683882266,
      "loss": 6.098,
      "mean_token_accuracy": 0.1738639682531357,
      "num_tokens": 20499277.0,
      "step": 8085
    },
    {
      "entropy": 6.2867530345916744,
      "epoch": 0.9336410848240047,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004924365629677031,
      "loss": 6.1016,
      "mean_token_accuracy": 0.1721029207110405,
      "num_tokens": 20510789.0,
      "step": 8090
    },
    {
      "entropy": 6.264944648742675,
      "epoch": 0.9342181188690133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492425950233809,
      "loss": 6.0445,
      "mean_token_accuracy": 0.17780997157096862,
      "num_tokens": 20523273.0,
      "step": 8095
    },
    {
      "entropy": 6.350660991668701,
      "epoch": 0.9347951529140219,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000492415330186901,
      "loss": 6.1288,
      "mean_token_accuracy": 0.1704434037208557,
      "num_tokens": 20535466.0,
      "step": 8100
    },
    {
      "entropy": 6.417326831817627,
      "epoch": 0.9353721869590306,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004924047028273364,
      "loss": 6.1848,
      "mean_token_accuracy": 0.16572478860616685,
      "num_tokens": 20548108.0,
      "step": 8105
    },
    {
      "entropy": 6.274970197677613,
      "epoch": 0.9359492210040392,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004923940681554725,
      "loss": 6.042,
      "mean_token_accuracy": 0.17275072634220123,
      "num_tokens": 20559859.0,
      "step": 8110
    },
    {
      "entropy": 6.299347448348999,
      "epoch": 0.9365262550490479,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004923834261716672,
      "loss": 6.1575,
      "mean_token_accuracy": 0.16803875118494033,
      "num_tokens": 20572569.0,
      "step": 8115
    },
    {
      "entropy": 6.3162522315979,
      "epoch": 0.9371032890940566,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004923727768762782,
      "loss": 6.0713,
      "mean_token_accuracy": 0.16688956767320634,
      "num_tokens": 20585014.0,
      "step": 8120
    },
    {
      "entropy": 6.309860801696777,
      "epoch": 0.9376803231390652,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492362120269664,
      "loss": 6.1442,
      "mean_token_accuracy": 0.16940630078315735,
      "num_tokens": 20597216.0,
      "step": 8125
    },
    {
      "entropy": 6.226264476776123,
      "epoch": 0.9382573571840739,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004923514563521827,
      "loss": 6.0322,
      "mean_token_accuracy": 0.1835881695151329,
      "num_tokens": 20609193.0,
      "step": 8130
    },
    {
      "entropy": 6.315596437454223,
      "epoch": 0.9388343912290825,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004923407851241933,
      "loss": 6.0794,
      "mean_token_accuracy": 0.16989486366510392,
      "num_tokens": 20620777.0,
      "step": 8135
    },
    {
      "entropy": 6.278818798065186,
      "epoch": 0.9394114252740912,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004923301065860545,
      "loss": 6.0672,
      "mean_token_accuracy": 0.1740580141544342,
      "num_tokens": 20633244.0,
      "step": 8140
    },
    {
      "entropy": 6.323602867126465,
      "epoch": 0.9399884593190998,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004923194207381254,
      "loss": 6.074,
      "mean_token_accuracy": 0.17614427506923674,
      "num_tokens": 20645622.0,
      "step": 8145
    },
    {
      "entropy": 6.351393985748291,
      "epoch": 0.9405654933641084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004923087275807656,
      "loss": 6.1534,
      "mean_token_accuracy": 0.16767651587724686,
      "num_tokens": 20658793.0,
      "step": 8150
    },
    {
      "entropy": 6.271270132064819,
      "epoch": 0.9411425274091172,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004922980271143347,
      "loss": 5.975,
      "mean_token_accuracy": 0.18465787321329116,
      "num_tokens": 20673009.0,
      "step": 8155
    },
    {
      "entropy": 6.3714597702026365,
      "epoch": 0.9417195614541258,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922873193391927,
      "loss": 6.1473,
      "mean_token_accuracy": 0.1721123456954956,
      "num_tokens": 20685061.0,
      "step": 8160
    },
    {
      "entropy": 6.293173694610596,
      "epoch": 0.9422965954991345,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004922766042556994,
      "loss": 5.9916,
      "mean_token_accuracy": 0.17162948995828628,
      "num_tokens": 20696854.0,
      "step": 8165
    },
    {
      "entropy": 6.2723010063171385,
      "epoch": 0.9428736295441431,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004922658818642156,
      "loss": 6.1031,
      "mean_token_accuracy": 0.17009749859571457,
      "num_tokens": 20709259.0,
      "step": 8170
    },
    {
      "entropy": 6.281980180740357,
      "epoch": 0.9434506635891518,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004922551521651018,
      "loss": 5.937,
      "mean_token_accuracy": 0.18290375918149948,
      "num_tokens": 20722382.0,
      "step": 8175
    },
    {
      "entropy": 6.220030355453491,
      "epoch": 0.9440276976341604,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922444151587189,
      "loss": 5.9785,
      "mean_token_accuracy": 0.17980871349573135,
      "num_tokens": 20734809.0,
      "step": 8180
    },
    {
      "entropy": 6.218851041793823,
      "epoch": 0.944604731679169,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004922336708454279,
      "loss": 5.9741,
      "mean_token_accuracy": 0.17323816418647767,
      "num_tokens": 20747490.0,
      "step": 8185
    },
    {
      "entropy": 6.332640361785889,
      "epoch": 0.9451817657241778,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004922229192255903,
      "loss": 6.0511,
      "mean_token_accuracy": 0.17083201110363005,
      "num_tokens": 20759398.0,
      "step": 8190
    },
    {
      "entropy": 6.267602491378784,
      "epoch": 0.9457587997691864,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004922121602995678,
      "loss": 6.0851,
      "mean_token_accuracy": 0.17635809034109115,
      "num_tokens": 20773057.0,
      "step": 8195
    },
    {
      "entropy": 6.374321413040161,
      "epoch": 0.9463358338141951,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004922013940677221,
      "loss": 6.0769,
      "mean_token_accuracy": 0.16364725679159164,
      "num_tokens": 20786084.0,
      "step": 8200
    },
    {
      "entropy": 6.315356111526489,
      "epoch": 0.9469128678592037,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004921906205304155,
      "loss": 6.0102,
      "mean_token_accuracy": 0.17808901518583298,
      "num_tokens": 20798459.0,
      "step": 8205
    },
    {
      "entropy": 6.141188144683838,
      "epoch": 0.9474899019042123,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004921798396880101,
      "loss": 5.8281,
      "mean_token_accuracy": 0.18906592279672624,
      "num_tokens": 20811242.0,
      "step": 8210
    },
    {
      "entropy": 6.202764272689819,
      "epoch": 0.948066935949221,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004921690515408688,
      "loss": 6.067,
      "mean_token_accuracy": 0.17071353644132614,
      "num_tokens": 20824757.0,
      "step": 8215
    },
    {
      "entropy": 6.367983055114746,
      "epoch": 0.9486439699942296,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004921582560893543,
      "loss": 6.0756,
      "mean_token_accuracy": 0.16784874945878983,
      "num_tokens": 20836794.0,
      "step": 8220
    },
    {
      "entropy": 6.376722764968872,
      "epoch": 0.9492210040392384,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004921474533338297,
      "loss": 6.0104,
      "mean_token_accuracy": 0.1712158054113388,
      "num_tokens": 20849240.0,
      "step": 8225
    },
    {
      "entropy": 6.283626890182495,
      "epoch": 0.949798038084247,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004921366432746585,
      "loss": 6.0678,
      "mean_token_accuracy": 0.1724574536085129,
      "num_tokens": 20861362.0,
      "step": 8230
    },
    {
      "entropy": 6.29132342338562,
      "epoch": 0.9503750721292556,
      "grad_norm": 0.890625,
      "learning_rate": 0.000492125825912204,
      "loss": 6.0582,
      "mean_token_accuracy": 0.17874690741300583,
      "num_tokens": 20874925.0,
      "step": 8235
    },
    {
      "entropy": 6.3366481304168705,
      "epoch": 0.9509521061742643,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004921150012468302,
      "loss": 6.0554,
      "mean_token_accuracy": 0.17673903256654738,
      "num_tokens": 20888098.0,
      "step": 8240
    },
    {
      "entropy": 6.247784423828125,
      "epoch": 0.9515291402192729,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004921041692789013,
      "loss": 6.0584,
      "mean_token_accuracy": 0.17359829992055892,
      "num_tokens": 20901125.0,
      "step": 8245
    },
    {
      "entropy": 6.343004846572876,
      "epoch": 0.9521061742642816,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004920933300087813,
      "loss": 6.1112,
      "mean_token_accuracy": 0.1725790947675705,
      "num_tokens": 20914815.0,
      "step": 8250
    },
    {
      "entropy": 6.197666168212891,
      "epoch": 0.9526832083092902,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004920824834368353,
      "loss": 5.9648,
      "mean_token_accuracy": 0.18590367138385772,
      "num_tokens": 20928153.0,
      "step": 8255
    },
    {
      "entropy": 6.39654803276062,
      "epoch": 0.953260242354299,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004920716295634275,
      "loss": 6.1398,
      "mean_token_accuracy": 0.16140637397766114,
      "num_tokens": 20940601.0,
      "step": 8260
    },
    {
      "entropy": 6.348662281036377,
      "epoch": 0.9538372763993076,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920607683889235,
      "loss": 6.1483,
      "mean_token_accuracy": 0.17179345786571504,
      "num_tokens": 20953158.0,
      "step": 8265
    },
    {
      "entropy": 6.324293088912964,
      "epoch": 0.9544143104443162,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004920498999136882,
      "loss": 6.1126,
      "mean_token_accuracy": 0.1703899696469307,
      "num_tokens": 20966572.0,
      "step": 8270
    },
    {
      "entropy": 6.383297109603882,
      "epoch": 0.9549913444893249,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004920390241380874,
      "loss": 6.0931,
      "mean_token_accuracy": 0.17156865298748017,
      "num_tokens": 20978432.0,
      "step": 8275
    },
    {
      "entropy": 6.306451511383057,
      "epoch": 0.9555683785343335,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004920281410624868,
      "loss": 6.1197,
      "mean_token_accuracy": 0.1742856591939926,
      "num_tokens": 20991454.0,
      "step": 8280
    },
    {
      "entropy": 6.289869689941407,
      "epoch": 0.9561454125793422,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004920172506872525,
      "loss": 6.0127,
      "mean_token_accuracy": 0.1764762818813324,
      "num_tokens": 21004111.0,
      "step": 8285
    },
    {
      "entropy": 6.243640756607055,
      "epoch": 0.9567224466243508,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004920063530127508,
      "loss": 5.9837,
      "mean_token_accuracy": 0.1683119535446167,
      "num_tokens": 21017523.0,
      "step": 8290
    },
    {
      "entropy": 6.320230484008789,
      "epoch": 0.9572994806693594,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919954480393482,
      "loss": 6.0787,
      "mean_token_accuracy": 0.16999112963676452,
      "num_tokens": 21028723.0,
      "step": 8295
    },
    {
      "entropy": 6.296166467666626,
      "epoch": 0.9578765147143682,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004919845357674114,
      "loss": 6.0685,
      "mean_token_accuracy": 0.17331511229276658,
      "num_tokens": 21042205.0,
      "step": 8300
    },
    {
      "entropy": 6.308388900756836,
      "epoch": 0.9584535487593768,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004919736161973076,
      "loss": 6.011,
      "mean_token_accuracy": 0.16991052478551866,
      "num_tokens": 21053798.0,
      "step": 8305
    },
    {
      "entropy": 6.3203400611877445,
      "epoch": 0.9590305828043855,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491962689329404,
      "loss": 6.1504,
      "mean_token_accuracy": 0.1653846397995949,
      "num_tokens": 21068333.0,
      "step": 8310
    },
    {
      "entropy": 6.308880424499511,
      "epoch": 0.9596076168493941,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004919517551640681,
      "loss": 6.0126,
      "mean_token_accuracy": 0.1809369921684265,
      "num_tokens": 21079484.0,
      "step": 8315
    },
    {
      "entropy": 6.274625682830811,
      "epoch": 0.9601846508944027,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004919408137016677,
      "loss": 6.0472,
      "mean_token_accuracy": 0.17769970893859863,
      "num_tokens": 21090940.0,
      "step": 8320
    },
    {
      "entropy": 6.387520265579224,
      "epoch": 0.9607616849394114,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004919298649425708,
      "loss": 6.0893,
      "mean_token_accuracy": 0.16682010740041733,
      "num_tokens": 21104351.0,
      "step": 8325
    },
    {
      "entropy": 6.356888055801392,
      "epoch": 0.96133871898442,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004919189088871456,
      "loss": 6.0665,
      "mean_token_accuracy": 0.16898371130228043,
      "num_tokens": 21116272.0,
      "step": 8330
    },
    {
      "entropy": 6.327161407470703,
      "epoch": 0.9619157530294288,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004919079455357608,
      "loss": 6.1205,
      "mean_token_accuracy": 0.1648918569087982,
      "num_tokens": 21129448.0,
      "step": 8335
    },
    {
      "entropy": 6.334046459197998,
      "epoch": 0.9624927870744374,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004918969748887847,
      "loss": 6.1167,
      "mean_token_accuracy": 0.1638977497816086,
      "num_tokens": 21143668.0,
      "step": 8340
    },
    {
      "entropy": 6.35278091430664,
      "epoch": 0.9630698211194461,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004918859969465868,
      "loss": 6.1176,
      "mean_token_accuracy": 0.16923247128725052,
      "num_tokens": 21156662.0,
      "step": 8345
    },
    {
      "entropy": 6.312923526763916,
      "epoch": 0.9636468551644547,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004918750117095361,
      "loss": 6.0627,
      "mean_token_accuracy": 0.17304722517728804,
      "num_tokens": 21170271.0,
      "step": 8350
    },
    {
      "entropy": 6.278135585784912,
      "epoch": 0.9642238892094633,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004918640191780021,
      "loss": 6.1084,
      "mean_token_accuracy": 0.16698332279920577,
      "num_tokens": 21183878.0,
      "step": 8355
    },
    {
      "entropy": 6.361699485778809,
      "epoch": 0.964800923254472,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004918530193523546,
      "loss": 6.0209,
      "mean_token_accuracy": 0.176391077041626,
      "num_tokens": 21196088.0,
      "step": 8360
    },
    {
      "entropy": 6.3334063529968265,
      "epoch": 0.9653779572994806,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004918420122329634,
      "loss": 6.1499,
      "mean_token_accuracy": 0.16920819878578186,
      "num_tokens": 21208640.0,
      "step": 8365
    },
    {
      "entropy": 6.326233673095703,
      "epoch": 0.9659549913444894,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004918309978201989,
      "loss": 5.9552,
      "mean_token_accuracy": 0.18688852488994598,
      "num_tokens": 21221246.0,
      "step": 8370
    },
    {
      "entropy": 6.275239896774292,
      "epoch": 0.966532025389498,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004918199761144315,
      "loss": 6.0152,
      "mean_token_accuracy": 0.1766079217195511,
      "num_tokens": 21233834.0,
      "step": 8375
    },
    {
      "entropy": 6.3196817398071286,
      "epoch": 0.9671090594345066,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004918089471160318,
      "loss": 6.0489,
      "mean_token_accuracy": 0.17631532549858092,
      "num_tokens": 21247388.0,
      "step": 8380
    },
    {
      "entropy": 6.332034635543823,
      "epoch": 0.9676860934795153,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004917979108253709,
      "loss": 6.0613,
      "mean_token_accuracy": 0.16909884214401244,
      "num_tokens": 21259800.0,
      "step": 8385
    },
    {
      "entropy": 6.333093881607056,
      "epoch": 0.9682631275245239,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049178686724282,
      "loss": 6.0779,
      "mean_token_accuracy": 0.172726571559906,
      "num_tokens": 21271776.0,
      "step": 8390
    },
    {
      "entropy": 6.270829153060913,
      "epoch": 0.9688401615695326,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004917758163687506,
      "loss": 5.9621,
      "mean_token_accuracy": 0.18276852816343309,
      "num_tokens": 21285368.0,
      "step": 8395
    },
    {
      "entropy": 6.193240785598755,
      "epoch": 0.9694171956145412,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004917647582035341,
      "loss": 5.967,
      "mean_token_accuracy": 0.18024432510137559,
      "num_tokens": 21299077.0,
      "step": 8400
    },
    {
      "entropy": 6.326466274261475,
      "epoch": 0.96999422965955,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004917536927475428,
      "loss": 6.064,
      "mean_token_accuracy": 0.17469258606433868,
      "num_tokens": 21312379.0,
      "step": 8405
    },
    {
      "entropy": 6.2522468090057375,
      "epoch": 0.9705712637045586,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004917426200011486,
      "loss": 6.0416,
      "mean_token_accuracy": 0.1785849466919899,
      "num_tokens": 21325530.0,
      "step": 8410
    },
    {
      "entropy": 6.165889263153076,
      "epoch": 0.9711482977495672,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000491731539964724,
      "loss": 5.9184,
      "mean_token_accuracy": 0.186752113699913,
      "num_tokens": 21339200.0,
      "step": 8415
    },
    {
      "entropy": 6.362313652038575,
      "epoch": 0.9717253317945759,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004917204526386418,
      "loss": 5.984,
      "mean_token_accuracy": 0.18137681484222412,
      "num_tokens": 21351072.0,
      "step": 8420
    },
    {
      "entropy": 6.2359185218811035,
      "epoch": 0.9723023658395845,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004917093580232748,
      "loss": 6.0345,
      "mean_token_accuracy": 0.17593223750591278,
      "num_tokens": 21363691.0,
      "step": 8425
    },
    {
      "entropy": 6.281664705276489,
      "epoch": 0.9728793998845932,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004916982561189962,
      "loss": 6.0475,
      "mean_token_accuracy": 0.16599519401788712,
      "num_tokens": 21376896.0,
      "step": 8430
    },
    {
      "entropy": 6.335173654556274,
      "epoch": 0.9734564339296018,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916871469261794,
      "loss": 6.0794,
      "mean_token_accuracy": 0.17554948329925538,
      "num_tokens": 21388590.0,
      "step": 8435
    },
    {
      "entropy": 6.348338985443116,
      "epoch": 0.9740334679746105,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000491676030445198,
      "loss": 6.1219,
      "mean_token_accuracy": 0.1677918940782547,
      "num_tokens": 21401912.0,
      "step": 8440
    },
    {
      "entropy": 6.2882709980010985,
      "epoch": 0.9746105020196192,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004916649066764259,
      "loss": 6.0742,
      "mean_token_accuracy": 0.17222648561000825,
      "num_tokens": 21415472.0,
      "step": 8445
    },
    {
      "entropy": 6.2841771125793455,
      "epoch": 0.9751875360646278,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916537756202375,
      "loss": 6.038,
      "mean_token_accuracy": 0.17215794324874878,
      "num_tokens": 21426758.0,
      "step": 8450
    },
    {
      "entropy": 6.274518728256226,
      "epoch": 0.9757645701096365,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004916426372770069,
      "loss": 6.0206,
      "mean_token_accuracy": 0.18222465366125107,
      "num_tokens": 21439347.0,
      "step": 8455
    },
    {
      "entropy": 6.3539186954498295,
      "epoch": 0.9763416041546451,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916314916471087,
      "loss": 6.131,
      "mean_token_accuracy": 0.16433806121349334,
      "num_tokens": 21451359.0,
      "step": 8460
    },
    {
      "entropy": 6.407370901107788,
      "epoch": 0.9769186381996537,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004916203387309179,
      "loss": 6.1147,
      "mean_token_accuracy": 0.16536137759685515,
      "num_tokens": 21464926.0,
      "step": 8465
    },
    {
      "entropy": 6.312540197372437,
      "epoch": 0.9774956722446624,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004916091785288096,
      "loss": 6.0678,
      "mean_token_accuracy": 0.17171259671449662,
      "num_tokens": 21477234.0,
      "step": 8470
    },
    {
      "entropy": 6.203128433227539,
      "epoch": 0.9780727062896711,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004915980110411593,
      "loss": 5.929,
      "mean_token_accuracy": 0.18144314885139465,
      "num_tokens": 21488751.0,
      "step": 8475
    },
    {
      "entropy": 6.318572950363159,
      "epoch": 0.9786497403346798,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004915868362683425,
      "loss": 6.0491,
      "mean_token_accuracy": 0.17302265018224716,
      "num_tokens": 21500532.0,
      "step": 8480
    },
    {
      "entropy": 6.28663215637207,
      "epoch": 0.9792267743796884,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004915756542107349,
      "loss": 6.0377,
      "mean_token_accuracy": 0.17060410231351852,
      "num_tokens": 21513725.0,
      "step": 8485
    },
    {
      "entropy": 6.2164019584655765,
      "epoch": 0.979803808424697,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004915644648687127,
      "loss": 5.9456,
      "mean_token_accuracy": 0.1819642022252083,
      "num_tokens": 21526886.0,
      "step": 8490
    },
    {
      "entropy": 6.178979873657227,
      "epoch": 0.9803808424697057,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004915532682426524,
      "loss": 5.9224,
      "mean_token_accuracy": 0.18136737495660782,
      "num_tokens": 21539278.0,
      "step": 8495
    },
    {
      "entropy": 6.337500524520874,
      "epoch": 0.9809578765147143,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915420643329306,
      "loss": 6.151,
      "mean_token_accuracy": 0.17058085203170775,
      "num_tokens": 21551864.0,
      "step": 8500
    },
    {
      "entropy": 6.3601514339447025,
      "epoch": 0.981534910559723,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000491530853139924,
      "loss": 6.0833,
      "mean_token_accuracy": 0.16988922506570817,
      "num_tokens": 21564398.0,
      "step": 8505
    },
    {
      "entropy": 6.386910438537598,
      "epoch": 0.9821119446047317,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004915196346640095,
      "loss": 6.1054,
      "mean_token_accuracy": 0.16802452653646469,
      "num_tokens": 21575998.0,
      "step": 8510
    },
    {
      "entropy": 6.291682386398316,
      "epoch": 0.9826889786497404,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004915084089055648,
      "loss": 6.062,
      "mean_token_accuracy": 0.17027097046375275,
      "num_tokens": 21589243.0,
      "step": 8515
    },
    {
      "entropy": 6.201304721832275,
      "epoch": 0.983266012694749,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004914971758649673,
      "loss": 6.001,
      "mean_token_accuracy": 0.17721253782510757,
      "num_tokens": 21601229.0,
      "step": 8520
    },
    {
      "entropy": 6.328898525238037,
      "epoch": 0.9838430467397576,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004914859355425948,
      "loss": 6.039,
      "mean_token_accuracy": 0.17448335587978364,
      "num_tokens": 21613270.0,
      "step": 8525
    },
    {
      "entropy": 6.2084362506866455,
      "epoch": 0.9844200807847663,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004914746879388255,
      "loss": 5.9712,
      "mean_token_accuracy": 0.1804818794131279,
      "num_tokens": 21626463.0,
      "step": 8530
    },
    {
      "entropy": 6.28183126449585,
      "epoch": 0.9849971148297749,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004914634330540373,
      "loss": 6.0825,
      "mean_token_accuracy": 0.17816960513591767,
      "num_tokens": 21638172.0,
      "step": 8535
    },
    {
      "entropy": 6.308590936660766,
      "epoch": 0.9855741488747836,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004914521708886093,
      "loss": 6.0068,
      "mean_token_accuracy": 0.17926101088523866,
      "num_tokens": 21650015.0,
      "step": 8540
    },
    {
      "entropy": 6.3298423290252686,
      "epoch": 0.9861511829197923,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004914409014429201,
      "loss": 6.0853,
      "mean_token_accuracy": 0.17151839137077332,
      "num_tokens": 21662502.0,
      "step": 8545
    },
    {
      "entropy": 6.267358779907227,
      "epoch": 0.986728216964801,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004914296247173486,
      "loss": 5.9973,
      "mean_token_accuracy": 0.1838891997933388,
      "num_tokens": 21675691.0,
      "step": 8550
    },
    {
      "entropy": 6.28891806602478,
      "epoch": 0.9873052510098096,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004914183407122741,
      "loss": 6.0695,
      "mean_token_accuracy": 0.16591113954782485,
      "num_tokens": 21687904.0,
      "step": 8555
    },
    {
      "entropy": 6.3173764705657955,
      "epoch": 0.9878822850548182,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004914070494280763,
      "loss": 6.1318,
      "mean_token_accuracy": 0.16826688945293428,
      "num_tokens": 21700371.0,
      "step": 8560
    },
    {
      "entropy": 6.266348457336425,
      "epoch": 0.9884593190998269,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004913957508651349,
      "loss": 6.0203,
      "mean_token_accuracy": 0.18244925290346145,
      "num_tokens": 21715235.0,
      "step": 8565
    },
    {
      "entropy": 6.240448999404907,
      "epoch": 0.9890363531448355,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004913844450238299,
      "loss": 5.9989,
      "mean_token_accuracy": 0.17710728347301483,
      "num_tokens": 21727994.0,
      "step": 8570
    },
    {
      "entropy": 6.298191690444947,
      "epoch": 0.9896133871898442,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004913731319045416,
      "loss": 5.9943,
      "mean_token_accuracy": 0.1736165091395378,
      "num_tokens": 21739894.0,
      "step": 8575
    },
    {
      "entropy": 6.322504472732544,
      "epoch": 0.9901904212348529,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004913618115076505,
      "loss": 6.1325,
      "mean_token_accuracy": 0.17039846032857894,
      "num_tokens": 21752957.0,
      "step": 8580
    },
    {
      "entropy": 6.2790844440460205,
      "epoch": 0.9907674552798615,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004913504838335372,
      "loss": 6.0043,
      "mean_token_accuracy": 0.17115072757005692,
      "num_tokens": 21765547.0,
      "step": 8585
    },
    {
      "entropy": 6.294105577468872,
      "epoch": 0.9913444893248702,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004913391488825829,
      "loss": 6.1039,
      "mean_token_accuracy": 0.17850189357995988,
      "num_tokens": 21778817.0,
      "step": 8590
    },
    {
      "entropy": 6.261435699462891,
      "epoch": 0.9919215233698788,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004913278066551689,
      "loss": 5.9907,
      "mean_token_accuracy": 0.17613508254289628,
      "num_tokens": 21791002.0,
      "step": 8595
    },
    {
      "entropy": 6.29529767036438,
      "epoch": 0.9924985574148875,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004913164571516765,
      "loss": 6.08,
      "mean_token_accuracy": 0.17442494481801987,
      "num_tokens": 21803877.0,
      "step": 8600
    },
    {
      "entropy": 6.319041204452515,
      "epoch": 0.9930755914598961,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004913051003724876,
      "loss": 5.9999,
      "mean_token_accuracy": 0.17097305953502656,
      "num_tokens": 21815884.0,
      "step": 8605
    },
    {
      "entropy": 6.247610139846802,
      "epoch": 0.9936526255049047,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004912937363179839,
      "loss": 6.0451,
      "mean_token_accuracy": 0.17411776036024093,
      "num_tokens": 21826990.0,
      "step": 8610
    },
    {
      "entropy": 6.261379289627075,
      "epoch": 0.9942296595499135,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491282364988548,
      "loss": 6.019,
      "mean_token_accuracy": 0.17805099189281465,
      "num_tokens": 21838835.0,
      "step": 8615
    },
    {
      "entropy": 6.382070112228393,
      "epoch": 0.9948066935949221,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004912709863845621,
      "loss": 6.134,
      "mean_token_accuracy": 0.17056389898061752,
      "num_tokens": 21850924.0,
      "step": 8620
    },
    {
      "entropy": 6.325075387954712,
      "epoch": 0.9953837276399308,
      "grad_norm": 0.875,
      "learning_rate": 0.000491259600506409,
      "loss": 6.0986,
      "mean_token_accuracy": 0.1749351978302002,
      "num_tokens": 21862720.0,
      "step": 8625
    },
    {
      "entropy": 6.298249101638794,
      "epoch": 0.9959607616849394,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004912482073544716,
      "loss": 6.0499,
      "mean_token_accuracy": 0.17579894959926606,
      "num_tokens": 21874655.0,
      "step": 8630
    },
    {
      "entropy": 6.1955633640289305,
      "epoch": 0.996537795729948,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004912368069291333,
      "loss": 6.0124,
      "mean_token_accuracy": 0.17971490770578386,
      "num_tokens": 21886911.0,
      "step": 8635
    },
    {
      "entropy": 6.2078932762146,
      "epoch": 0.9971148297749567,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004912253992307773,
      "loss": 5.9897,
      "mean_token_accuracy": 0.18102246075868605,
      "num_tokens": 21899614.0,
      "step": 8640
    },
    {
      "entropy": 6.3586358547210695,
      "epoch": 0.9976918638199653,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004912139842597875,
      "loss": 6.0146,
      "mean_token_accuracy": 0.17147439420223237,
      "num_tokens": 21912756.0,
      "step": 8645
    },
    {
      "entropy": 6.226405668258667,
      "epoch": 0.9982688978649741,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004912025620165477,
      "loss": 6.0197,
      "mean_token_accuracy": 0.16808625310659409,
      "num_tokens": 21924911.0,
      "step": 8650
    },
    {
      "entropy": 6.230767774581909,
      "epoch": 0.9988459319099827,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004911911325014421,
      "loss": 5.9436,
      "mean_token_accuracy": 0.17855229377746581,
      "num_tokens": 21938204.0,
      "step": 8655
    },
    {
      "entropy": 6.353689670562744,
      "epoch": 0.9994229659549914,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004911796957148552,
      "loss": 6.041,
      "mean_token_accuracy": 0.1740615800023079,
      "num_tokens": 21951285.0,
      "step": 8660
    },
    {
      "entropy": 6.28210768699646,
      "epoch": 1.0,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004911682516571715,
      "loss": 6.0783,
      "mean_token_accuracy": 0.17323559373617173,
      "num_tokens": 21963117.0,
      "step": 8665
    },
    {
      "entropy": 6.305393648147583,
      "epoch": 1.0005770340450086,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004911568003287761,
      "loss": 5.965,
      "mean_token_accuracy": 0.1732805535197258,
      "num_tokens": 21976103.0,
      "step": 8670
    },
    {
      "entropy": 6.221843481063843,
      "epoch": 1.0011540680900173,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004911453417300541,
      "loss": 5.8612,
      "mean_token_accuracy": 0.19035064578056335,
      "num_tokens": 21987905.0,
      "step": 8675
    },
    {
      "entropy": 6.191161012649536,
      "epoch": 1.001731102135026,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004911338758613909,
      "loss": 5.9456,
      "mean_token_accuracy": 0.17842523753643036,
      "num_tokens": 22000872.0,
      "step": 8680
    },
    {
      "entropy": 6.293917179107666,
      "epoch": 1.0023081361800346,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004911224027231722,
      "loss": 5.9699,
      "mean_token_accuracy": 0.1744038790464401,
      "num_tokens": 22012725.0,
      "step": 8685
    },
    {
      "entropy": 6.259350252151489,
      "epoch": 1.0028851702250432,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004911109223157838,
      "loss": 5.904,
      "mean_token_accuracy": 0.1862815409898758,
      "num_tokens": 22024802.0,
      "step": 8690
    },
    {
      "entropy": 6.319523811340332,
      "epoch": 1.0034622042700518,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004910994346396118,
      "loss": 5.981,
      "mean_token_accuracy": 0.17476661801338195,
      "num_tokens": 22037481.0,
      "step": 8695
    },
    {
      "entropy": 6.2829381942749025,
      "epoch": 1.0040392383150605,
      "grad_norm": 0.875,
      "learning_rate": 0.0004910879396950427,
      "loss": 5.9693,
      "mean_token_accuracy": 0.17762275487184526,
      "num_tokens": 22049387.0,
      "step": 8700
    },
    {
      "entropy": 6.314607334136963,
      "epoch": 1.0046162723600693,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000491076437482463,
      "loss": 5.9782,
      "mean_token_accuracy": 0.16941831558942794,
      "num_tokens": 22061323.0,
      "step": 8705
    },
    {
      "entropy": 6.300982809066772,
      "epoch": 1.005193306405078,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004910649280022599,
      "loss": 5.9084,
      "mean_token_accuracy": 0.18293041437864305,
      "num_tokens": 22073163.0,
      "step": 8710
    },
    {
      "entropy": 6.185075283050537,
      "epoch": 1.0057703404500866,
      "grad_norm": 0.875,
      "learning_rate": 0.0004910534112548201,
      "loss": 5.9069,
      "mean_token_accuracy": 0.1787968948483467,
      "num_tokens": 22085583.0,
      "step": 8715
    },
    {
      "entropy": 6.318300724029541,
      "epoch": 1.0063473744950953,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004910418872405312,
      "loss": 6.0127,
      "mean_token_accuracy": 0.17371083050966263,
      "num_tokens": 22098241.0,
      "step": 8720
    },
    {
      "entropy": 6.292492246627807,
      "epoch": 1.006924408540104,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004910303559597806,
      "loss": 5.8986,
      "mean_token_accuracy": 0.17350029051303864,
      "num_tokens": 22111626.0,
      "step": 8725
    },
    {
      "entropy": 6.224344396591187,
      "epoch": 1.0075014425851125,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004910188174129564,
      "loss": 5.926,
      "mean_token_accuracy": 0.17827400118112563,
      "num_tokens": 22124653.0,
      "step": 8730
    },
    {
      "entropy": 6.3021481990814205,
      "epoch": 1.0080784766301212,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004910072716004466,
      "loss": 5.9503,
      "mean_token_accuracy": 0.18166403472423553,
      "num_tokens": 22137154.0,
      "step": 8735
    },
    {
      "entropy": 6.329882717132568,
      "epoch": 1.0086555106751298,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004909957185226394,
      "loss": 6.0546,
      "mean_token_accuracy": 0.16671145260334014,
      "num_tokens": 22149374.0,
      "step": 8740
    },
    {
      "entropy": 6.277902889251709,
      "epoch": 1.0092325447201385,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004909841581799236,
      "loss": 6.0218,
      "mean_token_accuracy": 0.17590236365795137,
      "num_tokens": 22161988.0,
      "step": 8745
    },
    {
      "entropy": 6.3510805606842045,
      "epoch": 1.009809578765147,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004909725905726879,
      "loss": 6.0754,
      "mean_token_accuracy": 0.1719086691737175,
      "num_tokens": 22173490.0,
      "step": 8750
    },
    {
      "entropy": 6.277780532836914,
      "epoch": 1.0103866128101557,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004909610157013214,
      "loss": 6.0044,
      "mean_token_accuracy": 0.1760912850499153,
      "num_tokens": 22185065.0,
      "step": 8755
    },
    {
      "entropy": 6.2296899318695065,
      "epoch": 1.0109636468551644,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004909494335662134,
      "loss": 5.942,
      "mean_token_accuracy": 0.18478337675333023,
      "num_tokens": 22198325.0,
      "step": 8760
    },
    {
      "entropy": 6.181908226013183,
      "epoch": 1.011540680900173,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004909378441677535,
      "loss": 5.8771,
      "mean_token_accuracy": 0.18725387752056122,
      "num_tokens": 22211827.0,
      "step": 8765
    },
    {
      "entropy": 6.248178100585937,
      "epoch": 1.0121177149451817,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004909262475063314,
      "loss": 5.9292,
      "mean_token_accuracy": 0.179417784512043,
      "num_tokens": 22223411.0,
      "step": 8770
    },
    {
      "entropy": 6.233440399169922,
      "epoch": 1.0126947489901905,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004909146435823373,
      "loss": 5.9381,
      "mean_token_accuracy": 0.17608700394630433,
      "num_tokens": 22237752.0,
      "step": 8775
    },
    {
      "entropy": 6.301976728439331,
      "epoch": 1.0132717830351992,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004909030323961613,
      "loss": 5.9904,
      "mean_token_accuracy": 0.17447181046009064,
      "num_tokens": 22249778.0,
      "step": 8780
    },
    {
      "entropy": 6.315885210037232,
      "epoch": 1.0138488170802078,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004908914139481942,
      "loss": 5.9566,
      "mean_token_accuracy": 0.17732033580541612,
      "num_tokens": 22262514.0,
      "step": 8785
    },
    {
      "entropy": 6.260607624053955,
      "epoch": 1.0144258511252164,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004908797882388265,
      "loss": 5.93,
      "mean_token_accuracy": 0.1733621686697006,
      "num_tokens": 22273662.0,
      "step": 8790
    },
    {
      "entropy": 6.292414379119873,
      "epoch": 1.015002885170225,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004908681552684495,
      "loss": 6.0007,
      "mean_token_accuracy": 0.17214885503053665,
      "num_tokens": 22286213.0,
      "step": 8795
    },
    {
      "entropy": 6.229160594940185,
      "epoch": 1.0155799192152337,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004908565150374542,
      "loss": 5.8057,
      "mean_token_accuracy": 0.18783130198717118,
      "num_tokens": 22299191.0,
      "step": 8800
    },
    {
      "entropy": 6.194849491119385,
      "epoch": 1.0161569532602424,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004908448675462323,
      "loss": 5.9018,
      "mean_token_accuracy": 0.18518732339143754,
      "num_tokens": 22312047.0,
      "step": 8805
    },
    {
      "entropy": 6.305113697052002,
      "epoch": 1.016733987305251,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004908332127951756,
      "loss": 6.0331,
      "mean_token_accuracy": 0.16996526718139648,
      "num_tokens": 22326069.0,
      "step": 8810
    },
    {
      "entropy": 6.318189334869385,
      "epoch": 1.0173110213502596,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004908215507846757,
      "loss": 5.971,
      "mean_token_accuracy": 0.17298663407564163,
      "num_tokens": 22338593.0,
      "step": 8815
    },
    {
      "entropy": 6.297438526153565,
      "epoch": 1.0178880553952683,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004908098815151254,
      "loss": 5.9302,
      "mean_token_accuracy": 0.18209951519966125,
      "num_tokens": 22352286.0,
      "step": 8820
    },
    {
      "entropy": 6.278217077255249,
      "epoch": 1.018465089440277,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004907982049869168,
      "loss": 5.9273,
      "mean_token_accuracy": 0.17886870801448823,
      "num_tokens": 22364270.0,
      "step": 8825
    },
    {
      "entropy": 6.247045516967773,
      "epoch": 1.0190421234852856,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004907865212004428,
      "loss": 5.9627,
      "mean_token_accuracy": 0.1733900472521782,
      "num_tokens": 22377799.0,
      "step": 8830
    },
    {
      "entropy": 6.289280033111572,
      "epoch": 1.0196191575302942,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004907748301560963,
      "loss": 5.9769,
      "mean_token_accuracy": 0.17523756325244905,
      "num_tokens": 22389913.0,
      "step": 8835
    },
    {
      "entropy": 6.289673471450806,
      "epoch": 1.0201961915753028,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004907631318542707,
      "loss": 5.9927,
      "mean_token_accuracy": 0.172393099963665,
      "num_tokens": 22401545.0,
      "step": 8840
    },
    {
      "entropy": 6.220170783996582,
      "epoch": 1.0207732256203117,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004907514262953594,
      "loss": 5.8904,
      "mean_token_accuracy": 0.17623564302921296,
      "num_tokens": 22414084.0,
      "step": 8845
    },
    {
      "entropy": 6.220952892303467,
      "epoch": 1.0213502596653203,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004907397134797559,
      "loss": 5.8784,
      "mean_token_accuracy": 0.18220734894275664,
      "num_tokens": 22427406.0,
      "step": 8850
    },
    {
      "entropy": 6.33605432510376,
      "epoch": 1.021927293710329,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004907279934078543,
      "loss": 5.9781,
      "mean_token_accuracy": 0.17461190819740297,
      "num_tokens": 22441118.0,
      "step": 8855
    },
    {
      "entropy": 6.343763256072998,
      "epoch": 1.0225043277553376,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004907162660800488,
      "loss": 6.0549,
      "mean_token_accuracy": 0.17441704422235488,
      "num_tokens": 22453799.0,
      "step": 8860
    },
    {
      "entropy": 6.295844173431396,
      "epoch": 1.0230813618003463,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004907045314967338,
      "loss": 6.0,
      "mean_token_accuracy": 0.17054615914821625,
      "num_tokens": 22466573.0,
      "step": 8865
    },
    {
      "entropy": 6.1969630241394045,
      "epoch": 1.023658395845355,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004906927896583041,
      "loss": 5.8697,
      "mean_token_accuracy": 0.1860299900174141,
      "num_tokens": 22478968.0,
      "step": 8870
    },
    {
      "entropy": 6.268518400192261,
      "epoch": 1.0242354298903635,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004906810405651546,
      "loss": 5.9102,
      "mean_token_accuracy": 0.1873992756009102,
      "num_tokens": 22490669.0,
      "step": 8875
    },
    {
      "entropy": 6.278630065917969,
      "epoch": 1.0248124639353722,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004906692842176803,
      "loss": 5.9661,
      "mean_token_accuracy": 0.17420146316289903,
      "num_tokens": 22502235.0,
      "step": 8880
    },
    {
      "entropy": 6.2680768966674805,
      "epoch": 1.0253894979803808,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004906575206162769,
      "loss": 5.9422,
      "mean_token_accuracy": 0.17769977003335952,
      "num_tokens": 22514535.0,
      "step": 8885
    },
    {
      "entropy": 6.268265438079834,
      "epoch": 1.0259665320253895,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004906457497613399,
      "loss": 5.9417,
      "mean_token_accuracy": 0.17981360256671905,
      "num_tokens": 22527773.0,
      "step": 8890
    },
    {
      "entropy": 6.209779596328735,
      "epoch": 1.026543566070398,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004906339716532651,
      "loss": 5.9262,
      "mean_token_accuracy": 0.18416949212551117,
      "num_tokens": 22541853.0,
      "step": 8895
    },
    {
      "entropy": 6.370106840133667,
      "epoch": 1.0271206001154067,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004906221862924488,
      "loss": 6.015,
      "mean_token_accuracy": 0.1671313539147377,
      "num_tokens": 22554473.0,
      "step": 8900
    },
    {
      "entropy": 6.311019086837769,
      "epoch": 1.0276976341604154,
      "grad_norm": 0.875,
      "learning_rate": 0.0004906103936792875,
      "loss": 5.9845,
      "mean_token_accuracy": 0.17602384239435195,
      "num_tokens": 22567107.0,
      "step": 8905
    },
    {
      "entropy": 6.276227331161499,
      "epoch": 1.028274668205424,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004905985938141777,
      "loss": 6.1394,
      "mean_token_accuracy": 0.16377924382686615,
      "num_tokens": 22579806.0,
      "step": 8910
    },
    {
      "entropy": 6.291239166259766,
      "epoch": 1.0288517022504329,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004905867866975163,
      "loss": 5.9196,
      "mean_token_accuracy": 0.17613937705755234,
      "num_tokens": 22593460.0,
      "step": 8915
    },
    {
      "entropy": 6.291932392120361,
      "epoch": 1.0294287362954415,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004905749723297003,
      "loss": 5.9951,
      "mean_token_accuracy": 0.17714100629091262,
      "num_tokens": 22607113.0,
      "step": 8920
    },
    {
      "entropy": 6.292923641204834,
      "epoch": 1.0300057703404502,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004905631507111271,
      "loss": 5.9214,
      "mean_token_accuracy": 0.1798023298382759,
      "num_tokens": 22619749.0,
      "step": 8925
    },
    {
      "entropy": 6.200394964218139,
      "epoch": 1.0305828043854588,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004905513218421946,
      "loss": 5.8833,
      "mean_token_accuracy": 0.18466779589653015,
      "num_tokens": 22632584.0,
      "step": 8930
    },
    {
      "entropy": 6.264637613296509,
      "epoch": 1.0311598384304674,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004905394857233004,
      "loss": 5.9781,
      "mean_token_accuracy": 0.17751743346452714,
      "num_tokens": 22645181.0,
      "step": 8935
    },
    {
      "entropy": 6.333766603469849,
      "epoch": 1.031736872475476,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004905276423548426,
      "loss": 5.992,
      "mean_token_accuracy": 0.176498681306839,
      "num_tokens": 22657666.0,
      "step": 8940
    },
    {
      "entropy": 6.2933393001556395,
      "epoch": 1.0323139065204847,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004905157917372197,
      "loss": 6.0347,
      "mean_token_accuracy": 0.1707567095756531,
      "num_tokens": 22670404.0,
      "step": 8945
    },
    {
      "entropy": 6.2833434581756595,
      "epoch": 1.0328909405654934,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004905039338708302,
      "loss": 6.0235,
      "mean_token_accuracy": 0.1708123952150345,
      "num_tokens": 22683558.0,
      "step": 8950
    },
    {
      "entropy": 6.310410118103027,
      "epoch": 1.033467974610502,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004904920687560728,
      "loss": 5.8975,
      "mean_token_accuracy": 0.18546259850263597,
      "num_tokens": 22694907.0,
      "step": 8955
    },
    {
      "entropy": 6.3052013397216795,
      "epoch": 1.0340450086555106,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004904801963933469,
      "loss": 6.0802,
      "mean_token_accuracy": 0.1663333311676979,
      "num_tokens": 22707447.0,
      "step": 8960
    },
    {
      "entropy": 6.238166952133179,
      "epoch": 1.0346220427005193,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004904683167830515,
      "loss": 5.9448,
      "mean_token_accuracy": 0.17683406323194503,
      "num_tokens": 22719150.0,
      "step": 8965
    },
    {
      "entropy": 6.339394569396973,
      "epoch": 1.035199076745528,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004904564299255862,
      "loss": 6.0261,
      "mean_token_accuracy": 0.16863227635622025,
      "num_tokens": 22732489.0,
      "step": 8970
    },
    {
      "entropy": 6.228077030181884,
      "epoch": 1.0357761107905366,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904445358213511,
      "loss": 5.8748,
      "mean_token_accuracy": 0.17816764116287231,
      "num_tokens": 22745543.0,
      "step": 8975
    },
    {
      "entropy": 6.313041257858276,
      "epoch": 1.0363531448355452,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004904326344707461,
      "loss": 5.978,
      "mean_token_accuracy": 0.17733895182609558,
      "num_tokens": 22757643.0,
      "step": 8980
    },
    {
      "entropy": 6.339817762374878,
      "epoch": 1.036930178880554,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004904207258741712,
      "loss": 5.9578,
      "mean_token_accuracy": 0.17927060276269913,
      "num_tokens": 22770707.0,
      "step": 8985
    },
    {
      "entropy": 6.245674991607666,
      "epoch": 1.0375072129255627,
      "grad_norm": 0.875,
      "learning_rate": 0.0004904088100320274,
      "loss": 5.9834,
      "mean_token_accuracy": 0.17738377153873444,
      "num_tokens": 22782630.0,
      "step": 8990
    },
    {
      "entropy": 6.296307802200317,
      "epoch": 1.0380842469705713,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004903968869447151,
      "loss": 6.0469,
      "mean_token_accuracy": 0.17303405553102494,
      "num_tokens": 22795693.0,
      "step": 8995
    },
    {
      "entropy": 6.346906518936157,
      "epoch": 1.03866128101558,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004903849566126356,
      "loss": 6.0456,
      "mean_token_accuracy": 0.17103580385446548,
      "num_tokens": 22807715.0,
      "step": 9000
    },
    {
      "epoch": 1.03866128101558,
      "eval_entropy": 6.095231879526897,
      "eval_loss": 6.064687252044678,
      "eval_mean_token_accuracy": 0.17800477821192418,
      "eval_num_tokens": 22807715.0,
      "eval_runtime": 17.5587,
      "eval_samples_per_second": 1602.399,
      "eval_steps_per_second": 200.3,
      "step": 9000
    },
    {
      "entropy": 6.246942234039307,
      "epoch": 1.0392383150605886,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004903730190361902,
      "loss": 5.9425,
      "mean_token_accuracy": 0.17914803624153136,
      "num_tokens": 22820133.0,
      "step": 9005
    },
    {
      "entropy": 6.198151731491089,
      "epoch": 1.0398153491055973,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049036107421578,
      "loss": 5.9234,
      "mean_token_accuracy": 0.18442054241895675,
      "num_tokens": 22833928.0,
      "step": 9010
    },
    {
      "entropy": 6.265505409240722,
      "epoch": 1.040392383150606,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004903491221518073,
      "loss": 5.856,
      "mean_token_accuracy": 0.1893016591668129,
      "num_tokens": 22847883.0,
      "step": 9015
    },
    {
      "entropy": 6.289188051223755,
      "epoch": 1.0409694171956145,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004903371628446737,
      "loss": 5.9931,
      "mean_token_accuracy": 0.17363590449094773,
      "num_tokens": 22861540.0,
      "step": 9020
    },
    {
      "entropy": 6.2801597118377686,
      "epoch": 1.0415464512406232,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004903251962947817,
      "loss": 5.9459,
      "mean_token_accuracy": 0.17561002373695372,
      "num_tokens": 22874778.0,
      "step": 9025
    },
    {
      "entropy": 6.28412446975708,
      "epoch": 1.0421234852856318,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004903132225025335,
      "loss": 5.9995,
      "mean_token_accuracy": 0.18134995996952058,
      "num_tokens": 22886868.0,
      "step": 9030
    },
    {
      "entropy": 6.261662769317627,
      "epoch": 1.0427005193306405,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004903012414683322,
      "loss": 5.9428,
      "mean_token_accuracy": 0.17143700122833253,
      "num_tokens": 22899549.0,
      "step": 9035
    },
    {
      "entropy": 6.270760011672974,
      "epoch": 1.043277553375649,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004902892531925805,
      "loss": 5.9291,
      "mean_token_accuracy": 0.18446052074432373,
      "num_tokens": 22912301.0,
      "step": 9040
    },
    {
      "entropy": 6.262466096878052,
      "epoch": 1.0438545874206577,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004902772576756818,
      "loss": 5.967,
      "mean_token_accuracy": 0.18024677634239197,
      "num_tokens": 22924784.0,
      "step": 9045
    },
    {
      "entropy": 6.320473146438599,
      "epoch": 1.0444316214656664,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004902652549180394,
      "loss": 5.9899,
      "mean_token_accuracy": 0.17371919751167297,
      "num_tokens": 22937095.0,
      "step": 9050
    },
    {
      "entropy": 6.2498283863067625,
      "epoch": 1.0450086555106752,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004902532449200571,
      "loss": 5.8967,
      "mean_token_accuracy": 0.18342950493097304,
      "num_tokens": 22951385.0,
      "step": 9055
    },
    {
      "entropy": 6.23943395614624,
      "epoch": 1.0455856895556839,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004902412276821386,
      "loss": 5.974,
      "mean_token_accuracy": 0.17356501370668412,
      "num_tokens": 22965228.0,
      "step": 9060
    },
    {
      "entropy": 6.268825960159302,
      "epoch": 1.0461627236006925,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004902292032046887,
      "loss": 5.9656,
      "mean_token_accuracy": 0.1805337980389595,
      "num_tokens": 22978376.0,
      "step": 9065
    },
    {
      "entropy": 6.259447956085205,
      "epoch": 1.0467397576457012,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004902171714881112,
      "loss": 5.9711,
      "mean_token_accuracy": 0.17842662781476976,
      "num_tokens": 22991572.0,
      "step": 9070
    },
    {
      "entropy": 6.283546304702758,
      "epoch": 1.0473167916907098,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004902051325328112,
      "loss": 5.9103,
      "mean_token_accuracy": 0.18165430724620818,
      "num_tokens": 23004001.0,
      "step": 9075
    },
    {
      "entropy": 6.168697500228882,
      "epoch": 1.0478938257357184,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004901930863391934,
      "loss": 5.8921,
      "mean_token_accuracy": 0.18466428816318511,
      "num_tokens": 23016383.0,
      "step": 9080
    },
    {
      "entropy": 6.283572483062744,
      "epoch": 1.048470859780727,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000490181032907663,
      "loss": 5.9918,
      "mean_token_accuracy": 0.1763824313879013,
      "num_tokens": 23027836.0,
      "step": 9085
    },
    {
      "entropy": 6.268266773223877,
      "epoch": 1.0490478938257357,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004901689722386255,
      "loss": 5.9519,
      "mean_token_accuracy": 0.17849545925855637,
      "num_tokens": 23040083.0,
      "step": 9090
    },
    {
      "entropy": 6.206085062026977,
      "epoch": 1.0496249278707444,
      "grad_norm": 0.875,
      "learning_rate": 0.0004901569043324864,
      "loss": 5.956,
      "mean_token_accuracy": 0.17446473687887193,
      "num_tokens": 23051232.0,
      "step": 9095
    },
    {
      "entropy": 6.342846012115478,
      "epoch": 1.050201961915753,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004901448291896518,
      "loss": 6.0435,
      "mean_token_accuracy": 0.16940606385469437,
      "num_tokens": 23063355.0,
      "step": 9100
    },
    {
      "entropy": 6.153590154647827,
      "epoch": 1.0507789959607616,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004901327468105277,
      "loss": 5.7403,
      "mean_token_accuracy": 0.19071755558252335,
      "num_tokens": 23076268.0,
      "step": 9105
    },
    {
      "entropy": 6.1897430419921875,
      "epoch": 1.0513560300057703,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004901206571955205,
      "loss": 5.8845,
      "mean_token_accuracy": 0.17913285046815872,
      "num_tokens": 23088753.0,
      "step": 9110
    },
    {
      "entropy": 6.238257980346679,
      "epoch": 1.051933064050779,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004901085603450369,
      "loss": 5.8386,
      "mean_token_accuracy": 0.1816864624619484,
      "num_tokens": 23100452.0,
      "step": 9115
    },
    {
      "entropy": 6.306304836273194,
      "epoch": 1.0525100980957875,
      "grad_norm": 0.875,
      "learning_rate": 0.0004900964562594838,
      "loss": 5.9569,
      "mean_token_accuracy": 0.17561482191085814,
      "num_tokens": 23113368.0,
      "step": 9120
    },
    {
      "entropy": 6.187850761413574,
      "epoch": 1.0530871321407964,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000490084344939268,
      "loss": 5.8885,
      "mean_token_accuracy": 0.18586096316576003,
      "num_tokens": 23125940.0,
      "step": 9125
    },
    {
      "entropy": 6.194670629501343,
      "epoch": 1.053664166185805,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004900722263847973,
      "loss": 5.9313,
      "mean_token_accuracy": 0.1809296876192093,
      "num_tokens": 23137282.0,
      "step": 9130
    },
    {
      "entropy": 6.347890901565552,
      "epoch": 1.0542412002308137,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004900601005964791,
      "loss": 6.0232,
      "mean_token_accuracy": 0.17643271684646605,
      "num_tokens": 23149569.0,
      "step": 9135
    },
    {
      "entropy": 6.258325386047363,
      "epoch": 1.0548182342758223,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004900479675747212,
      "loss": 5.9952,
      "mean_token_accuracy": 0.17973504662513734,
      "num_tokens": 23163290.0,
      "step": 9140
    },
    {
      "entropy": 6.257934617996216,
      "epoch": 1.055395268320831,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004900358273199316,
      "loss": 5.9462,
      "mean_token_accuracy": 0.18773574084043504,
      "num_tokens": 23175092.0,
      "step": 9145
    },
    {
      "entropy": 6.276286506652832,
      "epoch": 1.0559723023658396,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000490023679832519,
      "loss": 5.9008,
      "mean_token_accuracy": 0.18411117047071457,
      "num_tokens": 23189064.0,
      "step": 9150
    },
    {
      "entropy": 6.260979461669922,
      "epoch": 1.0565493364108482,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004900115251128916,
      "loss": 5.9751,
      "mean_token_accuracy": 0.17391649186611174,
      "num_tokens": 23202619.0,
      "step": 9155
    },
    {
      "entropy": 6.218464660644531,
      "epoch": 1.057126370455857,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004899993631614583,
      "loss": 5.8705,
      "mean_token_accuracy": 0.18403320908546447,
      "num_tokens": 23214575.0,
      "step": 9160
    },
    {
      "entropy": 6.314262628555298,
      "epoch": 1.0577034045008655,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004899871939786283,
      "loss": 6.0101,
      "mean_token_accuracy": 0.1684699386358261,
      "num_tokens": 23226526.0,
      "step": 9165
    },
    {
      "entropy": 6.226275396347046,
      "epoch": 1.0582804385458742,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004899750175648107,
      "loss": 5.9576,
      "mean_token_accuracy": 0.1776348978281021,
      "num_tokens": 23239942.0,
      "step": 9170
    },
    {
      "entropy": 6.232362556457519,
      "epoch": 1.0588574725908828,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004899628339204154,
      "loss": 5.939,
      "mean_token_accuracy": 0.1749611333012581,
      "num_tokens": 23252798.0,
      "step": 9175
    },
    {
      "entropy": 6.219648933410644,
      "epoch": 1.0594345066358914,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004899506430458518,
      "loss": 5.888,
      "mean_token_accuracy": 0.18320101052522658,
      "num_tokens": 23266195.0,
      "step": 9180
    },
    {
      "entropy": 6.185371923446655,
      "epoch": 1.0600115406809,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004899384449415302,
      "loss": 5.8427,
      "mean_token_accuracy": 0.19028060883283615,
      "num_tokens": 23279057.0,
      "step": 9185
    },
    {
      "entropy": 6.347293090820313,
      "epoch": 1.0605885747259087,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004899262396078606,
      "loss": 6.0036,
      "mean_token_accuracy": 0.1756329968571663,
      "num_tokens": 23292011.0,
      "step": 9190
    },
    {
      "entropy": 6.275160551071167,
      "epoch": 1.0611656087709176,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004899140270452539,
      "loss": 5.8963,
      "mean_token_accuracy": 0.18479495793581008,
      "num_tokens": 23303719.0,
      "step": 9195
    },
    {
      "entropy": 6.241651916503907,
      "epoch": 1.0617426428159262,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004899018072541204,
      "loss": 5.9823,
      "mean_token_accuracy": 0.17531984448432922,
      "num_tokens": 23317291.0,
      "step": 9200
    },
    {
      "entropy": 6.2734925746917725,
      "epoch": 1.0623196768609349,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004898895802348715,
      "loss": 5.943,
      "mean_token_accuracy": 0.1795913726091385,
      "num_tokens": 23329546.0,
      "step": 9205
    },
    {
      "entropy": 6.225551462173462,
      "epoch": 1.0628967109059435,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004898773459879183,
      "loss": 5.8881,
      "mean_token_accuracy": 0.1821628764271736,
      "num_tokens": 23342171.0,
      "step": 9210
    },
    {
      "entropy": 6.237482500076294,
      "epoch": 1.0634737449509521,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898651045136724,
      "loss": 5.9468,
      "mean_token_accuracy": 0.1700466424226761,
      "num_tokens": 23355400.0,
      "step": 9215
    },
    {
      "entropy": 6.307815933227539,
      "epoch": 1.0640507789959608,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004898528558125453,
      "loss": 5.9052,
      "mean_token_accuracy": 0.18111446797847747,
      "num_tokens": 23368011.0,
      "step": 9220
    },
    {
      "entropy": 6.21876630783081,
      "epoch": 1.0646278130409694,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004898405998849492,
      "loss": 5.9666,
      "mean_token_accuracy": 0.1833662658929825,
      "num_tokens": 23380938.0,
      "step": 9225
    },
    {
      "entropy": 6.289742565155029,
      "epoch": 1.065204847085978,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898283367312962,
      "loss": 6.0358,
      "mean_token_accuracy": 0.17567623853683473,
      "num_tokens": 23392880.0,
      "step": 9230
    },
    {
      "entropy": 6.310242128372193,
      "epoch": 1.0657818811309867,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004898160663519988,
      "loss": 5.9322,
      "mean_token_accuracy": 0.17994374781847,
      "num_tokens": 23407809.0,
      "step": 9235
    },
    {
      "entropy": 6.303219223022461,
      "epoch": 1.0663589151759953,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004898037887474697,
      "loss": 5.9794,
      "mean_token_accuracy": 0.175724858045578,
      "num_tokens": 23419908.0,
      "step": 9240
    },
    {
      "entropy": 6.304628515243531,
      "epoch": 1.066935949221004,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004897915039181219,
      "loss": 6.0448,
      "mean_token_accuracy": 0.1720645770430565,
      "num_tokens": 23431743.0,
      "step": 9245
    },
    {
      "entropy": 6.353899431228638,
      "epoch": 1.0675129832660126,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004897792118643685,
      "loss": 5.9316,
      "mean_token_accuracy": 0.17956120222806932,
      "num_tokens": 23444234.0,
      "step": 9250
    },
    {
      "entropy": 6.237563753128052,
      "epoch": 1.0680900173110213,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004897669125866231,
      "loss": 5.8941,
      "mean_token_accuracy": 0.18154074549674987,
      "num_tokens": 23457393.0,
      "step": 9255
    },
    {
      "entropy": 6.1858323097229,
      "epoch": 1.06866705135603,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004897546060852993,
      "loss": 5.983,
      "mean_token_accuracy": 0.1752806708216667,
      "num_tokens": 23471520.0,
      "step": 9260
    },
    {
      "entropy": 6.276487874984741,
      "epoch": 1.0692440854010385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004897422923608108,
      "loss": 5.9377,
      "mean_token_accuracy": 0.1784738451242447,
      "num_tokens": 23484211.0,
      "step": 9265
    },
    {
      "entropy": 6.308072280883789,
      "epoch": 1.0698211194460474,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004897299714135721,
      "loss": 5.9875,
      "mean_token_accuracy": 0.17716077715158463,
      "num_tokens": 23497110.0,
      "step": 9270
    },
    {
      "entropy": 6.25221061706543,
      "epoch": 1.070398153491056,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004897176432439974,
      "loss": 5.9619,
      "mean_token_accuracy": 0.17784075886011125,
      "num_tokens": 23508803.0,
      "step": 9275
    },
    {
      "entropy": 6.21022596359253,
      "epoch": 1.0709751875360647,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004897053078525016,
      "loss": 5.8666,
      "mean_token_accuracy": 0.186790432035923,
      "num_tokens": 23522209.0,
      "step": 9280
    },
    {
      "entropy": 6.287808609008789,
      "epoch": 1.0715522215810733,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004896929652394993,
      "loss": 5.9444,
      "mean_token_accuracy": 0.17486472874879838,
      "num_tokens": 23534728.0,
      "step": 9285
    },
    {
      "entropy": 6.237624979019165,
      "epoch": 1.072129255626082,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004896806154054057,
      "loss": 6.0091,
      "mean_token_accuracy": 0.17814430147409438,
      "num_tokens": 23548389.0,
      "step": 9290
    },
    {
      "entropy": 6.26085901260376,
      "epoch": 1.0727062896710906,
      "grad_norm": 0.875,
      "learning_rate": 0.0004896682583506363,
      "loss": 5.908,
      "mean_token_accuracy": 0.1784102812409401,
      "num_tokens": 23560260.0,
      "step": 9295
    },
    {
      "entropy": 6.253527116775513,
      "epoch": 1.0732833237160992,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004896558940756067,
      "loss": 5.9347,
      "mean_token_accuracy": 0.1778605744242668,
      "num_tokens": 23572771.0,
      "step": 9300
    },
    {
      "entropy": 6.302005052566528,
      "epoch": 1.0738603577611079,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004896435225807327,
      "loss": 5.9718,
      "mean_token_accuracy": 0.17853919118642808,
      "num_tokens": 23584769.0,
      "step": 9305
    },
    {
      "entropy": 6.230500602722168,
      "epoch": 1.0744373918061165,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004896311438664304,
      "loss": 5.9174,
      "mean_token_accuracy": 0.18021211475133897,
      "num_tokens": 23598912.0,
      "step": 9310
    },
    {
      "entropy": 6.2525266170501705,
      "epoch": 1.0750144258511252,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004896187579331163,
      "loss": 5.9643,
      "mean_token_accuracy": 0.17552366554737092,
      "num_tokens": 23611781.0,
      "step": 9315
    },
    {
      "entropy": 6.244886016845703,
      "epoch": 1.0755914598961338,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004896063647812068,
      "loss": 5.9983,
      "mean_token_accuracy": 0.17546378821134567,
      "num_tokens": 23624116.0,
      "step": 9320
    },
    {
      "entropy": 6.286900043487549,
      "epoch": 1.0761684939411424,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004895939644111189,
      "loss": 5.9845,
      "mean_token_accuracy": 0.17510428428649902,
      "num_tokens": 23637456.0,
      "step": 9325
    },
    {
      "entropy": 6.249389886856079,
      "epoch": 1.076745527986151,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004895815568232694,
      "loss": 5.9668,
      "mean_token_accuracy": 0.1736954465508461,
      "num_tokens": 23649572.0,
      "step": 9330
    },
    {
      "entropy": 6.250957822799682,
      "epoch": 1.07732256203116,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004895691420180759,
      "loss": 5.9146,
      "mean_token_accuracy": 0.1833432734012604,
      "num_tokens": 23662934.0,
      "step": 9335
    },
    {
      "entropy": 6.184030055999756,
      "epoch": 1.0778995960761686,
      "grad_norm": 0.8125,
      "learning_rate": 0.000489556719995956,
      "loss": 5.8019,
      "mean_token_accuracy": 0.18784038573503495,
      "num_tokens": 23676258.0,
      "step": 9340
    },
    {
      "entropy": 6.3182549476623535,
      "epoch": 1.0784766301211772,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004895442907573274,
      "loss": 5.9606,
      "mean_token_accuracy": 0.17187910676002502,
      "num_tokens": 23689361.0,
      "step": 9345
    },
    {
      "entropy": 6.260778760910034,
      "epoch": 1.0790536641661859,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000489531854302608,
      "loss": 5.9308,
      "mean_token_accuracy": 0.17885816544294358,
      "num_tokens": 23702157.0,
      "step": 9350
    },
    {
      "entropy": 6.2573935985565186,
      "epoch": 1.0796306982111945,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004895194106322164,
      "loss": 5.9519,
      "mean_token_accuracy": 0.18166753500699998,
      "num_tokens": 23713846.0,
      "step": 9355
    },
    {
      "entropy": 6.187077331542969,
      "epoch": 1.0802077322562031,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004895069597465709,
      "loss": 5.8583,
      "mean_token_accuracy": 0.18885054141283036,
      "num_tokens": 23727685.0,
      "step": 9360
    },
    {
      "entropy": 6.219455862045288,
      "epoch": 1.0807847663012118,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004894945016460904,
      "loss": 5.8971,
      "mean_token_accuracy": 0.18692795187234879,
      "num_tokens": 23740806.0,
      "step": 9365
    },
    {
      "entropy": 6.238830184936523,
      "epoch": 1.0813618003462204,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004894820363311938,
      "loss": 5.888,
      "mean_token_accuracy": 0.18222576528787612,
      "num_tokens": 23752968.0,
      "step": 9370
    },
    {
      "entropy": 6.290104246139526,
      "epoch": 1.081938834391229,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004894695638023005,
      "loss": 6.0113,
      "mean_token_accuracy": 0.1772119954228401,
      "num_tokens": 23765141.0,
      "step": 9375
    },
    {
      "entropy": 6.247321557998657,
      "epoch": 1.0825158684362377,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048945708405983,
      "loss": 5.8983,
      "mean_token_accuracy": 0.17490099221467972,
      "num_tokens": 23778294.0,
      "step": 9380
    },
    {
      "entropy": 6.238208818435669,
      "epoch": 1.0830929024812463,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004894445971042019,
      "loss": 5.9281,
      "mean_token_accuracy": 0.1833486407995224,
      "num_tokens": 23790687.0,
      "step": 9385
    },
    {
      "entropy": 6.223427867889404,
      "epoch": 1.083669936526255,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004894321029358364,
      "loss": 5.8995,
      "mean_token_accuracy": 0.1771431177854538,
      "num_tokens": 23803569.0,
      "step": 9390
    },
    {
      "entropy": 6.22196626663208,
      "epoch": 1.0842469705712636,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004894196015551536,
      "loss": 5.9422,
      "mean_token_accuracy": 0.17300954908132554,
      "num_tokens": 23816693.0,
      "step": 9395
    },
    {
      "entropy": 6.262089967727661,
      "epoch": 1.0848240046162723,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000489407092962574,
      "loss": 5.9755,
      "mean_token_accuracy": 0.1807078331708908,
      "num_tokens": 23829139.0,
      "step": 9400
    },
    {
      "entropy": 6.244575166702271,
      "epoch": 1.085401038661281,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004893945771585183,
      "loss": 5.9667,
      "mean_token_accuracy": 0.18163262009620668,
      "num_tokens": 23842489.0,
      "step": 9405
    },
    {
      "entropy": 6.280972623825074,
      "epoch": 1.0859780727062898,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004893820541434075,
      "loss": 5.9288,
      "mean_token_accuracy": 0.1811828762292862,
      "num_tokens": 23854555.0,
      "step": 9410
    },
    {
      "entropy": 6.287322998046875,
      "epoch": 1.0865551067512984,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004893695239176629,
      "loss": 5.9481,
      "mean_token_accuracy": 0.17474596500396727,
      "num_tokens": 23868294.0,
      "step": 9415
    },
    {
      "entropy": 6.219698524475097,
      "epoch": 1.087132140796307,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004893569864817057,
      "loss": 5.8219,
      "mean_token_accuracy": 0.18769705891609192,
      "num_tokens": 23880475.0,
      "step": 9420
    },
    {
      "entropy": 6.300324106216431,
      "epoch": 1.0877091748413157,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004893444418359579,
      "loss": 6.0413,
      "mean_token_accuracy": 0.17764434069395066,
      "num_tokens": 23893008.0,
      "step": 9425
    },
    {
      "entropy": 6.2920409679412845,
      "epoch": 1.0882862088863243,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000489331889980841,
      "loss": 5.9756,
      "mean_token_accuracy": 0.17832292914390563,
      "num_tokens": 23906216.0,
      "step": 9430
    },
    {
      "entropy": 6.258578729629517,
      "epoch": 1.088863242931333,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004893193309167778,
      "loss": 5.9054,
      "mean_token_accuracy": 0.17965265810489656,
      "num_tokens": 23918978.0,
      "step": 9435
    },
    {
      "entropy": 6.262158203125,
      "epoch": 1.0894402769763416,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004893067646441902,
      "loss": 5.9469,
      "mean_token_accuracy": 0.17881031781435014,
      "num_tokens": 23931536.0,
      "step": 9440
    },
    {
      "entropy": 6.285189199447632,
      "epoch": 1.0900173110213502,
      "grad_norm": 0.8125,
      "learning_rate": 0.000489294191163501,
      "loss": 5.9702,
      "mean_token_accuracy": 0.1747586250305176,
      "num_tokens": 23945364.0,
      "step": 9445
    },
    {
      "entropy": 6.261805629730224,
      "epoch": 1.0905943450663589,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892816104751331,
      "loss": 5.9385,
      "mean_token_accuracy": 0.17939407527446746,
      "num_tokens": 23956654.0,
      "step": 9450
    },
    {
      "entropy": 6.364009046554566,
      "epoch": 1.0911713791113675,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004892690225795096,
      "loss": 6.0002,
      "mean_token_accuracy": 0.17271680682897567,
      "num_tokens": 23970902.0,
      "step": 9455
    },
    {
      "entropy": 6.364234161376953,
      "epoch": 1.0917484131563762,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004892564274770542,
      "loss": 6.0098,
      "mean_token_accuracy": 0.17067109793424606,
      "num_tokens": 23984312.0,
      "step": 9460
    },
    {
      "entropy": 6.2884523391723635,
      "epoch": 1.0923254472013848,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892438251681901,
      "loss": 5.9216,
      "mean_token_accuracy": 0.17963262945413588,
      "num_tokens": 23996040.0,
      "step": 9465
    },
    {
      "entropy": 6.197104406356812,
      "epoch": 1.0929024812463934,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004892312156533413,
      "loss": 5.8741,
      "mean_token_accuracy": 0.18320828825235366,
      "num_tokens": 24008737.0,
      "step": 9470
    },
    {
      "entropy": 6.250469303131103,
      "epoch": 1.0934795152914023,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892185989329321,
      "loss": 5.9258,
      "mean_token_accuracy": 0.18122901767492294,
      "num_tokens": 24021735.0,
      "step": 9475
    },
    {
      "entropy": 6.245228815078735,
      "epoch": 1.094056549336411,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004892059750073868,
      "loss": 5.9017,
      "mean_token_accuracy": 0.18107449859380723,
      "num_tokens": 24034391.0,
      "step": 9480
    },
    {
      "entropy": 6.247367858886719,
      "epoch": 1.0946335833814196,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004891933438771299,
      "loss": 5.9474,
      "mean_token_accuracy": 0.1786785215139389,
      "num_tokens": 24047331.0,
      "step": 9485
    },
    {
      "entropy": 6.277433919906616,
      "epoch": 1.0952106174264282,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004891807055425862,
      "loss": 5.9482,
      "mean_token_accuracy": 0.18003502041101455,
      "num_tokens": 24060111.0,
      "step": 9490
    },
    {
      "entropy": 6.210342884063721,
      "epoch": 1.0957876514714369,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004891680600041809,
      "loss": 5.961,
      "mean_token_accuracy": 0.1785256028175354,
      "num_tokens": 24072275.0,
      "step": 9495
    },
    {
      "entropy": 6.2774945259094235,
      "epoch": 1.0963646855164455,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004891554072623392,
      "loss": 5.9307,
      "mean_token_accuracy": 0.17721709311008454,
      "num_tokens": 24084647.0,
      "step": 9500
    },
    {
      "entropy": 6.23449592590332,
      "epoch": 1.0969417195614541,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004891427473174869,
      "loss": 5.8189,
      "mean_token_accuracy": 0.19999233931303023,
      "num_tokens": 24097565.0,
      "step": 9505
    },
    {
      "entropy": 6.25209789276123,
      "epoch": 1.0975187536064628,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004891300801700496,
      "loss": 5.9335,
      "mean_token_accuracy": 0.18282418251037597,
      "num_tokens": 24111603.0,
      "step": 9510
    },
    {
      "entropy": 6.259485149383545,
      "epoch": 1.0980957876514714,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004891174058204534,
      "loss": 5.8634,
      "mean_token_accuracy": 0.18556420505046844,
      "num_tokens": 24122763.0,
      "step": 9515
    },
    {
      "entropy": 6.221113395690918,
      "epoch": 1.09867282169648,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004891047242691246,
      "loss": 5.8904,
      "mean_token_accuracy": 0.18249062448740005,
      "num_tokens": 24134591.0,
      "step": 9520
    },
    {
      "entropy": 6.282765626907349,
      "epoch": 1.0992498557414887,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004890920355164897,
      "loss": 5.9997,
      "mean_token_accuracy": 0.1753741830587387,
      "num_tokens": 24147574.0,
      "step": 9525
    },
    {
      "entropy": 6.3097083568573,
      "epoch": 1.0998268897864973,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004890793395629756,
      "loss": 5.9537,
      "mean_token_accuracy": 0.17608107179403304,
      "num_tokens": 24159760.0,
      "step": 9530
    },
    {
      "entropy": 6.215789079666138,
      "epoch": 1.100403923831506,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004890666364090093,
      "loss": 5.9092,
      "mean_token_accuracy": 0.1857651188969612,
      "num_tokens": 24173577.0,
      "step": 9535
    },
    {
      "entropy": 6.2631007671356205,
      "epoch": 1.1009809578765146,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004890539260550181,
      "loss": 5.9339,
      "mean_token_accuracy": 0.18159197866916657,
      "num_tokens": 24185180.0,
      "step": 9540
    },
    {
      "entropy": 6.223391771316528,
      "epoch": 1.1015579919215233,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004890412085014292,
      "loss": 5.9155,
      "mean_token_accuracy": 0.17636099606752395,
      "num_tokens": 24198967.0,
      "step": 9545
    },
    {
      "entropy": 6.245272588729859,
      "epoch": 1.1021350259665321,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004890284837486706,
      "loss": 5.9933,
      "mean_token_accuracy": 0.17728378176689147,
      "num_tokens": 24210770.0,
      "step": 9550
    },
    {
      "entropy": 6.24344425201416,
      "epoch": 1.1027120600115408,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004890157517971704,
      "loss": 5.9213,
      "mean_token_accuracy": 0.17984783202409743,
      "num_tokens": 24222970.0,
      "step": 9555
    },
    {
      "entropy": 6.303714466094971,
      "epoch": 1.1032890940565494,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004890030126473566,
      "loss": 5.9768,
      "mean_token_accuracy": 0.17428198158740998,
      "num_tokens": 24235591.0,
      "step": 9560
    },
    {
      "entropy": 6.269661855697632,
      "epoch": 1.103866128101558,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889902662996578,
      "loss": 5.9264,
      "mean_token_accuracy": 0.1798974707722664,
      "num_tokens": 24248844.0,
      "step": 9565
    },
    {
      "entropy": 6.260860395431519,
      "epoch": 1.1044431621465667,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004889775127545027,
      "loss": 5.9247,
      "mean_token_accuracy": 0.18392487317323686,
      "num_tokens": 24260267.0,
      "step": 9570
    },
    {
      "entropy": 6.317773914337158,
      "epoch": 1.1050201961915753,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004889647520123202,
      "loss": 5.9876,
      "mean_token_accuracy": 0.17627719342708587,
      "num_tokens": 24273019.0,
      "step": 9575
    },
    {
      "entropy": 6.271059656143189,
      "epoch": 1.105597230236584,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004889519840735396,
      "loss": 5.8994,
      "mean_token_accuracy": 0.18268461227416993,
      "num_tokens": 24284805.0,
      "step": 9580
    },
    {
      "entropy": 6.235453796386719,
      "epoch": 1.1061742642815926,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004889392089385903,
      "loss": 5.9807,
      "mean_token_accuracy": 0.1702512338757515,
      "num_tokens": 24297080.0,
      "step": 9585
    },
    {
      "entropy": 6.291859579086304,
      "epoch": 1.1067512983266012,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004889264266079019,
      "loss": 5.9423,
      "mean_token_accuracy": 0.1780322164297104,
      "num_tokens": 24308237.0,
      "step": 9590
    },
    {
      "entropy": 6.294753837585449,
      "epoch": 1.1073283323716099,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004889136370819045,
      "loss": 5.9285,
      "mean_token_accuracy": 0.17732006907463074,
      "num_tokens": 24322435.0,
      "step": 9595
    },
    {
      "entropy": 6.284805250167847,
      "epoch": 1.1079053664166185,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004889008403610281,
      "loss": 5.949,
      "mean_token_accuracy": 0.17276936620473862,
      "num_tokens": 24335372.0,
      "step": 9600
    },
    {
      "entropy": 6.239071464538574,
      "epoch": 1.1084824004616272,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004888880364457033,
      "loss": 5.8372,
      "mean_token_accuracy": 0.18595589846372604,
      "num_tokens": 24347956.0,
      "step": 9605
    },
    {
      "entropy": 6.267660903930664,
      "epoch": 1.1090594345066358,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004888752253363604,
      "loss": 5.9154,
      "mean_token_accuracy": 0.17928190678358077,
      "num_tokens": 24360378.0,
      "step": 9610
    },
    {
      "entropy": 6.2103160381317135,
      "epoch": 1.1096364685516447,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004888624070334308,
      "loss": 5.9451,
      "mean_token_accuracy": 0.1764599710702896,
      "num_tokens": 24372492.0,
      "step": 9615
    },
    {
      "entropy": 6.338330411911011,
      "epoch": 1.1102135025966533,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004888495815373452,
      "loss": 5.9821,
      "mean_token_accuracy": 0.17369418740272521,
      "num_tokens": 24384634.0,
      "step": 9620
    },
    {
      "entropy": 6.301195240020752,
      "epoch": 1.110790536641662,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004888367488485351,
      "loss": 5.9912,
      "mean_token_accuracy": 0.17869668155908586,
      "num_tokens": 24397117.0,
      "step": 9625
    },
    {
      "entropy": 6.26541485786438,
      "epoch": 1.1113675706866706,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004888239089674323,
      "loss": 5.9391,
      "mean_token_accuracy": 0.18358063846826553,
      "num_tokens": 24409071.0,
      "step": 9630
    },
    {
      "entropy": 6.195077991485595,
      "epoch": 1.1119446047316792,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004888110618944686,
      "loss": 5.9033,
      "mean_token_accuracy": 0.17729021608829498,
      "num_tokens": 24422016.0,
      "step": 9635
    },
    {
      "entropy": 6.241928386688232,
      "epoch": 1.1125216387766879,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004887982076300759,
      "loss": 5.9767,
      "mean_token_accuracy": 0.17864816784858703,
      "num_tokens": 24433356.0,
      "step": 9640
    },
    {
      "entropy": 6.205294370651245,
      "epoch": 1.1130986728216965,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0004887853461746867,
      "loss": 5.8722,
      "mean_token_accuracy": 0.18485870957374573,
      "num_tokens": 24449027.0,
      "step": 9645
    },
    {
      "entropy": 6.290260124206543,
      "epoch": 1.1136757068667051,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004887724775287336,
      "loss": 5.9492,
      "mean_token_accuracy": 0.17543125450611113,
      "num_tokens": 24461592.0,
      "step": 9650
    },
    {
      "entropy": 6.217285299301148,
      "epoch": 1.1142527409117138,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004887596016926494,
      "loss": 5.9454,
      "mean_token_accuracy": 0.1770663306117058,
      "num_tokens": 24474003.0,
      "step": 9655
    },
    {
      "entropy": 6.249671125411988,
      "epoch": 1.1148297749567224,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004887467186668673,
      "loss": 5.8903,
      "mean_token_accuracy": 0.1817844644188881,
      "num_tokens": 24487096.0,
      "step": 9660
    },
    {
      "entropy": 6.210844850540161,
      "epoch": 1.115406809001731,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004887338284518204,
      "loss": 5.8311,
      "mean_token_accuracy": 0.19033624082803727,
      "num_tokens": 24499685.0,
      "step": 9665
    },
    {
      "entropy": 6.265480661392212,
      "epoch": 1.1159838430467397,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004887209310479423,
      "loss": 5.9627,
      "mean_token_accuracy": 0.17856135815382004,
      "num_tokens": 24512372.0,
      "step": 9670
    },
    {
      "entropy": 6.109089088439942,
      "epoch": 1.1165608770917483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004887080264556668,
      "loss": 5.6992,
      "mean_token_accuracy": 0.19630496203899384,
      "num_tokens": 24524694.0,
      "step": 9675
    },
    {
      "entropy": 6.202250719070435,
      "epoch": 1.117137911136757,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004886951146754282,
      "loss": 5.8167,
      "mean_token_accuracy": 0.18757638931274415,
      "num_tokens": 24537779.0,
      "step": 9680
    },
    {
      "entropy": 6.218533182144165,
      "epoch": 1.1177149451817656,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004886821957076603,
      "loss": 5.8706,
      "mean_token_accuracy": 0.18529890924692155,
      "num_tokens": 24551012.0,
      "step": 9685
    },
    {
      "entropy": 6.220413780212402,
      "epoch": 1.1182919792267745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488669269552798,
      "loss": 5.8784,
      "mean_token_accuracy": 0.17863376289606095,
      "num_tokens": 24562464.0,
      "step": 9690
    },
    {
      "entropy": 6.295958089828491,
      "epoch": 1.1188690132717831,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000488656336211276,
      "loss": 5.8816,
      "mean_token_accuracy": 0.18366558402776717,
      "num_tokens": 24574223.0,
      "step": 9695
    },
    {
      "entropy": 6.268126726150513,
      "epoch": 1.1194460473167918,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004886433956835292,
      "loss": 5.9385,
      "mean_token_accuracy": 0.18187979757785797,
      "num_tokens": 24586464.0,
      "step": 9700
    },
    {
      "entropy": 6.195445489883423,
      "epoch": 1.1200230813618004,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004886304479699929,
      "loss": 5.9321,
      "mean_token_accuracy": 0.17413788586854934,
      "num_tokens": 24599071.0,
      "step": 9705
    },
    {
      "entropy": 6.211527633666992,
      "epoch": 1.120600115406809,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004886174930711027,
      "loss": 5.8169,
      "mean_token_accuracy": 0.19079080671072007,
      "num_tokens": 24611302.0,
      "step": 9710
    },
    {
      "entropy": 6.238886785507202,
      "epoch": 1.1211771494518177,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004886045309872941,
      "loss": 5.9838,
      "mean_token_accuracy": 0.17684394866228104,
      "num_tokens": 24624725.0,
      "step": 9715
    },
    {
      "entropy": 6.259770631790161,
      "epoch": 1.1217541834968263,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004885915617190034,
      "loss": 5.992,
      "mean_token_accuracy": 0.1820837065577507,
      "num_tokens": 24638797.0,
      "step": 9720
    },
    {
      "entropy": 6.302796792984009,
      "epoch": 1.122331217541835,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004885785852666664,
      "loss": 5.9229,
      "mean_token_accuracy": 0.17878200858831406,
      "num_tokens": 24651462.0,
      "step": 9725
    },
    {
      "entropy": 6.206921529769898,
      "epoch": 1.1229082515868436,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004885656016307199,
      "loss": 5.9213,
      "mean_token_accuracy": 0.18251071125268936,
      "num_tokens": 24664351.0,
      "step": 9730
    },
    {
      "entropy": 6.265410995483398,
      "epoch": 1.1234852856318522,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004885526108116004,
      "loss": 5.9288,
      "mean_token_accuracy": 0.18047355115413666,
      "num_tokens": 24677332.0,
      "step": 9735
    },
    {
      "entropy": 6.259086179733276,
      "epoch": 1.1240623196768609,
      "grad_norm": 0.796875,
      "learning_rate": 0.000488539612809745,
      "loss": 5.9438,
      "mean_token_accuracy": 0.18085369616746902,
      "num_tokens": 24690408.0,
      "step": 9740
    },
    {
      "entropy": 6.236653900146484,
      "epoch": 1.1246393537218695,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004885266076255907,
      "loss": 6.0472,
      "mean_token_accuracy": 0.17337984293699266,
      "num_tokens": 24703776.0,
      "step": 9745
    },
    {
      "entropy": 6.252508497238159,
      "epoch": 1.1252163877668782,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000488513595259575,
      "loss": 5.8047,
      "mean_token_accuracy": 0.1864628329873085,
      "num_tokens": 24716385.0,
      "step": 9750
    },
    {
      "entropy": 6.190405654907226,
      "epoch": 1.125793421811887,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004885005757121357,
      "loss": 5.8653,
      "mean_token_accuracy": 0.1836153119802475,
      "num_tokens": 24729055.0,
      "step": 9755
    },
    {
      "entropy": 6.235238027572632,
      "epoch": 1.1263704558568954,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004884875489837105,
      "loss": 5.8967,
      "mean_token_accuracy": 0.18671600222587587,
      "num_tokens": 24741168.0,
      "step": 9760
    },
    {
      "entropy": 6.274339389801026,
      "epoch": 1.1269474899019043,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004884745150747378,
      "loss": 5.878,
      "mean_token_accuracy": 0.18690308183431625,
      "num_tokens": 24753682.0,
      "step": 9765
    },
    {
      "entropy": 6.100393867492675,
      "epoch": 1.127524523946913,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004884614739856556,
      "loss": 5.8941,
      "mean_token_accuracy": 0.18680865913629532,
      "num_tokens": 24767528.0,
      "step": 9770
    },
    {
      "entropy": 6.275787830352783,
      "epoch": 1.1281015579919216,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004884484257169028,
      "loss": 5.9828,
      "mean_token_accuracy": 0.17152058184146882,
      "num_tokens": 24780989.0,
      "step": 9775
    },
    {
      "entropy": 6.291332054138183,
      "epoch": 1.1286785920369302,
      "grad_norm": 0.875,
      "learning_rate": 0.0004884353702689183,
      "loss": 5.9483,
      "mean_token_accuracy": 0.17670078277587892,
      "num_tokens": 24792595.0,
      "step": 9780
    },
    {
      "entropy": 6.212571907043457,
      "epoch": 1.1292556260819389,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004884223076421411,
      "loss": 5.8811,
      "mean_token_accuracy": 0.17897730320692062,
      "num_tokens": 24805665.0,
      "step": 9785
    },
    {
      "entropy": 6.260483598709106,
      "epoch": 1.1298326601269475,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004884092378370106,
      "loss": 5.8229,
      "mean_token_accuracy": 0.1905406802892685,
      "num_tokens": 24818276.0,
      "step": 9790
    },
    {
      "entropy": 6.23786735534668,
      "epoch": 1.1304096941719561,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004883961608539664,
      "loss": 5.9127,
      "mean_token_accuracy": 0.17826142013072968,
      "num_tokens": 24831828.0,
      "step": 9795
    },
    {
      "entropy": 6.209011077880859,
      "epoch": 1.1309867282169648,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004883830766934484,
      "loss": 5.8813,
      "mean_token_accuracy": 0.18617275655269622,
      "num_tokens": 24844597.0,
      "step": 9800
    },
    {
      "entropy": 6.194504928588867,
      "epoch": 1.1315637622619734,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004883699853558965,
      "loss": 5.89,
      "mean_token_accuracy": 0.18529684394598006,
      "num_tokens": 24856136.0,
      "step": 9805
    },
    {
      "entropy": 6.2990494728088375,
      "epoch": 1.132140796306982,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004883568868417511,
      "loss": 6.0407,
      "mean_token_accuracy": 0.17000253051519393,
      "num_tokens": 24870888.0,
      "step": 9810
    },
    {
      "entropy": 6.2523195266723635,
      "epoch": 1.1327178303519907,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004883437811514528,
      "loss": 5.9622,
      "mean_token_accuracy": 0.1797739028930664,
      "num_tokens": 24884239.0,
      "step": 9815
    },
    {
      "entropy": 6.243176984786987,
      "epoch": 1.1332948643969993,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004883306682854424,
      "loss": 5.8781,
      "mean_token_accuracy": 0.18946582525968553,
      "num_tokens": 24896370.0,
      "step": 9820
    },
    {
      "entropy": 6.19445104598999,
      "epoch": 1.133871898442008,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004883175482441611,
      "loss": 5.9536,
      "mean_token_accuracy": 0.1723155528306961,
      "num_tokens": 24908868.0,
      "step": 9825
    },
    {
      "entropy": 6.236796855926514,
      "epoch": 1.1344489324870168,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004883044210280499,
      "loss": 6.0054,
      "mean_token_accuracy": 0.18114019483327864,
      "num_tokens": 24920630.0,
      "step": 9830
    },
    {
      "entropy": 6.168635559082031,
      "epoch": 1.1350259665320255,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004882912866375505,
      "loss": 5.8071,
      "mean_token_accuracy": 0.18634059578180312,
      "num_tokens": 24932787.0,
      "step": 9835
    },
    {
      "entropy": 6.263698530197144,
      "epoch": 1.1356030005770341,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048827814507310455,
      "loss": 5.8698,
      "mean_token_accuracy": 0.18093537539243698,
      "num_tokens": 24944436.0,
      "step": 9840
    },
    {
      "entropy": 6.259082365036011,
      "epoch": 1.1361800346220428,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048826499633515416,
      "loss": 5.873,
      "mean_token_accuracy": 0.18810746669769288,
      "num_tokens": 24956944.0,
      "step": 9845
    },
    {
      "entropy": 6.270704078674316,
      "epoch": 1.1367570686670514,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00048825184042414156,
      "loss": 5.9421,
      "mean_token_accuracy": 0.18319830447435378,
      "num_tokens": 24971540.0,
      "step": 9850
    },
    {
      "entropy": 6.285923004150391,
      "epoch": 1.13733410271206,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004882386773405092,
      "loss": 5.9922,
      "mean_token_accuracy": 0.1748587116599083,
      "num_tokens": 24984056.0,
      "step": 9855
    },
    {
      "entropy": 6.226939535140991,
      "epoch": 1.1379111367570687,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004882255070846999,
      "loss": 5.9907,
      "mean_token_accuracy": 0.1781661555171013,
      "num_tokens": 24995625.0,
      "step": 9860
    },
    {
      "entropy": 6.298963260650635,
      "epoch": 1.1384881708020773,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00048821232965715663,
      "loss": 5.8656,
      "mean_token_accuracy": 0.1751459628343582,
      "num_tokens": 25007336.0,
      "step": 9865
    },
    {
      "entropy": 6.229776239395141,
      "epoch": 1.139065204847086,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004881991450583225,
      "loss": 5.9496,
      "mean_token_accuracy": 0.18053314387798308,
      "num_tokens": 25019446.0,
      "step": 9870
    },
    {
      "entropy": 6.266025638580322,
      "epoch": 1.1396422388920946,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000488185953288641,
      "loss": 5.8927,
      "mean_token_accuracy": 0.1791336715221405,
      "num_tokens": 25031518.0,
      "step": 9875
    },
    {
      "entropy": 6.200574684143066,
      "epoch": 1.1402192729371032,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004881727543485559,
      "loss": 5.8407,
      "mean_token_accuracy": 0.1861478477716446,
      "num_tokens": 25043365.0,
      "step": 9880
    },
    {
      "entropy": 6.250562047958374,
      "epoch": 1.1407963069821119,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048815954823851107,
      "loss": 5.9886,
      "mean_token_accuracy": 0.17195338010787964,
      "num_tokens": 25055748.0,
      "step": 9885
    },
    {
      "entropy": 6.2705894947052006,
      "epoch": 1.1413733410271205,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048814633495895067,
      "loss": 5.8634,
      "mean_token_accuracy": 0.17923913896083832,
      "num_tokens": 25067904.0,
      "step": 9890
    },
    {
      "entropy": 6.219842052459716,
      "epoch": 1.1419503750721294,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004881331145103192,
      "loss": 5.916,
      "mean_token_accuracy": 0.1817471206188202,
      "num_tokens": 25081584.0,
      "step": 9895
    },
    {
      "entropy": 6.233426237106324,
      "epoch": 1.1425274091171378,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004881198868930614,
      "loss": 5.9405,
      "mean_token_accuracy": 0.17513066232204438,
      "num_tokens": 25094555.0,
      "step": 9900
    },
    {
      "entropy": 6.314224767684936,
      "epoch": 1.1431044431621467,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048810665210762195,
      "loss": 5.9795,
      "mean_token_accuracy": 0.17922903299331666,
      "num_tokens": 25108405.0,
      "step": 9905
    },
    {
      "entropy": 6.2468496799469,
      "epoch": 1.1436814772071553,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048809341015444615,
      "loss": 5.9598,
      "mean_token_accuracy": 0.17613538652658461,
      "num_tokens": 25121404.0,
      "step": 9910
    },
    {
      "entropy": 6.228936100006104,
      "epoch": 1.144258511252164,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048808016103397934,
      "loss": 5.9006,
      "mean_token_accuracy": 0.18569639325141907,
      "num_tokens": 25134773.0,
      "step": 9915
    },
    {
      "entropy": 6.231281471252442,
      "epoch": 1.1448355452971726,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004880669047466671,
      "loss": 5.8676,
      "mean_token_accuracy": 0.18478555381298065,
      "num_tokens": 25147873.0,
      "step": 9920
    },
    {
      "entropy": 6.158187294006348,
      "epoch": 1.1454125793421812,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048805364129295554,
      "loss": 5.8132,
      "mean_token_accuracy": 0.19686105996370315,
      "num_tokens": 25162081.0,
      "step": 9925
    },
    {
      "entropy": 6.31539568901062,
      "epoch": 1.1459896133871899,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048804037067329037,
      "loss": 6.0098,
      "mean_token_accuracy": 0.17687641978263854,
      "num_tokens": 25174203.0,
      "step": 9930
    },
    {
      "entropy": 6.2253416061401365,
      "epoch": 1.1465666474321985,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004880270928881183,
      "loss": 5.9902,
      "mean_token_accuracy": 0.1790986031293869,
      "num_tokens": 25187238.0,
      "step": 9935
    },
    {
      "entropy": 6.233942222595215,
      "epoch": 1.1471436814772071,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004880138079378857,
      "loss": 5.8489,
      "mean_token_accuracy": 0.18493434339761733,
      "num_tokens": 25199456.0,
      "step": 9940
    },
    {
      "entropy": 6.232425022125244,
      "epoch": 1.1477207155222158,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004880005158230395,
      "loss": 5.9085,
      "mean_token_accuracy": 0.1807033970952034,
      "num_tokens": 25215163.0,
      "step": 9945
    },
    {
      "entropy": 6.203218793869018,
      "epoch": 1.1482977495672244,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004879872165440268,
      "loss": 5.9218,
      "mean_token_accuracy": 0.18384969383478164,
      "num_tokens": 25226946.0,
      "step": 9950
    },
    {
      "entropy": 6.203146028518677,
      "epoch": 1.148874783612233,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004879739101012948,
      "loss": 5.8349,
      "mean_token_accuracy": 0.18173687607049943,
      "num_tokens": 25238691.0,
      "step": 9955
    },
    {
      "entropy": 6.153835201263428,
      "epoch": 1.1494518176572417,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004879605964952911,
      "loss": 5.831,
      "mean_token_accuracy": 0.18132368177175523,
      "num_tokens": 25251513.0,
      "step": 9960
    },
    {
      "entropy": 6.217679882049561,
      "epoch": 1.1500288517022503,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048794727572646366,
      "loss": 5.7929,
      "mean_token_accuracy": 0.18657493889331817,
      "num_tokens": 25263513.0,
      "step": 9965
    },
    {
      "entropy": 6.207217168807984,
      "epoch": 1.1506058857472592,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004879339477952603,
      "loss": 5.9144,
      "mean_token_accuracy": 0.17993103563785554,
      "num_tokens": 25276989.0,
      "step": 9970
    },
    {
      "entropy": 6.237392568588257,
      "epoch": 1.1511829197922678,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048792061270212935,
      "loss": 5.9348,
      "mean_token_accuracy": 0.18304591476917267,
      "num_tokens": 25289474.0,
      "step": 9975
    },
    {
      "entropy": 6.265247774124146,
      "epoch": 1.1517599538372765,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004879072704475193,
      "loss": 5.8999,
      "mean_token_accuracy": 0.1862184226512909,
      "num_tokens": 25302102.0,
      "step": 9980
    },
    {
      "entropy": 6.185840034484864,
      "epoch": 1.1523369878822851,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048789392103187906,
      "loss": 5.8916,
      "mean_token_accuracy": 0.1849305585026741,
      "num_tokens": 25315645.0,
      "step": 9985
    },
    {
      "entropy": 6.2735367774963375,
      "epoch": 1.1529140219272938,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004878805644556575,
      "loss": 5.9489,
      "mean_token_accuracy": 0.1791803017258644,
      "num_tokens": 25328272.0,
      "step": 9990
    },
    {
      "entropy": 6.214619493484497,
      "epoch": 1.1534910559723024,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000487867200719304,
      "loss": 5.8995,
      "mean_token_accuracy": 0.18508705645799636,
      "num_tokens": 25340449.0,
      "step": 9995
    },
    {
      "entropy": 6.209858751296997,
      "epoch": 1.154068090017311,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004878538298232678,
      "loss": 5.8365,
      "mean_token_accuracy": 0.1894098103046417,
      "num_tokens": 25353232.0,
      "step": 10000
    },
    {
      "entropy": 6.244632434844971,
      "epoch": 1.1546451240623197,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004878404517679988,
      "loss": 5.9236,
      "mean_token_accuracy": 0.18128351867198944,
      "num_tokens": 25364575.0,
      "step": 10005
    },
    {
      "entropy": 6.134302568435669,
      "epoch": 1.1552221581073283,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048782706655394695,
      "loss": 5.849,
      "mean_token_accuracy": 0.18665627390146255,
      "num_tokens": 25377992.0,
      "step": 10010
    },
    {
      "entropy": 6.2707353115081785,
      "epoch": 1.155799192152337,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004878136741815624,
      "loss": 5.9444,
      "mean_token_accuracy": 0.18252215087413787,
      "num_tokens": 25391168.0,
      "step": 10015
    },
    {
      "entropy": 6.199631929397583,
      "epoch": 1.1563762261973456,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004878002746512956,
      "loss": 5.8329,
      "mean_token_accuracy": 0.18418675661087036,
      "num_tokens": 25405211.0,
      "step": 10020
    },
    {
      "entropy": 6.231357145309448,
      "epoch": 1.1569532602423542,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004877868679635974,
      "loss": 5.9151,
      "mean_token_accuracy": 0.18079279214143754,
      "num_tokens": 25416902.0,
      "step": 10025
    },
    {
      "entropy": 6.238675165176391,
      "epoch": 1.1575302942873629,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004877734541189185,
      "loss": 5.9455,
      "mean_token_accuracy": 0.1775543436408043,
      "num_tokens": 25429071.0,
      "step": 10030
    },
    {
      "entropy": 6.1994078159332275,
      "epoch": 1.1581073283323717,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048776003311771013,
      "loss": 5.8572,
      "mean_token_accuracy": 0.18071705847978592,
      "num_tokens": 25441153.0,
      "step": 10035
    },
    {
      "entropy": 6.2144859790802,
      "epoch": 1.1586843623773802,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004877466049604238,
      "loss": 5.8553,
      "mean_token_accuracy": 0.18691859543323516,
      "num_tokens": 25454302.0,
      "step": 10040
    },
    {
      "entropy": 6.275542545318603,
      "epoch": 1.159261396422389,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048773316964751106,
      "loss": 5.9295,
      "mean_token_accuracy": 0.17534056007862092,
      "num_tokens": 25466507.0,
      "step": 10045
    },
    {
      "entropy": 6.235187721252442,
      "epoch": 1.1598384304673977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004877197271794239,
      "loss": 5.916,
      "mean_token_accuracy": 0.1796540692448616,
      "num_tokens": 25479952.0,
      "step": 10050
    },
    {
      "entropy": 6.247036266326904,
      "epoch": 1.1604154645124063,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004877062775566142,
      "loss": 5.9867,
      "mean_token_accuracy": 0.16934431195259095,
      "num_tokens": 25491952.0,
      "step": 10055
    },
    {
      "entropy": 6.271330976486206,
      "epoch": 1.160992498557415,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048769282077953464,
      "loss": 5.8806,
      "mean_token_accuracy": 0.17862699180841446,
      "num_tokens": 25505053.0,
      "step": 10060
    },
    {
      "entropy": 6.282383728027344,
      "epoch": 1.1615695326024236,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048767935684863775,
      "loss": 5.9767,
      "mean_token_accuracy": 0.17835345566272737,
      "num_tokens": 25517367.0,
      "step": 10065
    },
    {
      "entropy": 6.28428087234497,
      "epoch": 1.1621465666474322,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004876658857643762,
      "loss": 5.9016,
      "mean_token_accuracy": 0.17964425235986708,
      "num_tokens": 25528697.0,
      "step": 10070
    },
    {
      "entropy": 6.275952768325806,
      "epoch": 1.1627236006924409,
      "grad_norm": 0.875,
      "learning_rate": 0.0004876524075272034,
      "loss": 5.9288,
      "mean_token_accuracy": 0.1738267332315445,
      "num_tokens": 25541109.0,
      "step": 10075
    },
    {
      "entropy": 6.159681749343872,
      "epoch": 1.1633006347374495,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048763892213757234,
      "loss": 5.8333,
      "mean_token_accuracy": 0.1890805646777153,
      "num_tokens": 25553870.0,
      "step": 10080
    },
    {
      "entropy": 6.260570430755616,
      "epoch": 1.1638776687824581,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048762542959593683,
      "loss": 5.9446,
      "mean_token_accuracy": 0.17460126280784607,
      "num_tokens": 25566048.0,
      "step": 10085
    },
    {
      "entropy": 6.178368759155274,
      "epoch": 1.1644547028274668,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004876119299027506,
      "loss": 5.8383,
      "mean_token_accuracy": 0.18482700139284133,
      "num_tokens": 25578282.0,
      "step": 10090
    },
    {
      "entropy": 6.1744965553283695,
      "epoch": 1.1650317368724754,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048759842305846766,
      "loss": 5.8239,
      "mean_token_accuracy": 0.18501157760620118,
      "num_tokens": 25590359.0,
      "step": 10095
    },
    {
      "entropy": 6.315312004089355,
      "epoch": 1.165608770917484,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004875849090635425,
      "loss": 5.9787,
      "mean_token_accuracy": 0.17710129916667938,
      "num_tokens": 25603256.0,
      "step": 10100
    },
    {
      "entropy": 6.200481605529785,
      "epoch": 1.1661858049624927,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048757138791842943,
      "loss": 5.9544,
      "mean_token_accuracy": 0.17941863387823104,
      "num_tokens": 25616844.0,
      "step": 10105
    },
    {
      "entropy": 6.236193895339966,
      "epoch": 1.1667628390075016,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004875578596235832,
      "loss": 5.8176,
      "mean_token_accuracy": 0.1952678456902504,
      "num_tokens": 25630490.0,
      "step": 10110
    },
    {
      "entropy": 6.274014806747436,
      "epoch": 1.1673398730525102,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004875443241794591,
      "loss": 5.8916,
      "mean_token_accuracy": 0.18010980784893035,
      "num_tokens": 25645082.0,
      "step": 10115
    },
    {
      "entropy": 6.220429372787476,
      "epoch": 1.1679169070975188,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048753078158651227,
      "loss": 5.9519,
      "mean_token_accuracy": 0.17907483726739884,
      "num_tokens": 25658108.0,
      "step": 10120
    },
    {
      "entropy": 6.217692613601685,
      "epoch": 1.1684939411425275,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000487517231845198,
      "loss": 5.9192,
      "mean_token_accuracy": 0.17816399335861205,
      "num_tokens": 25669883.0,
      "step": 10125
    },
    {
      "entropy": 6.268486404418946,
      "epoch": 1.169070975187536,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004875036749559724,
      "loss": 5.9209,
      "mean_token_accuracy": 0.17892272621393204,
      "num_tokens": 25683393.0,
      "step": 10130
    },
    {
      "entropy": 6.286698532104492,
      "epoch": 1.1696480092325447,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048749011091929115,
      "loss": 5.9222,
      "mean_token_accuracy": 0.17769978046417237,
      "num_tokens": 25696564.0,
      "step": 10135
    },
    {
      "entropy": 6.2154289245605465,
      "epoch": 1.1702250432775534,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004874765397356105,
      "loss": 5.8347,
      "mean_token_accuracy": 0.17812687456607817,
      "num_tokens": 25709350.0,
      "step": 10140
    },
    {
      "entropy": 6.24783148765564,
      "epoch": 1.170802077322562,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004874629614053871,
      "loss": 5.894,
      "mean_token_accuracy": 0.17832050025463103,
      "num_tokens": 25721559.0,
      "step": 10145
    },
    {
      "entropy": 6.291457319259644,
      "epoch": 1.1713791113675707,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004874493759290775,
      "loss": 6.0272,
      "mean_token_accuracy": 0.17179853171110154,
      "num_tokens": 25734185.0,
      "step": 10150
    },
    {
      "entropy": 6.232964324951172,
      "epoch": 1.1719561454125793,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048743578330713854,
      "loss": 5.8595,
      "mean_token_accuracy": 0.18776810318231582,
      "num_tokens": 25746456.0,
      "step": 10155
    },
    {
      "entropy": 6.168695068359375,
      "epoch": 1.172533179457588,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004874221835400277,
      "loss": 5.8924,
      "mean_token_accuracy": 0.181561279296875,
      "num_tokens": 25759893.0,
      "step": 10160
    },
    {
      "entropy": 6.282555675506591,
      "epoch": 1.1731102135025966,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004874085766282022,
      "loss": 5.9917,
      "mean_token_accuracy": 0.17544945776462556,
      "num_tokens": 25772574.0,
      "step": 10165
    },
    {
      "entropy": 6.2115216732025145,
      "epoch": 1.1736872475476052,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048739496257211966,
      "loss": 5.881,
      "mean_token_accuracy": 0.18557531386613846,
      "num_tokens": 25785748.0,
      "step": 10170
    },
    {
      "entropy": 6.232867574691772,
      "epoch": 1.1742642815926139,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048738134137223815,
      "loss": 5.8806,
      "mean_token_accuracy": 0.18154890239238738,
      "num_tokens": 25799073.0,
      "step": 10175
    },
    {
      "entropy": 6.291546297073364,
      "epoch": 1.1748413156376225,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048736771302901566,
      "loss": 5.8926,
      "mean_token_accuracy": 0.17969266921281815,
      "num_tokens": 25810896.0,
      "step": 10180
    },
    {
      "entropy": 6.231689500808716,
      "epoch": 1.1754183496826314,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048735407754291063,
      "loss": 5.889,
      "mean_token_accuracy": 0.1797415569424629,
      "num_tokens": 25824728.0,
      "step": 10185
    },
    {
      "entropy": 6.2198854923248295,
      "epoch": 1.17599538372764,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048734043491438175,
      "loss": 5.9486,
      "mean_token_accuracy": 0.17400774508714675,
      "num_tokens": 25837569.0,
      "step": 10190
    },
    {
      "entropy": 6.308599996566772,
      "epoch": 1.1765724177726486,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048732678514388773,
      "loss": 5.9849,
      "mean_token_accuracy": 0.17587724328041077,
      "num_tokens": 25850232.0,
      "step": 10195
    },
    {
      "entropy": 6.242422294616699,
      "epoch": 1.1771494518176573,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004873131282318878,
      "loss": 5.9386,
      "mean_token_accuracy": 0.17981125563383102,
      "num_tokens": 25862038.0,
      "step": 10200
    },
    {
      "entropy": 6.305715847015381,
      "epoch": 1.177726485862666,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048729946417884126,
      "loss": 6.0184,
      "mean_token_accuracy": 0.17254866808652877,
      "num_tokens": 25874337.0,
      "step": 10205
    },
    {
      "entropy": 6.239461851119995,
      "epoch": 1.1783035199076746,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004872857929852076,
      "loss": 5.8907,
      "mean_token_accuracy": 0.18850106447935105,
      "num_tokens": 25888175.0,
      "step": 10210
    },
    {
      "entropy": 6.2550124168396,
      "epoch": 1.1788805539526832,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004872721146514469,
      "loss": 5.8429,
      "mean_token_accuracy": 0.18456741124391557,
      "num_tokens": 25900747.0,
      "step": 10215
    },
    {
      "entropy": 6.202442073822022,
      "epoch": 1.1794575879976918,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000487258429178019,
      "loss": 5.8995,
      "mean_token_accuracy": 0.1773380905389786,
      "num_tokens": 25913010.0,
      "step": 10220
    },
    {
      "entropy": 6.213880586624145,
      "epoch": 1.1800346220427005,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048724473656538427,
      "loss": 5.8449,
      "mean_token_accuracy": 0.18177816569805144,
      "num_tokens": 25925757.0,
      "step": 10225
    },
    {
      "entropy": 6.273483371734619,
      "epoch": 1.1806116560877091,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004872310368140032,
      "loss": 5.937,
      "mean_token_accuracy": 0.1752402275800705,
      "num_tokens": 25939574.0,
      "step": 10230
    },
    {
      "entropy": 6.249111223220825,
      "epoch": 1.1811886901327178,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004872173299243368,
      "loss": 5.965,
      "mean_token_accuracy": 0.17713995277881622,
      "num_tokens": 25951142.0,
      "step": 10235
    },
    {
      "entropy": 6.2365669250488285,
      "epoch": 1.1817657241777264,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048720361589684575,
      "loss": 5.9241,
      "mean_token_accuracy": 0.17792513072490693,
      "num_tokens": 25962575.0,
      "step": 10240
    },
    {
      "entropy": 6.23852686882019,
      "epoch": 1.182342758222735,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048718989473199147,
      "loss": 5.9164,
      "mean_token_accuracy": 0.1835549220442772,
      "num_tokens": 25977012.0,
      "step": 10245
    },
    {
      "entropy": 6.216757869720459,
      "epoch": 1.182919792267744,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004871761664302356,
      "loss": 5.9332,
      "mean_token_accuracy": 0.18358702212572098,
      "num_tokens": 25991520.0,
      "step": 10250
    },
    {
      "entropy": 6.2142578125,
      "epoch": 1.1834968263127525,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004871624309920397,
      "loss": 5.8071,
      "mean_token_accuracy": 0.18625397980213165,
      "num_tokens": 26004644.0,
      "step": 10255
    },
    {
      "entropy": 6.1872588157653805,
      "epoch": 1.1840738603577612,
      "grad_norm": 0.875,
      "learning_rate": 0.00048714868841786586,
      "loss": 5.8511,
      "mean_token_accuracy": 0.18240738660097122,
      "num_tokens": 26016452.0,
      "step": 10260
    },
    {
      "entropy": 6.218697023391724,
      "epoch": 1.1846508944027698,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048713493870817626,
      "loss": 5.8881,
      "mean_token_accuracy": 0.18670417070388795,
      "num_tokens": 26028675.0,
      "step": 10265
    },
    {
      "entropy": 6.277977418899536,
      "epoch": 1.1852279284477785,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048712118186343336,
      "loss": 5.9415,
      "mean_token_accuracy": 0.17847172915935516,
      "num_tokens": 26041561.0,
      "step": 10270
    },
    {
      "entropy": 6.124319458007813,
      "epoch": 1.185804962492787,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048710741788409985,
      "loss": 5.7724,
      "mean_token_accuracy": 0.18766767531633377,
      "num_tokens": 26052808.0,
      "step": 10275
    },
    {
      "entropy": 6.268497562408447,
      "epoch": 1.1863819965377957,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004870936467706387,
      "loss": 5.9033,
      "mean_token_accuracy": 0.18367504924535752,
      "num_tokens": 26065624.0,
      "step": 10280
    },
    {
      "entropy": 6.202836418151856,
      "epoch": 1.1869590305828044,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004870798685235131,
      "loss": 5.8474,
      "mean_token_accuracy": 0.18399796932935714,
      "num_tokens": 26078075.0,
      "step": 10285
    },
    {
      "entropy": 6.239213514328003,
      "epoch": 1.187536064627813,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048706608314318654,
      "loss": 5.9215,
      "mean_token_accuracy": 0.18444866240024566,
      "num_tokens": 26090454.0,
      "step": 10290
    },
    {
      "entropy": 6.196172046661377,
      "epoch": 1.1881130986728217,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004870522906301225,
      "loss": 5.8467,
      "mean_token_accuracy": 0.1871015638113022,
      "num_tokens": 26103516.0,
      "step": 10295
    },
    {
      "entropy": 6.2758753299713135,
      "epoch": 1.1886901327178303,
      "grad_norm": 0.84375,
      "learning_rate": 0.000487038490984785,
      "loss": 5.9353,
      "mean_token_accuracy": 0.1768675282597542,
      "num_tokens": 26116420.0,
      "step": 10300
    },
    {
      "entropy": 6.208378410339355,
      "epoch": 1.189267166762839,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048702468420763826,
      "loss": 5.8908,
      "mean_token_accuracy": 0.18248891085386276,
      "num_tokens": 26129685.0,
      "step": 10305
    },
    {
      "entropy": 6.291346502304077,
      "epoch": 1.1898442008078476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048701087029914657,
      "loss": 5.9175,
      "mean_token_accuracy": 0.17836499810218812,
      "num_tokens": 26142909.0,
      "step": 10310
    },
    {
      "entropy": 6.265020275115967,
      "epoch": 1.1904212348528562,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004869970492597745,
      "loss": 5.914,
      "mean_token_accuracy": 0.18202037662267684,
      "num_tokens": 26155229.0,
      "step": 10315
    },
    {
      "entropy": 6.238765478134155,
      "epoch": 1.1909982688978649,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004869832210899871,
      "loss": 5.9151,
      "mean_token_accuracy": 0.1824778661131859,
      "num_tokens": 26168309.0,
      "step": 10320
    },
    {
      "entropy": 6.2530022144317625,
      "epoch": 1.1915753029428737,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048696938579024927,
      "loss": 5.8972,
      "mean_token_accuracy": 0.1859513446688652,
      "num_tokens": 26180189.0,
      "step": 10325
    },
    {
      "entropy": 6.141672563552857,
      "epoch": 1.1921523369878824,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048695554336102644,
      "loss": 5.8512,
      "mean_token_accuracy": 0.18196981102228166,
      "num_tokens": 26193125.0,
      "step": 10330
    },
    {
      "entropy": 6.245037460327149,
      "epoch": 1.192729371032891,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048694169380278417,
      "loss": 5.929,
      "mean_token_accuracy": 0.17541455328464509,
      "num_tokens": 26206179.0,
      "step": 10335
    },
    {
      "entropy": 6.311059665679932,
      "epoch": 1.1933064050778996,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004869278371159884,
      "loss": 5.8602,
      "mean_token_accuracy": 0.1836206927895546,
      "num_tokens": 26217887.0,
      "step": 10340
    },
    {
      "entropy": 6.2669158458709715,
      "epoch": 1.1938834391229083,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048691397330110503,
      "loss": 5.8926,
      "mean_token_accuracy": 0.18323934674263,
      "num_tokens": 26231280.0,
      "step": 10345
    },
    {
      "entropy": 6.106601095199585,
      "epoch": 1.194460473167917,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004869001023586005,
      "loss": 5.7311,
      "mean_token_accuracy": 0.19937066733837128,
      "num_tokens": 26245565.0,
      "step": 10350
    },
    {
      "entropy": 6.180907964706421,
      "epoch": 1.1950375072129256,
      "grad_norm": 0.875,
      "learning_rate": 0.0004868862242889413,
      "loss": 5.8993,
      "mean_token_accuracy": 0.18012045174837113,
      "num_tokens": 26257313.0,
      "step": 10355
    },
    {
      "entropy": 6.206119441986084,
      "epoch": 1.1956145412579342,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000486872339092594,
      "loss": 5.8711,
      "mean_token_accuracy": 0.18440091609954834,
      "num_tokens": 26270566.0,
      "step": 10360
    },
    {
      "entropy": 6.232692050933838,
      "epoch": 1.1961915753029428,
      "grad_norm": 0.890625,
      "learning_rate": 0.000486858446770026,
      "loss": 5.9784,
      "mean_token_accuracy": 0.17337860018014908,
      "num_tokens": 26283490.0,
      "step": 10365
    },
    {
      "entropy": 6.255571603775024,
      "epoch": 1.1967686093479515,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004868445473217043,
      "loss": 5.904,
      "mean_token_accuracy": 0.17833744287490844,
      "num_tokens": 26297702.0,
      "step": 10370
    },
    {
      "entropy": 6.221910762786865,
      "epoch": 1.1973456433929601,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004868306407480965,
      "loss": 5.84,
      "mean_token_accuracy": 0.18549697399139403,
      "num_tokens": 26309543.0,
      "step": 10375
    },
    {
      "entropy": 6.223928642272949,
      "epoch": 1.1979226774379688,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048681672704967036,
      "loss": 5.9945,
      "mean_token_accuracy": 0.17395448684692383,
      "num_tokens": 26322259.0,
      "step": 10380
    },
    {
      "entropy": 6.256245803833008,
      "epoch": 1.1984997114829774,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004868028062268938,
      "loss": 5.8578,
      "mean_token_accuracy": 0.18279649764299394,
      "num_tokens": 26334985.0,
      "step": 10385
    },
    {
      "entropy": 6.132374000549317,
      "epoch": 1.1990767455279863,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048678887828023504,
      "loss": 5.6968,
      "mean_token_accuracy": 0.1954952985048294,
      "num_tokens": 26348441.0,
      "step": 10390
    },
    {
      "entropy": 6.240096712112427,
      "epoch": 1.1996537795729947,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004867749432101626,
      "loss": 5.9171,
      "mean_token_accuracy": 0.17672910243272782,
      "num_tokens": 26360848.0,
      "step": 10395
    },
    {
      "entropy": 6.11644811630249,
      "epoch": 1.2002308136180035,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004867610010171451,
      "loss": 5.8365,
      "mean_token_accuracy": 0.19645826071500777,
      "num_tokens": 26373704.0,
      "step": 10400
    },
    {
      "entropy": 6.1811017990112305,
      "epoch": 1.2008078476630122,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048674705170165147,
      "loss": 5.8723,
      "mean_token_accuracy": 0.18627698421478273,
      "num_tokens": 26387035.0,
      "step": 10405
    },
    {
      "entropy": 6.201904392242431,
      "epoch": 1.2013848817080208,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000486733095264151,
      "loss": 5.8028,
      "mean_token_accuracy": 0.1883980467915535,
      "num_tokens": 26399064.0,
      "step": 10410
    },
    {
      "entropy": 6.199052333831787,
      "epoch": 1.2019619157530295,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048671913170511306,
      "loss": 5.9041,
      "mean_token_accuracy": 0.18148902356624602,
      "num_tokens": 26411930.0,
      "step": 10415
    },
    {
      "entropy": 6.176446390151978,
      "epoch": 1.202538949798038,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004867051610250073,
      "loss": 5.8383,
      "mean_token_accuracy": 0.182884781062603,
      "num_tokens": 26425855.0,
      "step": 10420
    },
    {
      "entropy": 6.269433307647705,
      "epoch": 1.2031159838430467,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004866911832243035,
      "loss": 5.9339,
      "mean_token_accuracy": 0.1812303602695465,
      "num_tokens": 26437649.0,
      "step": 10425
    },
    {
      "entropy": 6.196122121810913,
      "epoch": 1.2036930178880554,
      "grad_norm": 0.875,
      "learning_rate": 0.00048667719830347203,
      "loss": 5.8448,
      "mean_token_accuracy": 0.17881242781877518,
      "num_tokens": 26449503.0,
      "step": 10430
    },
    {
      "entropy": 6.241456317901611,
      "epoch": 1.204270051933064,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048666320626298307,
      "loss": 5.8891,
      "mean_token_accuracy": 0.1883440360426903,
      "num_tokens": 26462199.0,
      "step": 10435
    },
    {
      "entropy": 6.2513649463653564,
      "epoch": 1.2048470859780727,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048664920710330735,
      "loss": 5.9099,
      "mean_token_accuracy": 0.1840690106153488,
      "num_tokens": 26475295.0,
      "step": 10440
    },
    {
      "entropy": 6.204084205627441,
      "epoch": 1.2054241200230813,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048663520082491564,
      "loss": 5.9022,
      "mean_token_accuracy": 0.180405630171299,
      "num_tokens": 26488846.0,
      "step": 10445
    },
    {
      "entropy": 6.275696706771851,
      "epoch": 1.20600115406809,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004866211874282791,
      "loss": 5.9076,
      "mean_token_accuracy": 0.18316664248704911,
      "num_tokens": 26501645.0,
      "step": 10450
    },
    {
      "entropy": 6.200248670578003,
      "epoch": 1.2065781881130986,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000486607166913869,
      "loss": 5.8792,
      "mean_token_accuracy": 0.1819622129201889,
      "num_tokens": 26513494.0,
      "step": 10455
    },
    {
      "entropy": 6.2503838539123535,
      "epoch": 1.2071552221581072,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048659313928215705,
      "loss": 5.8613,
      "mean_token_accuracy": 0.18637897372245787,
      "num_tokens": 26525931.0,
      "step": 10460
    },
    {
      "entropy": 6.265742921829224,
      "epoch": 1.207732256203116,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004865791045336149,
      "loss": 5.9361,
      "mean_token_accuracy": 0.18052163273096083,
      "num_tokens": 26539445.0,
      "step": 10465
    },
    {
      "entropy": 6.248571872711182,
      "epoch": 1.2083092902481247,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004865650626687146,
      "loss": 5.9364,
      "mean_token_accuracy": 0.17711347937583924,
      "num_tokens": 26551758.0,
      "step": 10470
    },
    {
      "entropy": 6.264990186691284,
      "epoch": 1.2088863242931334,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048655101368792866,
      "loss": 5.8603,
      "mean_token_accuracy": 0.18293217122554778,
      "num_tokens": 26564756.0,
      "step": 10475
    },
    {
      "entropy": 6.166375780105591,
      "epoch": 1.209463358338142,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004865369575917293,
      "loss": 5.839,
      "mean_token_accuracy": 0.1933852419257164,
      "num_tokens": 26579152.0,
      "step": 10480
    },
    {
      "entropy": 6.183269500732422,
      "epoch": 1.2100403923831506,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048652289438058953,
      "loss": 5.8587,
      "mean_token_accuracy": 0.18614097386598588,
      "num_tokens": 26590749.0,
      "step": 10485
    },
    {
      "entropy": 6.316128540039062,
      "epoch": 1.2106174264281593,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048650882405498226,
      "loss": 5.9641,
      "mean_token_accuracy": 0.17796555310487747,
      "num_tokens": 26603859.0,
      "step": 10490
    },
    {
      "entropy": 6.090430736541748,
      "epoch": 1.211194460473168,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004864947466153808,
      "loss": 5.7678,
      "mean_token_accuracy": 0.1986262321472168,
      "num_tokens": 26616002.0,
      "step": 10495
    },
    {
      "entropy": 6.1785181045532225,
      "epoch": 1.2117714945181766,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004864806620622585,
      "loss": 5.8986,
      "mean_token_accuracy": 0.18132064789533614,
      "num_tokens": 26629092.0,
      "step": 10500
    },
    {
      "entropy": 6.254997444152832,
      "epoch": 1.2123485285631852,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004864665703960892,
      "loss": 5.8689,
      "mean_token_accuracy": 0.1796957314014435,
      "num_tokens": 26642830.0,
      "step": 10505
    },
    {
      "entropy": 6.295885705947876,
      "epoch": 1.2129255626081938,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004864524716173469,
      "loss": 6.0447,
      "mean_token_accuracy": 0.17379555404186248,
      "num_tokens": 26656421.0,
      "step": 10510
    },
    {
      "entropy": 6.173092699050903,
      "epoch": 1.2135025966532025,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004864383657265058,
      "loss": 5.8253,
      "mean_token_accuracy": 0.1844503790140152,
      "num_tokens": 26669267.0,
      "step": 10515
    },
    {
      "entropy": 6.273373222351074,
      "epoch": 1.2140796306982111,
      "grad_norm": 0.796875,
      "learning_rate": 0.00048642425272404016,
      "loss": 5.8785,
      "mean_token_accuracy": 0.1848890408873558,
      "num_tokens": 26681074.0,
      "step": 10520
    },
    {
      "entropy": 6.1193849563598635,
      "epoch": 1.2146566647432198,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004864101326104248,
      "loss": 5.8354,
      "mean_token_accuracy": 0.187041799724102,
      "num_tokens": 26694772.0,
      "step": 10525
    },
    {
      "entropy": 6.247080326080322,
      "epoch": 1.2152336987882286,
      "grad_norm": 0.875,
      "learning_rate": 0.0004863960053861348,
      "loss": 5.9271,
      "mean_token_accuracy": 0.1849571779370308,
      "num_tokens": 26706173.0,
      "step": 10530
    },
    {
      "entropy": 6.2308114528656,
      "epoch": 1.215810732833237,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048638187105164507,
      "loss": 5.957,
      "mean_token_accuracy": 0.1795593425631523,
      "num_tokens": 26718273.0,
      "step": 10535
    },
    {
      "entropy": 6.255645513534546,
      "epoch": 1.216387766878246,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004863677296074311,
      "loss": 5.9524,
      "mean_token_accuracy": 0.18110742717981337,
      "num_tokens": 26730781.0,
      "step": 10540
    },
    {
      "entropy": 6.186445093154907,
      "epoch": 1.2169648009232545,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004863535810539686,
      "loss": 5.8922,
      "mean_token_accuracy": 0.1831481859087944,
      "num_tokens": 26743661.0,
      "step": 10545
    },
    {
      "entropy": 6.173359346389771,
      "epoch": 1.2175418349682632,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048633942539173325,
      "loss": 5.8275,
      "mean_token_accuracy": 0.18538701087236403,
      "num_tokens": 26755307.0,
      "step": 10550
    },
    {
      "entropy": 6.202890348434448,
      "epoch": 1.2181188690132718,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048632526262120144,
      "loss": 5.8605,
      "mean_token_accuracy": 0.18347341567277908,
      "num_tokens": 26769679.0,
      "step": 10555
    },
    {
      "entropy": 6.191610479354859,
      "epoch": 1.2186959030582805,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004863110927428493,
      "loss": 5.7906,
      "mean_token_accuracy": 0.1873555064201355,
      "num_tokens": 26782141.0,
      "step": 10560
    },
    {
      "entropy": 6.241343355178833,
      "epoch": 1.219272937103289,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004862969157571536,
      "loss": 5.911,
      "mean_token_accuracy": 0.18459155261516572,
      "num_tokens": 26794721.0,
      "step": 10565
    },
    {
      "entropy": 6.21027102470398,
      "epoch": 1.2198499711482977,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048628273166459105,
      "loss": 5.8213,
      "mean_token_accuracy": 0.19047395139932632,
      "num_tokens": 26807844.0,
      "step": 10570
    },
    {
      "entropy": 6.227565145492553,
      "epoch": 1.2204270051933064,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048626854046563876,
      "loss": 5.9259,
      "mean_token_accuracy": 0.17979119420051576,
      "num_tokens": 26820150.0,
      "step": 10575
    },
    {
      "entropy": 6.240106105804443,
      "epoch": 1.221004039238315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048625434216077404,
      "loss": 5.8868,
      "mean_token_accuracy": 0.18591456562280656,
      "num_tokens": 26833949.0,
      "step": 10580
    },
    {
      "entropy": 6.161479520797729,
      "epoch": 1.2215810732833237,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00048624013675047453,
      "loss": 5.7959,
      "mean_token_accuracy": 0.18247667253017424,
      "num_tokens": 26847637.0,
      "step": 10585
    },
    {
      "entropy": 6.214194679260254,
      "epoch": 1.2221581073283323,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048622592423521783,
      "loss": 5.8994,
      "mean_token_accuracy": 0.18749276101589202,
      "num_tokens": 26859173.0,
      "step": 10590
    },
    {
      "entropy": 6.244353103637695,
      "epoch": 1.222735141373341,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004862117046154822,
      "loss": 5.9012,
      "mean_token_accuracy": 0.18018726706504823,
      "num_tokens": 26870883.0,
      "step": 10595
    },
    {
      "entropy": 6.300807285308838,
      "epoch": 1.2233121754183496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048619747789174577,
      "loss": 5.9537,
      "mean_token_accuracy": 0.18005219101905823,
      "num_tokens": 26882813.0,
      "step": 10600
    },
    {
      "entropy": 6.131255769729615,
      "epoch": 1.2238892094633584,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004861832440644871,
      "loss": 5.8613,
      "mean_token_accuracy": 0.1861237496137619,
      "num_tokens": 26893928.0,
      "step": 10605
    },
    {
      "entropy": 6.113092756271362,
      "epoch": 1.224466243508367,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048616900313418486,
      "loss": 5.8353,
      "mean_token_accuracy": 0.1916848599910736,
      "num_tokens": 26907431.0,
      "step": 10610
    },
    {
      "entropy": 6.303126001358033,
      "epoch": 1.2250432775533757,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048615475510131815,
      "loss": 5.9126,
      "mean_token_accuracy": 0.1822982147336006,
      "num_tokens": 26920278.0,
      "step": 10615
    },
    {
      "entropy": 6.298568868637085,
      "epoch": 1.2256203115983844,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048614049996636614,
      "loss": 6.002,
      "mean_token_accuracy": 0.17795106321573256,
      "num_tokens": 26933348.0,
      "step": 10620
    },
    {
      "entropy": 6.2169867038726805,
      "epoch": 1.226197345643393,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048612623772980825,
      "loss": 5.9471,
      "mean_token_accuracy": 0.17671473771333696,
      "num_tokens": 26946017.0,
      "step": 10625
    },
    {
      "entropy": 6.132002639770508,
      "epoch": 1.2267743796884016,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048611196839212426,
      "loss": 5.6903,
      "mean_token_accuracy": 0.19564807564020156,
      "num_tokens": 26960230.0,
      "step": 10630
    },
    {
      "entropy": 6.2894245147705075,
      "epoch": 1.2273514137334103,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000486097691953794,
      "loss": 5.9397,
      "mean_token_accuracy": 0.17599738091230394,
      "num_tokens": 26971811.0,
      "step": 10635
    },
    {
      "entropy": 6.284972667694092,
      "epoch": 1.227928447778419,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048608340841529784,
      "loss": 5.9203,
      "mean_token_accuracy": 0.18293810933828353,
      "num_tokens": 26983098.0,
      "step": 10640
    },
    {
      "entropy": 6.240820074081421,
      "epoch": 1.2285054818234276,
      "grad_norm": 0.921875,
      "learning_rate": 0.000486069117777116,
      "loss": 5.9231,
      "mean_token_accuracy": 0.1746153563261032,
      "num_tokens": 26996410.0,
      "step": 10645
    },
    {
      "entropy": 6.319471168518066,
      "epoch": 1.2290825158684362,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004860548200397293,
      "loss": 5.9207,
      "mean_token_accuracy": 0.17919676005840302,
      "num_tokens": 27009329.0,
      "step": 10650
    },
    {
      "entropy": 6.196267890930176,
      "epoch": 1.2296595499134448,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048604051520361846,
      "loss": 5.8489,
      "mean_token_accuracy": 0.18131981194019317,
      "num_tokens": 27022261.0,
      "step": 10655
    },
    {
      "entropy": 6.207855939865112,
      "epoch": 1.2302365839584535,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048602620326926484,
      "loss": 5.9439,
      "mean_token_accuracy": 0.1832319661974907,
      "num_tokens": 27035006.0,
      "step": 10660
    },
    {
      "entropy": 6.195567846298218,
      "epoch": 1.2308136180034621,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048601188423714965,
      "loss": 5.8238,
      "mean_token_accuracy": 0.1868258848786354,
      "num_tokens": 27048481.0,
      "step": 10665
    },
    {
      "entropy": 6.300758409500122,
      "epoch": 1.231390652048471,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004859975581077545,
      "loss": 5.9905,
      "mean_token_accuracy": 0.17319414615631104,
      "num_tokens": 27061083.0,
      "step": 10670
    },
    {
      "entropy": 6.235935974121094,
      "epoch": 1.2319676860934794,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004859832248815614,
      "loss": 5.9045,
      "mean_token_accuracy": 0.18003612607717515,
      "num_tokens": 27073547.0,
      "step": 10675
    },
    {
      "entropy": 6.194489908218384,
      "epoch": 1.2325447201384883,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004859688845590522,
      "loss": 5.8974,
      "mean_token_accuracy": 0.18143948465585708,
      "num_tokens": 27085739.0,
      "step": 10680
    },
    {
      "entropy": 6.187116813659668,
      "epoch": 1.233121754183497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048595453714070944,
      "loss": 5.88,
      "mean_token_accuracy": 0.18256208598613738,
      "num_tokens": 27097331.0,
      "step": 10685
    },
    {
      "entropy": 6.350992250442505,
      "epoch": 1.2336987882285055,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004859401826270156,
      "loss": 5.9313,
      "mean_token_accuracy": 0.18188985139131547,
      "num_tokens": 27109181.0,
      "step": 10690
    },
    {
      "entropy": 6.230800533294678,
      "epoch": 1.2342758222735142,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004859258210184536,
      "loss": 5.8251,
      "mean_token_accuracy": 0.18658582121133804,
      "num_tokens": 27120978.0,
      "step": 10695
    },
    {
      "entropy": 6.2771892070770265,
      "epoch": 1.2348528563185228,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048591145231550623,
      "loss": 5.8354,
      "mean_token_accuracy": 0.18821250200271605,
      "num_tokens": 27135243.0,
      "step": 10700
    },
    {
      "entropy": 6.128650569915772,
      "epoch": 1.2354298903635315,
      "grad_norm": 0.765625,
      "learning_rate": 0.00048589707651865697,
      "loss": 5.8103,
      "mean_token_accuracy": 0.189788256585598,
      "num_tokens": 27149366.0,
      "step": 10705
    },
    {
      "entropy": 6.217971324920654,
      "epoch": 1.23600692440854,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004858826936283893,
      "loss": 5.9109,
      "mean_token_accuracy": 0.17731088995933533,
      "num_tokens": 27162552.0,
      "step": 10710
    },
    {
      "entropy": 6.217400026321411,
      "epoch": 1.2365839584535487,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000485868303645187,
      "loss": 5.8907,
      "mean_token_accuracy": 0.18079452216625214,
      "num_tokens": 27175312.0,
      "step": 10715
    },
    {
      "entropy": 6.233800840377808,
      "epoch": 1.2371609924985574,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048585390656953397,
      "loss": 5.8684,
      "mean_token_accuracy": 0.1765349343419075,
      "num_tokens": 27187730.0,
      "step": 10720
    },
    {
      "entropy": 6.221810245513916,
      "epoch": 1.237738026543566,
      "grad_norm": 0.875,
      "learning_rate": 0.0004858395024019147,
      "loss": 5.8529,
      "mean_token_accuracy": 0.18980882316827774,
      "num_tokens": 27200191.0,
      "step": 10725
    },
    {
      "entropy": 6.221706438064575,
      "epoch": 1.2383150605885747,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004858250911428133,
      "loss": 5.9038,
      "mean_token_accuracy": 0.18302361220121383,
      "num_tokens": 27211905.0,
      "step": 10730
    },
    {
      "entropy": 6.225879859924317,
      "epoch": 1.2388920946335833,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004858106727927148,
      "loss": 5.87,
      "mean_token_accuracy": 0.18463122099637985,
      "num_tokens": 27224153.0,
      "step": 10735
    },
    {
      "entropy": 6.233470916748047,
      "epoch": 1.239469128678592,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000485796247352104,
      "loss": 5.89,
      "mean_token_accuracy": 0.17813801318407058,
      "num_tokens": 27237167.0,
      "step": 10740
    },
    {
      "entropy": 6.235151910781861,
      "epoch": 1.2400461627236008,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004857818148214662,
      "loss": 5.9002,
      "mean_token_accuracy": 0.18422091603279114,
      "num_tokens": 27247956.0,
      "step": 10745
    },
    {
      "entropy": 6.242047309875488,
      "epoch": 1.2406231967686094,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004857673752012866,
      "loss": 5.924,
      "mean_token_accuracy": 0.17710370868444442,
      "num_tokens": 27260357.0,
      "step": 10750
    },
    {
      "entropy": 6.219243669509888,
      "epoch": 1.241200230813618,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048575292849205114,
      "loss": 5.822,
      "mean_token_accuracy": 0.18197387605905532,
      "num_tokens": 27273953.0,
      "step": 10755
    },
    {
      "entropy": 6.210616397857666,
      "epoch": 1.2417772648586267,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004857384746942456,
      "loss": 5.8956,
      "mean_token_accuracy": 0.1815459206700325,
      "num_tokens": 27285550.0,
      "step": 10760
    },
    {
      "entropy": 6.28066577911377,
      "epoch": 1.2423542989036354,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004857240138083562,
      "loss": 5.962,
      "mean_token_accuracy": 0.17763375639915466,
      "num_tokens": 27297817.0,
      "step": 10765
    },
    {
      "entropy": 6.252126264572143,
      "epoch": 1.242931332948644,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004857095458348692,
      "loss": 5.9957,
      "mean_token_accuracy": 0.17702541649341583,
      "num_tokens": 27310398.0,
      "step": 10770
    },
    {
      "entropy": 6.253581094741821,
      "epoch": 1.2435083669936526,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048569507077427134,
      "loss": 5.9272,
      "mean_token_accuracy": 0.17670738697052002,
      "num_tokens": 27322462.0,
      "step": 10775
    },
    {
      "entropy": 6.377768230438233,
      "epoch": 1.2440854010386613,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004856805886270494,
      "loss": 5.9682,
      "mean_token_accuracy": 0.1788796678185463,
      "num_tokens": 27335301.0,
      "step": 10780
    },
    {
      "entropy": 6.179411315917969,
      "epoch": 1.24466243508367,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004856660993936905,
      "loss": 5.8209,
      "mean_token_accuracy": 0.1899016574025154,
      "num_tokens": 27347425.0,
      "step": 10785
    },
    {
      "entropy": 6.2287742614746096,
      "epoch": 1.2452394691286786,
      "grad_norm": 0.859375,
      "learning_rate": 0.000485651603074682,
      "loss": 5.9149,
      "mean_token_accuracy": 0.18241002708673476,
      "num_tokens": 27361149.0,
      "step": 10790
    },
    {
      "entropy": 6.24484658241272,
      "epoch": 1.2458165031736872,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004856370996705115,
      "loss": 5.8894,
      "mean_token_accuracy": 0.1860358715057373,
      "num_tokens": 27373055.0,
      "step": 10795
    },
    {
      "entropy": 6.232689905166626,
      "epoch": 1.2463935372186958,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004856225891816667,
      "loss": 5.9433,
      "mean_token_accuracy": 0.18515971004962922,
      "num_tokens": 27385279.0,
      "step": 10800
    },
    {
      "entropy": 6.257841300964356,
      "epoch": 1.2469705712637045,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004856080716086358,
      "loss": 5.9111,
      "mean_token_accuracy": 0.18471490144729613,
      "num_tokens": 27399035.0,
      "step": 10805
    },
    {
      "entropy": 6.265076112747193,
      "epoch": 1.2475476053087133,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000485593546951907,
      "loss": 5.9207,
      "mean_token_accuracy": 0.18478413075208663,
      "num_tokens": 27410473.0,
      "step": 10810
    },
    {
      "entropy": 6.261572790145874,
      "epoch": 1.2481246393537218,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004855790152119688,
      "loss": 5.8996,
      "mean_token_accuracy": 0.1812854066491127,
      "num_tokens": 27422194.0,
      "step": 10815
    },
    {
      "entropy": 6.227007627487183,
      "epoch": 1.2487016733987306,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004855644763893102,
      "loss": 5.8966,
      "mean_token_accuracy": 0.18256665617227555,
      "num_tokens": 27434800.0,
      "step": 10820
    },
    {
      "entropy": 6.121037817001342,
      "epoch": 1.2492787074437393,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048554993048441987,
      "loss": 5.7659,
      "mean_token_accuracy": 0.19123974889516832,
      "num_tokens": 27447002.0,
      "step": 10825
    },
    {
      "entropy": 6.280801677703858,
      "epoch": 1.249855741488748,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004855353774977873,
      "loss": 5.985,
      "mean_token_accuracy": 0.17724091708660125,
      "num_tokens": 27461461.0,
      "step": 10830
    },
    {
      "entropy": 6.275634384155273,
      "epoch": 1.2504327755337565,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048552081742990184,
      "loss": 5.9565,
      "mean_token_accuracy": 0.17275320887565612,
      "num_tokens": 27475062.0,
      "step": 10835
    },
    {
      "entropy": 6.232098817825317,
      "epoch": 1.2510098095787652,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048550625028125327,
      "loss": 5.8458,
      "mean_token_accuracy": 0.19013809859752656,
      "num_tokens": 27487812.0,
      "step": 10840
    },
    {
      "entropy": 6.19635591506958,
      "epoch": 1.2515868436237738,
      "grad_norm": 0.875,
      "learning_rate": 0.0004854916760523315,
      "loss": 5.8639,
      "mean_token_accuracy": 0.1898317366838455,
      "num_tokens": 27501351.0,
      "step": 10845
    },
    {
      "entropy": 6.214985466003418,
      "epoch": 1.2521638776687825,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048547709474362684,
      "loss": 5.8483,
      "mean_token_accuracy": 0.18931108862161636,
      "num_tokens": 27514651.0,
      "step": 10850
    },
    {
      "entropy": 6.291312026977539,
      "epoch": 1.252740911713791,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004854625063556296,
      "loss": 5.972,
      "mean_token_accuracy": 0.17728905528783798,
      "num_tokens": 27527037.0,
      "step": 10855
    },
    {
      "entropy": 6.272031641006469,
      "epoch": 1.2533179457587997,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004854479108888305,
      "loss": 5.9503,
      "mean_token_accuracy": 0.18258428275585176,
      "num_tokens": 27540800.0,
      "step": 10860
    },
    {
      "entropy": 6.2498390674591064,
      "epoch": 1.2538949798038084,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048543330834372047,
      "loss": 5.8751,
      "mean_token_accuracy": 0.183306285738945,
      "num_tokens": 27553873.0,
      "step": 10865
    },
    {
      "entropy": 6.1702086448669435,
      "epoch": 1.254472013848817,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048541869872079064,
      "loss": 5.8717,
      "mean_token_accuracy": 0.18380814492702485,
      "num_tokens": 27566336.0,
      "step": 10870
    },
    {
      "entropy": 6.31038818359375,
      "epoch": 1.2550490478938257,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004854040820205324,
      "loss": 5.9946,
      "mean_token_accuracy": 0.1781650796532631,
      "num_tokens": 27578677.0,
      "step": 10875
    },
    {
      "entropy": 6.318043756484985,
      "epoch": 1.2556260819388343,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004853894582434373,
      "loss": 5.9378,
      "mean_token_accuracy": 0.1817599982023239,
      "num_tokens": 27590734.0,
      "step": 10880
    },
    {
      "entropy": 6.210410022735596,
      "epoch": 1.2562031159838432,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004853748273899974,
      "loss": 5.8776,
      "mean_token_accuracy": 0.17782250940799713,
      "num_tokens": 27603193.0,
      "step": 10885
    },
    {
      "entropy": 6.226425313949585,
      "epoch": 1.2567801500288516,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004853601894607046,
      "loss": 5.847,
      "mean_token_accuracy": 0.1877206265926361,
      "num_tokens": 27615248.0,
      "step": 10890
    },
    {
      "entropy": 6.278065395355225,
      "epoch": 1.2573571840738604,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004853455444560514,
      "loss": 5.9262,
      "mean_token_accuracy": 0.18514909446239472,
      "num_tokens": 27627384.0,
      "step": 10895
    },
    {
      "entropy": 6.228753566741943,
      "epoch": 1.257934218118869,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004853308923765302,
      "loss": 5.9141,
      "mean_token_accuracy": 0.1801733896136284,
      "num_tokens": 27641014.0,
      "step": 10900
    },
    {
      "entropy": 6.242538690567017,
      "epoch": 1.2585112521638777,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000485316233222634,
      "loss": 5.7988,
      "mean_token_accuracy": 0.19334478676319122,
      "num_tokens": 27652325.0,
      "step": 10905
    },
    {
      "entropy": 6.256791639328003,
      "epoch": 1.2590882862088864,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004853015669948557,
      "loss": 5.8981,
      "mean_token_accuracy": 0.18492160141468048,
      "num_tokens": 27664365.0,
      "step": 10910
    },
    {
      "entropy": 6.185790157318115,
      "epoch": 1.259665320253895,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048528689369368863,
      "loss": 5.7989,
      "mean_token_accuracy": 0.18811622709035875,
      "num_tokens": 27677657.0,
      "step": 10915
    },
    {
      "entropy": 6.161892127990723,
      "epoch": 1.2602423542989036,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004852722133196264,
      "loss": 5.7452,
      "mean_token_accuracy": 0.19919943064451218,
      "num_tokens": 27689598.0,
      "step": 10920
    },
    {
      "entropy": 6.165025758743286,
      "epoch": 1.2608193883439123,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004852575258731627,
      "loss": 5.7537,
      "mean_token_accuracy": 0.19810103625059128,
      "num_tokens": 27702573.0,
      "step": 10925
    },
    {
      "entropy": 6.231005859375,
      "epoch": 1.261396422388921,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004852428313547916,
      "loss": 5.9183,
      "mean_token_accuracy": 0.18382099121809006,
      "num_tokens": 27714506.0,
      "step": 10930
    },
    {
      "entropy": 6.0321409702301025,
      "epoch": 1.2619734564339296,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048522812976500726,
      "loss": 5.6025,
      "mean_token_accuracy": 0.2134798437356949,
      "num_tokens": 27728308.0,
      "step": 10935
    },
    {
      "entropy": 6.202177906036377,
      "epoch": 1.2625504904789382,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048521342110430417,
      "loss": 5.9004,
      "mean_token_accuracy": 0.18098054826259613,
      "num_tokens": 27741521.0,
      "step": 10940
    },
    {
      "entropy": 6.304830408096313,
      "epoch": 1.2631275245239468,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048519870537317713,
      "loss": 5.9345,
      "mean_token_accuracy": 0.17498082369565965,
      "num_tokens": 27754033.0,
      "step": 10945
    },
    {
      "entropy": 6.272612190246582,
      "epoch": 1.2637045585689557,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048518398257212103,
      "loss": 5.9259,
      "mean_token_accuracy": 0.18497760444879532,
      "num_tokens": 27766557.0,
      "step": 10950
    },
    {
      "entropy": 6.227681350708008,
      "epoch": 1.2642815926139641,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004851692527016311,
      "loss": 5.8712,
      "mean_token_accuracy": 0.19011031240224838,
      "num_tokens": 27778018.0,
      "step": 10955
    },
    {
      "entropy": 6.228446912765503,
      "epoch": 1.264858626658973,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004851545157622027,
      "loss": 5.8279,
      "mean_token_accuracy": 0.18843198865652083,
      "num_tokens": 27789514.0,
      "step": 10960
    },
    {
      "entropy": 6.25038104057312,
      "epoch": 1.2654356607039816,
      "grad_norm": 0.875,
      "learning_rate": 0.0004851397717543316,
      "loss": 5.8994,
      "mean_token_accuracy": 0.17876967042684555,
      "num_tokens": 27802759.0,
      "step": 10965
    },
    {
      "entropy": 6.183712816238403,
      "epoch": 1.2660126947489903,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004851250206785137,
      "loss": 5.8294,
      "mean_token_accuracy": 0.18825262486934663,
      "num_tokens": 27817064.0,
      "step": 10970
    },
    {
      "entropy": 6.237206506729126,
      "epoch": 1.266589728793999,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048511026253524503,
      "loss": 5.8577,
      "mean_token_accuracy": 0.18066271692514418,
      "num_tokens": 27830627.0,
      "step": 10975
    },
    {
      "entropy": 6.151304769515991,
      "epoch": 1.2671667628390075,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004850954973250221,
      "loss": 5.7718,
      "mean_token_accuracy": 0.19892093986272813,
      "num_tokens": 27842745.0,
      "step": 10980
    },
    {
      "entropy": 6.248844051361084,
      "epoch": 1.2677437968840162,
      "grad_norm": 0.875,
      "learning_rate": 0.00048508072504834144,
      "loss": 5.9213,
      "mean_token_accuracy": 0.18295619189739226,
      "num_tokens": 27855720.0,
      "step": 10985
    },
    {
      "entropy": 6.184025382995605,
      "epoch": 1.2683208309290248,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048506594570569997,
      "loss": 5.8103,
      "mean_token_accuracy": 0.1965300992131233,
      "num_tokens": 27869019.0,
      "step": 10990
    },
    {
      "entropy": 6.252641057968139,
      "epoch": 1.2688978649740335,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004850511592975947,
      "loss": 5.9526,
      "mean_token_accuracy": 0.17748470455408097,
      "num_tokens": 27883563.0,
      "step": 10995
    },
    {
      "entropy": 6.191432476043701,
      "epoch": 1.269474899019042,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850363658245231,
      "loss": 5.8903,
      "mean_token_accuracy": 0.18608094453811647,
      "num_tokens": 27895980.0,
      "step": 11000
    },
    {
      "entropy": 6.209790563583374,
      "epoch": 1.2700519330640507,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004850215652869826,
      "loss": 5.9486,
      "mean_token_accuracy": 0.17304446399211884,
      "num_tokens": 27907535.0,
      "step": 11005
    },
    {
      "entropy": 6.351325798034668,
      "epoch": 1.2706289671090594,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004850067576854711,
      "loss": 5.8758,
      "mean_token_accuracy": 0.1832320436835289,
      "num_tokens": 27920665.0,
      "step": 11010
    },
    {
      "entropy": 6.1678464889526365,
      "epoch": 1.271206001154068,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004849919430204867,
      "loss": 5.7929,
      "mean_token_accuracy": 0.19066481441259384,
      "num_tokens": 27932178.0,
      "step": 11015
    },
    {
      "entropy": 6.1721426486969,
      "epoch": 1.2717830351990767,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004849771212925275,
      "loss": 5.8402,
      "mean_token_accuracy": 0.18766386955976486,
      "num_tokens": 27945319.0,
      "step": 11020
    },
    {
      "entropy": 6.185254335403442,
      "epoch": 1.2723600692440855,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048496229250209223,
      "loss": 5.8514,
      "mean_token_accuracy": 0.18470791578292847,
      "num_tokens": 27957765.0,
      "step": 11025
    },
    {
      "entropy": 6.199320697784424,
      "epoch": 1.272937103289094,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004849474566496795,
      "loss": 5.8478,
      "mean_token_accuracy": 0.19055087566375734,
      "num_tokens": 27971306.0,
      "step": 11030
    },
    {
      "entropy": 6.1969986915588375,
      "epoch": 1.2735141373341028,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004849326137357883,
      "loss": 5.8027,
      "mean_token_accuracy": 0.1879406288266182,
      "num_tokens": 27982998.0,
      "step": 11035
    },
    {
      "entropy": 6.202544403076172,
      "epoch": 1.2740911713791114,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000484917763760918,
      "loss": 5.8982,
      "mean_token_accuracy": 0.1865037515759468,
      "num_tokens": 27995792.0,
      "step": 11040
    },
    {
      "entropy": 6.141378736495971,
      "epoch": 1.27466820542412,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048490290672556784,
      "loss": 5.7892,
      "mean_token_accuracy": 0.18931265473365783,
      "num_tokens": 28008882.0,
      "step": 11045
    },
    {
      "entropy": 6.212600755691528,
      "epoch": 1.2752452394691287,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004848880426302378,
      "loss": 5.9074,
      "mean_token_accuracy": 0.1816494032740593,
      "num_tokens": 28022768.0,
      "step": 11050
    },
    {
      "entropy": 6.20895791053772,
      "epoch": 1.2758222735141374,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004848731714754277,
      "loss": 5.8502,
      "mean_token_accuracy": 0.184775310754776,
      "num_tokens": 28035196.0,
      "step": 11055
    },
    {
      "entropy": 6.343476963043213,
      "epoch": 1.276399307559146,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004848582932616377,
      "loss": 6.0182,
      "mean_token_accuracy": 0.17365770190954208,
      "num_tokens": 28048089.0,
      "step": 11060
    },
    {
      "entropy": 6.255667972564697,
      "epoch": 1.2769763416041546,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004848434079893682,
      "loss": 5.9318,
      "mean_token_accuracy": 0.17477375119924546,
      "num_tokens": 28060140.0,
      "step": 11065
    },
    {
      "entropy": 6.303800296783447,
      "epoch": 1.2775533756491633,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004848285156591201,
      "loss": 5.8979,
      "mean_token_accuracy": 0.18160488307476044,
      "num_tokens": 28072448.0,
      "step": 11070
    },
    {
      "entropy": 6.316568803787232,
      "epoch": 1.278130409694172,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048481361627139384,
      "loss": 5.9613,
      "mean_token_accuracy": 0.17848495095968248,
      "num_tokens": 28084696.0,
      "step": 11075
    },
    {
      "entropy": 6.255594921112061,
      "epoch": 1.2787074437391805,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048479870982669096,
      "loss": 5.9127,
      "mean_token_accuracy": 0.18629123866558076,
      "num_tokens": 28097834.0,
      "step": 11080
    },
    {
      "entropy": 6.212059211730957,
      "epoch": 1.2792844777841892,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048478379632551266,
      "loss": 5.7972,
      "mean_token_accuracy": 0.1920081004500389,
      "num_tokens": 28110066.0,
      "step": 11085
    },
    {
      "entropy": 6.21475510597229,
      "epoch": 1.279861511829198,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004847688757683606,
      "loss": 5.9232,
      "mean_token_accuracy": 0.180953311920166,
      "num_tokens": 28122617.0,
      "step": 11090
    },
    {
      "entropy": 6.178725290298462,
      "epoch": 1.2804385458742065,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004847539481557366,
      "loss": 5.8263,
      "mean_token_accuracy": 0.18228112757205964,
      "num_tokens": 28134987.0,
      "step": 11095
    },
    {
      "entropy": 6.293797159194947,
      "epoch": 1.2810155799192153,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048473901348814277,
      "loss": 5.9023,
      "mean_token_accuracy": 0.17612725496292114,
      "num_tokens": 28147478.0,
      "step": 11100
    },
    {
      "entropy": 6.264359474182129,
      "epoch": 1.281592613964224,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004847240717660814,
      "loss": 5.9125,
      "mean_token_accuracy": 0.18425216227769853,
      "num_tokens": 28159010.0,
      "step": 11105
    },
    {
      "entropy": 6.238105440139771,
      "epoch": 1.2821696480092326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048470912299005493,
      "loss": 5.9471,
      "mean_token_accuracy": 0.18064728528261184,
      "num_tokens": 28170579.0,
      "step": 11110
    },
    {
      "entropy": 6.258795118331909,
      "epoch": 1.2827466820542412,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048469416716056635,
      "loss": 5.8843,
      "mean_token_accuracy": 0.18427431136369704,
      "num_tokens": 28182569.0,
      "step": 11115
    },
    {
      "entropy": 6.280900859832764,
      "epoch": 1.28332371609925,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004846792042781187,
      "loss": 5.9565,
      "mean_token_accuracy": 0.1773100271821022,
      "num_tokens": 28195736.0,
      "step": 11120
    },
    {
      "entropy": 6.253517961502075,
      "epoch": 1.2839007501442585,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048466423434321513,
      "loss": 5.926,
      "mean_token_accuracy": 0.18404542058706283,
      "num_tokens": 28210944.0,
      "step": 11125
    },
    {
      "entropy": 6.2450648784637455,
      "epoch": 1.2844777841892672,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048464925735635907,
      "loss": 5.8733,
      "mean_token_accuracy": 0.18082726001739502,
      "num_tokens": 28223674.0,
      "step": 11130
    },
    {
      "entropy": 6.2887615203857425,
      "epoch": 1.2850548182342758,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048463427331805445,
      "loss": 5.94,
      "mean_token_accuracy": 0.17404428273439407,
      "num_tokens": 28236643.0,
      "step": 11135
    },
    {
      "entropy": 6.275627374649048,
      "epoch": 1.2856318522792844,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004846192822288052,
      "loss": 5.9229,
      "mean_token_accuracy": 0.17755234092473984,
      "num_tokens": 28249678.0,
      "step": 11140
    },
    {
      "entropy": 6.224632358551025,
      "epoch": 1.286208886324293,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004846042840891155,
      "loss": 5.8678,
      "mean_token_accuracy": 0.18753730952739717,
      "num_tokens": 28262571.0,
      "step": 11145
    },
    {
      "entropy": 6.181311321258545,
      "epoch": 1.2867859203693017,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004845892788994899,
      "loss": 5.8787,
      "mean_token_accuracy": 0.179297511279583,
      "num_tokens": 28275450.0,
      "step": 11150
    },
    {
      "entropy": 6.265094852447509,
      "epoch": 1.2873629544143104,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004845742666604329,
      "loss": 5.8228,
      "mean_token_accuracy": 0.1867560401558876,
      "num_tokens": 28288224.0,
      "step": 11155
    },
    {
      "entropy": 6.103898048400879,
      "epoch": 1.287939988459319,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048455924737244953,
      "loss": 5.6738,
      "mean_token_accuracy": 0.20230050683021544,
      "num_tokens": 28300515.0,
      "step": 11160
    },
    {
      "entropy": 6.201824378967285,
      "epoch": 1.2885170225043279,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048454422103604505,
      "loss": 5.9034,
      "mean_token_accuracy": 0.18397419452667235,
      "num_tokens": 28313729.0,
      "step": 11165
    },
    {
      "entropy": 6.233884954452515,
      "epoch": 1.2890940565493363,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004845291876517247,
      "loss": 5.8423,
      "mean_token_accuracy": 0.1857234552502632,
      "num_tokens": 28325872.0,
      "step": 11170
    },
    {
      "entropy": 6.268154621124268,
      "epoch": 1.2896710905943451,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048451414721999424,
      "loss": 5.8745,
      "mean_token_accuracy": 0.18568265438079834,
      "num_tokens": 28337820.0,
      "step": 11175
    },
    {
      "entropy": 6.259863185882568,
      "epoch": 1.2902481246393538,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048449909974135954,
      "loss": 5.9131,
      "mean_token_accuracy": 0.1863590285181999,
      "num_tokens": 28350608.0,
      "step": 11180
    },
    {
      "entropy": 6.207197570800782,
      "epoch": 1.2908251586843624,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004844840452163267,
      "loss": 5.8339,
      "mean_token_accuracy": 0.18503543436527253,
      "num_tokens": 28362229.0,
      "step": 11185
    },
    {
      "entropy": 6.210993480682373,
      "epoch": 1.291402192729371,
      "grad_norm": 0.921875,
      "learning_rate": 0.000484468983645402,
      "loss": 5.8131,
      "mean_token_accuracy": 0.189563025534153,
      "num_tokens": 28374767.0,
      "step": 11190
    },
    {
      "entropy": 6.185582971572876,
      "epoch": 1.2919792267743797,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048445391502909213,
      "loss": 5.777,
      "mean_token_accuracy": 0.18247984796762468,
      "num_tokens": 28387401.0,
      "step": 11195
    },
    {
      "entropy": 6.194123315811157,
      "epoch": 1.2925562608193883,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048443883936790386,
      "loss": 5.8478,
      "mean_token_accuracy": 0.19127041101455688,
      "num_tokens": 28400123.0,
      "step": 11200
    },
    {
      "entropy": 6.253632545471191,
      "epoch": 1.293133294864397,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048442375666234427,
      "loss": 5.9125,
      "mean_token_accuracy": 0.1879119783639908,
      "num_tokens": 28411607.0,
      "step": 11205
    },
    {
      "entropy": 6.284770917892456,
      "epoch": 1.2937103289094056,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004844086669129206,
      "loss": 5.8841,
      "mean_token_accuracy": 0.17771966606378556,
      "num_tokens": 28424299.0,
      "step": 11210
    },
    {
      "entropy": 6.307318925857544,
      "epoch": 1.2942873629544143,
      "grad_norm": 0.875,
      "learning_rate": 0.00048439357012014045,
      "loss": 5.9641,
      "mean_token_accuracy": 0.1830985054373741,
      "num_tokens": 28438646.0,
      "step": 11215
    },
    {
      "entropy": 6.206006145477295,
      "epoch": 1.294864396999423,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004843784662845116,
      "loss": 5.8907,
      "mean_token_accuracy": 0.1812688261270523,
      "num_tokens": 28452870.0,
      "step": 11220
    },
    {
      "entropy": 6.300561141967774,
      "epoch": 1.2954414310444315,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000484363355406542,
      "loss": 5.9626,
      "mean_token_accuracy": 0.17714263945817948,
      "num_tokens": 28465391.0,
      "step": 11225
    },
    {
      "entropy": 6.21189923286438,
      "epoch": 1.2960184650894404,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048434823748674,
      "loss": 5.7972,
      "mean_token_accuracy": 0.18819179981946946,
      "num_tokens": 28477458.0,
      "step": 11230
    },
    {
      "entropy": 6.261479234695434,
      "epoch": 1.2965954991344488,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00048433311252561403,
      "loss": 5.8816,
      "mean_token_accuracy": 0.18019199073314668,
      "num_tokens": 28489092.0,
      "step": 11235
    },
    {
      "entropy": 6.266319799423218,
      "epoch": 1.2971725331794577,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004843179805236728,
      "loss": 5.8879,
      "mean_token_accuracy": 0.18743618875741958,
      "num_tokens": 28501990.0,
      "step": 11240
    },
    {
      "entropy": 6.242443323135376,
      "epoch": 1.297749567224466,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00048430284148142516,
      "loss": 5.8447,
      "mean_token_accuracy": 0.1851040929555893,
      "num_tokens": 28515984.0,
      "step": 11245
    },
    {
      "entropy": 6.177800846099854,
      "epoch": 1.298326601269475,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004842876953993805,
      "loss": 5.8171,
      "mean_token_accuracy": 0.19079625159502028,
      "num_tokens": 28528093.0,
      "step": 11250
    },
    {
      "entropy": 6.243556261062622,
      "epoch": 1.2989036353144836,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004842725422780482,
      "loss": 5.879,
      "mean_token_accuracy": 0.18425228148698808,
      "num_tokens": 28541177.0,
      "step": 11255
    },
    {
      "entropy": 6.187909698486328,
      "epoch": 1.2994806693594922,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004842573821179378,
      "loss": 5.8136,
      "mean_token_accuracy": 0.18927030265331268,
      "num_tokens": 28553033.0,
      "step": 11260
    },
    {
      "entropy": 6.083381700515747,
      "epoch": 1.3000577034045009,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004842422149195593,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1947258949279785,
      "num_tokens": 28565787.0,
      "step": 11265
    },
    {
      "entropy": 6.224492454528809,
      "epoch": 1.3006347374495095,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048422704068342294,
      "loss": 5.8696,
      "mean_token_accuracy": 0.1832335263490677,
      "num_tokens": 28578823.0,
      "step": 11270
    },
    {
      "entropy": 6.2317932605743405,
      "epoch": 1.3012117714945182,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004842118594100389,
      "loss": 5.8472,
      "mean_token_accuracy": 0.18649330884218215,
      "num_tokens": 28590873.0,
      "step": 11275
    },
    {
      "entropy": 6.2860107421875,
      "epoch": 1.3017888055395268,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048419667109991793,
      "loss": 5.8708,
      "mean_token_accuracy": 0.18558044582605362,
      "num_tokens": 28603202.0,
      "step": 11280
    },
    {
      "entropy": 6.266296434402466,
      "epoch": 1.3023658395845354,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048418147575357085,
      "loss": 5.9199,
      "mean_token_accuracy": 0.18047585636377333,
      "num_tokens": 28616803.0,
      "step": 11285
    },
    {
      "entropy": 6.238744926452637,
      "epoch": 1.302942873629544,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004841662733715087,
      "loss": 5.9029,
      "mean_token_accuracy": 0.18507481813430787,
      "num_tokens": 28629439.0,
      "step": 11290
    },
    {
      "entropy": 6.192346525192261,
      "epoch": 1.3035199076745527,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048415106395424294,
      "loss": 5.8385,
      "mean_token_accuracy": 0.19292256236076355,
      "num_tokens": 28642002.0,
      "step": 11295
    },
    {
      "entropy": 6.2264879703521725,
      "epoch": 1.3040969417195614,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048413584750228494,
      "loss": 5.913,
      "mean_token_accuracy": 0.17661560624837874,
      "num_tokens": 28653906.0,
      "step": 11300
    },
    {
      "entropy": 6.250672769546509,
      "epoch": 1.3046739757645702,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048412062401614653,
      "loss": 5.8615,
      "mean_token_accuracy": 0.18393090665340422,
      "num_tokens": 28666085.0,
      "step": 11305
    },
    {
      "entropy": 6.268711185455322,
      "epoch": 1.3052510098095786,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048410539349634,
      "loss": 5.893,
      "mean_token_accuracy": 0.1831775650382042,
      "num_tokens": 28679298.0,
      "step": 11310
    },
    {
      "entropy": 6.246642208099365,
      "epoch": 1.3058280438545875,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048409015594337725,
      "loss": 5.8228,
      "mean_token_accuracy": 0.1877551034092903,
      "num_tokens": 28691146.0,
      "step": 11315
    },
    {
      "entropy": 6.1628588199615475,
      "epoch": 1.3064050778995961,
      "grad_norm": 0.796875,
      "learning_rate": 0.000484074911357771,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1942190557718277,
      "num_tokens": 28705259.0,
      "step": 11320
    },
    {
      "entropy": 6.268603658676147,
      "epoch": 1.3069821119446048,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048405965974003404,
      "loss": 5.8633,
      "mean_token_accuracy": 0.18663452118635177,
      "num_tokens": 28718004.0,
      "step": 11325
    },
    {
      "entropy": 6.35073037147522,
      "epoch": 1.3075591459896134,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048404440109067927,
      "loss": 5.9283,
      "mean_token_accuracy": 0.174931201338768,
      "num_tokens": 28730442.0,
      "step": 11330
    },
    {
      "entropy": 6.185829591751099,
      "epoch": 1.308136180034622,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004840291354102198,
      "loss": 5.8559,
      "mean_token_accuracy": 0.1861432209610939,
      "num_tokens": 28743078.0,
      "step": 11335
    },
    {
      "entropy": 6.179704475402832,
      "epoch": 1.3087132140796307,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004840138626991693,
      "loss": 5.807,
      "mean_token_accuracy": 0.19521986246109008,
      "num_tokens": 28756975.0,
      "step": 11340
    },
    {
      "entropy": 6.215914249420166,
      "epoch": 1.3092902481246393,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004839985829580413,
      "loss": 5.8511,
      "mean_token_accuracy": 0.19149716794490815,
      "num_tokens": 28770830.0,
      "step": 11345
    },
    {
      "entropy": 6.169485092163086,
      "epoch": 1.309867282169648,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048398329618734977,
      "loss": 5.8638,
      "mean_token_accuracy": 0.18363995850086212,
      "num_tokens": 28782894.0,
      "step": 11350
    },
    {
      "entropy": 6.25293493270874,
      "epoch": 1.3104443162146566,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004839680023876089,
      "loss": 5.8783,
      "mean_token_accuracy": 0.1869498923420906,
      "num_tokens": 28796412.0,
      "step": 11355
    },
    {
      "entropy": 6.2621642589569095,
      "epoch": 1.3110213502596653,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004839527015593331,
      "loss": 5.8645,
      "mean_token_accuracy": 0.1865290731191635,
      "num_tokens": 28809004.0,
      "step": 11360
    },
    {
      "entropy": 6.1716225147247314,
      "epoch": 1.311598384304674,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048393739370303684,
      "loss": 5.8455,
      "mean_token_accuracy": 0.18741453289985657,
      "num_tokens": 28821104.0,
      "step": 11365
    },
    {
      "entropy": 6.2290661334991455,
      "epoch": 1.3121754183496828,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004839220788192353,
      "loss": 5.8466,
      "mean_token_accuracy": 0.18230463564395905,
      "num_tokens": 28833413.0,
      "step": 11370
    },
    {
      "entropy": 6.251428890228271,
      "epoch": 1.3127524523946912,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00048390675690844335,
      "loss": 5.9039,
      "mean_token_accuracy": 0.18322034180164337,
      "num_tokens": 28846233.0,
      "step": 11375
    },
    {
      "entropy": 6.216500854492187,
      "epoch": 1.3133294864397,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004838914279711764,
      "loss": 5.8916,
      "mean_token_accuracy": 0.18149366080760956,
      "num_tokens": 28859296.0,
      "step": 11380
    },
    {
      "entropy": 6.207100820541382,
      "epoch": 1.3139065204847085,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048387609200795003,
      "loss": 5.8516,
      "mean_token_accuracy": 0.18329866826534272,
      "num_tokens": 28872871.0,
      "step": 11385
    },
    {
      "entropy": 6.25659294128418,
      "epoch": 1.3144835545297173,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048386074901928,
      "loss": 5.862,
      "mean_token_accuracy": 0.1852646842598915,
      "num_tokens": 28885842.0,
      "step": 11390
    },
    {
      "entropy": 6.212846183776856,
      "epoch": 1.315060588574726,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004838453990056825,
      "loss": 5.8577,
      "mean_token_accuracy": 0.18432665020227432,
      "num_tokens": 28898030.0,
      "step": 11395
    },
    {
      "entropy": 6.225255060195923,
      "epoch": 1.3156376226197346,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048383004196767373,
      "loss": 5.9492,
      "mean_token_accuracy": 0.17624325752258302,
      "num_tokens": 28910562.0,
      "step": 11400
    },
    {
      "entropy": 6.380964565277099,
      "epoch": 1.3162146566647432,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004838146779057702,
      "loss": 5.9458,
      "mean_token_accuracy": 0.17678328305482865,
      "num_tokens": 28923169.0,
      "step": 11405
    },
    {
      "entropy": 6.263010120391845,
      "epoch": 1.3167916907097519,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004837993068204887,
      "loss": 5.9334,
      "mean_token_accuracy": 0.18236592710018157,
      "num_tokens": 28935840.0,
      "step": 11410
    },
    {
      "entropy": 6.144847440719604,
      "epoch": 1.3173687247547605,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048378392871234634,
      "loss": 5.7891,
      "mean_token_accuracy": 0.18429743200540544,
      "num_tokens": 28948533.0,
      "step": 11415
    },
    {
      "entropy": 6.21379075050354,
      "epoch": 1.3179457587997692,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004837685435818601,
      "loss": 5.8408,
      "mean_token_accuracy": 0.18431537300348283,
      "num_tokens": 28961060.0,
      "step": 11420
    },
    {
      "entropy": 6.227513217926026,
      "epoch": 1.3185227928447778,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004837531514295477,
      "loss": 5.8388,
      "mean_token_accuracy": 0.19210083335638045,
      "num_tokens": 28973570.0,
      "step": 11425
    },
    {
      "entropy": 6.132026720046997,
      "epoch": 1.3190998268897864,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004837377522559267,
      "loss": 5.8454,
      "mean_token_accuracy": 0.18970469385385513,
      "num_tokens": 28987519.0,
      "step": 11430
    },
    {
      "entropy": 6.258304595947266,
      "epoch": 1.319676860934795,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048372234606151507,
      "loss": 5.8492,
      "mean_token_accuracy": 0.1866730347275734,
      "num_tokens": 29000790.0,
      "step": 11435
    },
    {
      "entropy": 6.2377321243286135,
      "epoch": 1.3202538949798037,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048370693284683106,
      "loss": 5.8719,
      "mean_token_accuracy": 0.17453133761882783,
      "num_tokens": 29013263.0,
      "step": 11440
    },
    {
      "entropy": 6.212386226654052,
      "epoch": 1.3208309290248126,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048369151261239303,
      "loss": 5.8366,
      "mean_token_accuracy": 0.1929737776517868,
      "num_tokens": 29024958.0,
      "step": 11445
    },
    {
      "entropy": 6.214302921295166,
      "epoch": 1.321407963069821,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004836760853587196,
      "loss": 5.8504,
      "mean_token_accuracy": 0.1852180302143097,
      "num_tokens": 29037752.0,
      "step": 11450
    },
    {
      "entropy": 6.211385869979859,
      "epoch": 1.3219849971148299,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048366065108632967,
      "loss": 5.7534,
      "mean_token_accuracy": 0.197591532766819,
      "num_tokens": 29050872.0,
      "step": 11455
    },
    {
      "entropy": 6.272281837463379,
      "epoch": 1.3225620311598385,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048364520979574246,
      "loss": 5.8345,
      "mean_token_accuracy": 0.18267764002084733,
      "num_tokens": 29062472.0,
      "step": 11460
    },
    {
      "entropy": 6.18592324256897,
      "epoch": 1.3231390652048471,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048362976148747715,
      "loss": 5.8723,
      "mean_token_accuracy": 0.1919670134782791,
      "num_tokens": 29075392.0,
      "step": 11465
    },
    {
      "entropy": 6.214108085632324,
      "epoch": 1.3237160992498558,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004836143061620536,
      "loss": 5.8475,
      "mean_token_accuracy": 0.18289434164762497,
      "num_tokens": 29088221.0,
      "step": 11470
    },
    {
      "entropy": 6.248012638092041,
      "epoch": 1.3242931332948644,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004835988438199914,
      "loss": 5.8786,
      "mean_token_accuracy": 0.18585892021656036,
      "num_tokens": 29101050.0,
      "step": 11475
    },
    {
      "entropy": 6.236446809768677,
      "epoch": 1.324870167339873,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004835833744618107,
      "loss": 5.9016,
      "mean_token_accuracy": 0.17933612018823625,
      "num_tokens": 29113928.0,
      "step": 11480
    },
    {
      "entropy": 6.320403051376343,
      "epoch": 1.3254472013848817,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004835678980880318,
      "loss": 5.9642,
      "mean_token_accuracy": 0.1730928048491478,
      "num_tokens": 29127196.0,
      "step": 11485
    },
    {
      "entropy": 6.2192981243133545,
      "epoch": 1.3260242354298903,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004835524146991753,
      "loss": 5.8545,
      "mean_token_accuracy": 0.18383729159832002,
      "num_tokens": 29140498.0,
      "step": 11490
    },
    {
      "entropy": 6.183111572265625,
      "epoch": 1.326601269474899,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048353692429576185,
      "loss": 5.7086,
      "mean_token_accuracy": 0.2016303688287735,
      "num_tokens": 29153910.0,
      "step": 11495
    },
    {
      "entropy": 6.255474472045899,
      "epoch": 1.3271783035199076,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004835214268783126,
      "loss": 5.8995,
      "mean_token_accuracy": 0.1834321767091751,
      "num_tokens": 29166289.0,
      "step": 11500
    },
    {
      "entropy": 6.2197103023529055,
      "epoch": 1.3277553375649163,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048350592244734866,
      "loss": 5.8799,
      "mean_token_accuracy": 0.1853266701102257,
      "num_tokens": 29178321.0,
      "step": 11505
    },
    {
      "entropy": 6.246076345443726,
      "epoch": 1.3283323716099251,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004834904110033917,
      "loss": 5.8603,
      "mean_token_accuracy": 0.18980297744274138,
      "num_tokens": 29190396.0,
      "step": 11510
    },
    {
      "entropy": 6.23483362197876,
      "epoch": 1.3289094056549335,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048347489254696327,
      "loss": 5.8772,
      "mean_token_accuracy": 0.18688696175813674,
      "num_tokens": 29202698.0,
      "step": 11515
    },
    {
      "entropy": 6.149147701263428,
      "epoch": 1.3294864396999424,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004834593670785854,
      "loss": 5.7888,
      "mean_token_accuracy": 0.1987441748380661,
      "num_tokens": 29214508.0,
      "step": 11520
    },
    {
      "entropy": 6.277246332168579,
      "epoch": 1.3300634737449508,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048344383459878024,
      "loss": 5.9011,
      "mean_token_accuracy": 0.1866747960448265,
      "num_tokens": 29225683.0,
      "step": 11525
    },
    {
      "entropy": 6.153271293640136,
      "epoch": 1.3306405077899597,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048342829510807024,
      "loss": 5.7998,
      "mean_token_accuracy": 0.1977728173136711,
      "num_tokens": 29239497.0,
      "step": 11530
    },
    {
      "entropy": 6.213573217391968,
      "epoch": 1.3312175418349683,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000483412748606978,
      "loss": 5.8507,
      "mean_token_accuracy": 0.18818900287151336,
      "num_tokens": 29252230.0,
      "step": 11535
    },
    {
      "entropy": 6.210008716583252,
      "epoch": 1.331794575879977,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004833971950960266,
      "loss": 5.8267,
      "mean_token_accuracy": 0.18866318911314012,
      "num_tokens": 29265349.0,
      "step": 11540
    },
    {
      "entropy": 6.218953847885132,
      "epoch": 1.3323716099249856,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004833816345757391,
      "loss": 5.8652,
      "mean_token_accuracy": 0.1896584689617157,
      "num_tokens": 29278291.0,
      "step": 11545
    },
    {
      "entropy": 6.235172891616822,
      "epoch": 1.3329486439699942,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004833660670466387,
      "loss": 5.8905,
      "mean_token_accuracy": 0.17808593064546585,
      "num_tokens": 29290580.0,
      "step": 11550
    },
    {
      "entropy": 6.303551864624024,
      "epoch": 1.3335256780150029,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004833504925092492,
      "loss": 5.9973,
      "mean_token_accuracy": 0.1788480743765831,
      "num_tokens": 29305738.0,
      "step": 11555
    },
    {
      "entropy": 6.182866048812866,
      "epoch": 1.3341027120600115,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004833349109640944,
      "loss": 5.7962,
      "mean_token_accuracy": 0.19064209908246993,
      "num_tokens": 29316533.0,
      "step": 11560
    },
    {
      "entropy": 6.20551061630249,
      "epoch": 1.3346797461050202,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004833193224116983,
      "loss": 5.8346,
      "mean_token_accuracy": 0.19129739850759506,
      "num_tokens": 29329723.0,
      "step": 11565
    },
    {
      "entropy": 6.207838344573974,
      "epoch": 1.3352567801500288,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048330372685258526,
      "loss": 5.8875,
      "mean_token_accuracy": 0.18001511991024016,
      "num_tokens": 29342830.0,
      "step": 11570
    },
    {
      "entropy": 6.150080060958862,
      "epoch": 1.3358338141950374,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004832881242872799,
      "loss": 5.7713,
      "mean_token_accuracy": 0.1943075478076935,
      "num_tokens": 29354096.0,
      "step": 11575
    },
    {
      "entropy": 6.235388278961182,
      "epoch": 1.336410848240046,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004832725147163069,
      "loss": 5.8362,
      "mean_token_accuracy": 0.1873424157500267,
      "num_tokens": 29367173.0,
      "step": 11580
    },
    {
      "entropy": 6.226995658874512,
      "epoch": 1.336987882285055,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048325689814019134,
      "loss": 5.8474,
      "mean_token_accuracy": 0.1855469673871994,
      "num_tokens": 29378958.0,
      "step": 11585
    },
    {
      "entropy": 6.206862306594848,
      "epoch": 1.3375649163300634,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004832412745594585,
      "loss": 5.8358,
      "mean_token_accuracy": 0.19303939044475554,
      "num_tokens": 29391360.0,
      "step": 11590
    },
    {
      "entropy": 6.199012041091919,
      "epoch": 1.3381419503750722,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048322564397463376,
      "loss": 5.8957,
      "mean_token_accuracy": 0.18181481957435608,
      "num_tokens": 29404040.0,
      "step": 11595
    },
    {
      "entropy": 6.259762001037598,
      "epoch": 1.3387189844200809,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048321000638624296,
      "loss": 5.886,
      "mean_token_accuracy": 0.18082512617111207,
      "num_tokens": 29417960.0,
      "step": 11600
    },
    {
      "entropy": 6.22097544670105,
      "epoch": 1.3392960184650895,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000483194361794812,
      "loss": 5.8838,
      "mean_token_accuracy": 0.18197784274816514,
      "num_tokens": 29431453.0,
      "step": 11605
    },
    {
      "entropy": 6.2116952419281,
      "epoch": 1.3398730525100981,
      "grad_norm": 0.890625,
      "learning_rate": 0.000483178710200867,
      "loss": 5.904,
      "mean_token_accuracy": 0.18123240917921066,
      "num_tokens": 29443716.0,
      "step": 11610
    },
    {
      "entropy": 6.218241786956787,
      "epoch": 1.3404500865551068,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004831630516049346,
      "loss": 5.7846,
      "mean_token_accuracy": 0.1950230598449707,
      "num_tokens": 29455588.0,
      "step": 11615
    },
    {
      "entropy": 6.1530999660491945,
      "epoch": 1.3410271206001154,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004831473860075414,
      "loss": 5.6893,
      "mean_token_accuracy": 0.19766913950443268,
      "num_tokens": 29469472.0,
      "step": 11620
    },
    {
      "entropy": 6.2322492599487305,
      "epoch": 1.341604154645124,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048313171340921417,
      "loss": 5.809,
      "mean_token_accuracy": 0.1921806216239929,
      "num_tokens": 29482432.0,
      "step": 11625
    },
    {
      "entropy": 6.228823041915893,
      "epoch": 1.3421811886901327,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048311603381048025,
      "loss": 5.8078,
      "mean_token_accuracy": 0.19551759511232375,
      "num_tokens": 29495425.0,
      "step": 11630
    },
    {
      "entropy": 6.256144762039185,
      "epoch": 1.3427582227351413,
      "grad_norm": 0.875,
      "learning_rate": 0.0004831003472118668,
      "loss": 5.8569,
      "mean_token_accuracy": 0.19368956983089447,
      "num_tokens": 29507336.0,
      "step": 11635
    },
    {
      "entropy": 6.158148002624512,
      "epoch": 1.34333525678015,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004830846536139016,
      "loss": 5.8096,
      "mean_token_accuracy": 0.19053316712379456,
      "num_tokens": 29519857.0,
      "step": 11640
    },
    {
      "entropy": 6.183690452575684,
      "epoch": 1.3439122908251586,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004830689530171124,
      "loss": 5.8275,
      "mean_token_accuracy": 0.18617817461490632,
      "num_tokens": 29531842.0,
      "step": 11645
    },
    {
      "entropy": 6.257759141921997,
      "epoch": 1.3444893248701673,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004830532454220273,
      "loss": 5.8922,
      "mean_token_accuracy": 0.1862940013408661,
      "num_tokens": 29543721.0,
      "step": 11650
    },
    {
      "entropy": 6.311384057998657,
      "epoch": 1.345066358915176,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048303753082917474,
      "loss": 5.8791,
      "mean_token_accuracy": 0.18588151931762695,
      "num_tokens": 29555713.0,
      "step": 11655
    },
    {
      "entropy": 6.213400554656983,
      "epoch": 1.3456433929601848,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048302180923908306,
      "loss": 5.8924,
      "mean_token_accuracy": 0.18682309836149216,
      "num_tokens": 29568983.0,
      "step": 11660
    },
    {
      "entropy": 6.307500839233398,
      "epoch": 1.3462204270051932,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048300608065228126,
      "loss": 5.8907,
      "mean_token_accuracy": 0.18177999556064606,
      "num_tokens": 29580778.0,
      "step": 11665
    },
    {
      "entropy": 6.26597900390625,
      "epoch": 1.346797461050202,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048299034506929815,
      "loss": 5.8829,
      "mean_token_accuracy": 0.18175012171268462,
      "num_tokens": 29594395.0,
      "step": 11670
    },
    {
      "entropy": 6.110319137573242,
      "epoch": 1.3473744950952107,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048297460249066315,
      "loss": 5.7534,
      "mean_token_accuracy": 0.19732194393873215,
      "num_tokens": 29607588.0,
      "step": 11675
    },
    {
      "entropy": 6.257490873336792,
      "epoch": 1.3479515291402193,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004829588529169057,
      "loss": 5.8178,
      "mean_token_accuracy": 0.18612280637025833,
      "num_tokens": 29618822.0,
      "step": 11680
    },
    {
      "entropy": 6.099374485015869,
      "epoch": 1.348528563185228,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004829430963485555,
      "loss": 5.7855,
      "mean_token_accuracy": 0.19132813513278962,
      "num_tokens": 29631759.0,
      "step": 11685
    },
    {
      "entropy": 6.238083219528198,
      "epoch": 1.3491055972302366,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004829273327861426,
      "loss": 5.8745,
      "mean_token_accuracy": 0.17742790132761002,
      "num_tokens": 29644381.0,
      "step": 11690
    },
    {
      "entropy": 6.249943399429322,
      "epoch": 1.3496826312752452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004829115622301971,
      "loss": 5.8692,
      "mean_token_accuracy": 0.1858292281627655,
      "num_tokens": 29657370.0,
      "step": 11695
    },
    {
      "entropy": 6.246383142471314,
      "epoch": 1.3502596653202539,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048289578468124956,
      "loss": 5.9193,
      "mean_token_accuracy": 0.1809074342250824,
      "num_tokens": 29669438.0,
      "step": 11700
    },
    {
      "entropy": 6.245568752288818,
      "epoch": 1.3508366993652625,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048288000013983046,
      "loss": 5.9064,
      "mean_token_accuracy": 0.18217008113861083,
      "num_tokens": 29681277.0,
      "step": 11705
    },
    {
      "entropy": 6.138575077056885,
      "epoch": 1.3514137334102712,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048286420860647086,
      "loss": 5.7278,
      "mean_token_accuracy": 0.19579086750745772,
      "num_tokens": 29693523.0,
      "step": 11710
    },
    {
      "entropy": 6.325686311721801,
      "epoch": 1.3519907674552798,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048284841008170185,
      "loss": 5.9386,
      "mean_token_accuracy": 0.17788529247045518,
      "num_tokens": 29707589.0,
      "step": 11715
    },
    {
      "entropy": 6.273955011367798,
      "epoch": 1.3525678015002884,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048283260456605487,
      "loss": 5.9469,
      "mean_token_accuracy": 0.18046126067638396,
      "num_tokens": 29719267.0,
      "step": 11720
    },
    {
      "entropy": 6.128788423538208,
      "epoch": 1.3531448355452973,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004828167920600614,
      "loss": 5.7341,
      "mean_token_accuracy": 0.19797903150320054,
      "num_tokens": 29731575.0,
      "step": 11725
    },
    {
      "entropy": 6.258849859237671,
      "epoch": 1.3537218695903057,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004828009725642534,
      "loss": 5.9019,
      "mean_token_accuracy": 0.17968628257513047,
      "num_tokens": 29744762.0,
      "step": 11730
    },
    {
      "entropy": 6.276324510574341,
      "epoch": 1.3542989036353146,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004827851460791628,
      "loss": 5.949,
      "mean_token_accuracy": 0.18110539764165878,
      "num_tokens": 29758300.0,
      "step": 11735
    },
    {
      "entropy": 6.263505649566651,
      "epoch": 1.3548759376803232,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048276931260532213,
      "loss": 5.8362,
      "mean_token_accuracy": 0.18832190632820128,
      "num_tokens": 29771489.0,
      "step": 11740
    },
    {
      "entropy": 6.268861913681031,
      "epoch": 1.3554529717253319,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004827534721432639,
      "loss": 5.9354,
      "mean_token_accuracy": 0.1809426173567772,
      "num_tokens": 29783063.0,
      "step": 11745
    },
    {
      "entropy": 6.257713460922242,
      "epoch": 1.3560300057703405,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004827376246935207,
      "loss": 5.8859,
      "mean_token_accuracy": 0.18558690398931504,
      "num_tokens": 29797136.0,
      "step": 11750
    },
    {
      "entropy": 6.217473936080933,
      "epoch": 1.3566070398153491,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004827217702566257,
      "loss": 5.8472,
      "mean_token_accuracy": 0.18783827424049376,
      "num_tokens": 29809580.0,
      "step": 11755
    },
    {
      "entropy": 6.116797590255738,
      "epoch": 1.3571840738603578,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048270590883311217,
      "loss": 5.7526,
      "mean_token_accuracy": 0.19447557330131532,
      "num_tokens": 29822400.0,
      "step": 11760
    },
    {
      "entropy": 6.2492424011230465,
      "epoch": 1.3577611079053664,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048269004042351363,
      "loss": 5.8556,
      "mean_token_accuracy": 0.18541983067989348,
      "num_tokens": 29834392.0,
      "step": 11765
    },
    {
      "entropy": 6.279932928085327,
      "epoch": 1.358338141950375,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004826741650283637,
      "loss": 5.8702,
      "mean_token_accuracy": 0.1849477082490921,
      "num_tokens": 29846626.0,
      "step": 11770
    },
    {
      "entropy": 6.239205169677734,
      "epoch": 1.3589151759953837,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004826582826481963,
      "loss": 5.8251,
      "mean_token_accuracy": 0.1864626482129097,
      "num_tokens": 29858988.0,
      "step": 11775
    },
    {
      "entropy": 6.131876802444458,
      "epoch": 1.3594922100403923,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004826423932835458,
      "loss": 5.7869,
      "mean_token_accuracy": 0.19552639722824097,
      "num_tokens": 29871875.0,
      "step": 11780
    },
    {
      "entropy": 6.2368231296539305,
      "epoch": 1.360069244085401,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048262649693494653,
      "loss": 5.8254,
      "mean_token_accuracy": 0.19159042537212373,
      "num_tokens": 29884757.0,
      "step": 11785
    },
    {
      "entropy": 6.217382764816284,
      "epoch": 1.3606462781304096,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004826105936029332,
      "loss": 5.8219,
      "mean_token_accuracy": 0.1947940081357956,
      "num_tokens": 29897309.0,
      "step": 11790
    },
    {
      "entropy": 6.2693610191345215,
      "epoch": 1.3612233121754183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004825946832880406,
      "loss": 5.9147,
      "mean_token_accuracy": 0.17783207297325135,
      "num_tokens": 29910005.0,
      "step": 11795
    },
    {
      "entropy": 6.227682161331177,
      "epoch": 1.3618003462204271,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048257876599080404,
      "loss": 5.8764,
      "mean_token_accuracy": 0.1868819698691368,
      "num_tokens": 29921912.0,
      "step": 11800
    },
    {
      "entropy": 6.2378123760223385,
      "epoch": 1.3623773802654355,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048256284171175874,
      "loss": 5.8299,
      "mean_token_accuracy": 0.19049610197544098,
      "num_tokens": 29934399.0,
      "step": 11805
    },
    {
      "entropy": 6.229330158233642,
      "epoch": 1.3629544143104444,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048254691045144035,
      "loss": 5.8085,
      "mean_token_accuracy": 0.18707264810800553,
      "num_tokens": 29946660.0,
      "step": 11810
    },
    {
      "entropy": 6.191214084625244,
      "epoch": 1.363531448355453,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004825309722103848,
      "loss": 5.8303,
      "mean_token_accuracy": 0.18389316648244858,
      "num_tokens": 29958382.0,
      "step": 11815
    },
    {
      "entropy": 6.247525835037232,
      "epoch": 1.3641084824004617,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000482515026989128,
      "loss": 5.8047,
      "mean_token_accuracy": 0.1920720010995865,
      "num_tokens": 29971241.0,
      "step": 11820
    },
    {
      "entropy": 6.300128412246704,
      "epoch": 1.3646855164454703,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048249907478820634,
      "loss": 5.9112,
      "mean_token_accuracy": 0.18333859443664552,
      "num_tokens": 29984145.0,
      "step": 11825
    },
    {
      "entropy": 6.1681116104125975,
      "epoch": 1.365262550490479,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048248311560815637,
      "loss": 5.8,
      "mean_token_accuracy": 0.18586193919181823,
      "num_tokens": 29997092.0,
      "step": 11830
    },
    {
      "entropy": 6.128707933425903,
      "epoch": 1.3658395845354876,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004824671494495149,
      "loss": 5.8445,
      "mean_token_accuracy": 0.18968741744756698,
      "num_tokens": 30009369.0,
      "step": 11835
    },
    {
      "entropy": 6.221187734603882,
      "epoch": 1.3664166185804962,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004824511763128189,
      "loss": 5.813,
      "mean_token_accuracy": 0.18948614001274108,
      "num_tokens": 30021423.0,
      "step": 11840
    },
    {
      "entropy": 6.253768396377564,
      "epoch": 1.3669936526255049,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048243519619860567,
      "loss": 5.8647,
      "mean_token_accuracy": 0.18771067261695862,
      "num_tokens": 30034426.0,
      "step": 11845
    },
    {
      "entropy": 6.196330404281616,
      "epoch": 1.3675706866705135,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004824192091074126,
      "loss": 5.845,
      "mean_token_accuracy": 0.18435929864645004,
      "num_tokens": 30046687.0,
      "step": 11850
    },
    {
      "entropy": 6.265780258178711,
      "epoch": 1.3681477207155222,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004824032150397775,
      "loss": 5.9,
      "mean_token_accuracy": 0.17973716259002687,
      "num_tokens": 30058815.0,
      "step": 11855
    },
    {
      "entropy": 6.3070659160614015,
      "epoch": 1.3687247547605308,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048238721399623824,
      "loss": 5.9097,
      "mean_token_accuracy": 0.17965372353792192,
      "num_tokens": 30070715.0,
      "step": 11860
    },
    {
      "entropy": 6.236549425125122,
      "epoch": 1.3693017888055397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048237120597733316,
      "loss": 5.7561,
      "mean_token_accuracy": 0.19360620826482772,
      "num_tokens": 30083636.0,
      "step": 11865
    },
    {
      "entropy": 6.211022090911865,
      "epoch": 1.369878822850548,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004823551909836005,
      "loss": 5.9072,
      "mean_token_accuracy": 0.18094888627529143,
      "num_tokens": 30094956.0,
      "step": 11870
    },
    {
      "entropy": 6.240145587921143,
      "epoch": 1.370455856895557,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048233916901557896,
      "loss": 5.8076,
      "mean_token_accuracy": 0.18934722393751144,
      "num_tokens": 30107947.0,
      "step": 11875
    },
    {
      "entropy": 6.238693189620972,
      "epoch": 1.3710328909405656,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048232314007380753,
      "loss": 5.9136,
      "mean_token_accuracy": 0.17957172244787217,
      "num_tokens": 30120557.0,
      "step": 11880
    },
    {
      "entropy": 6.262637424468994,
      "epoch": 1.3716099249855742,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048230710415882524,
      "loss": 5.9169,
      "mean_token_accuracy": 0.18944373726844788,
      "num_tokens": 30131734.0,
      "step": 11885
    },
    {
      "entropy": 6.2398604393005375,
      "epoch": 1.3721869590305829,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048229106127117144,
      "loss": 5.8006,
      "mean_token_accuracy": 0.19295659214258193,
      "num_tokens": 30144025.0,
      "step": 11890
    },
    {
      "entropy": 6.238266324996948,
      "epoch": 1.3727639930755915,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004822750114113858,
      "loss": 5.8457,
      "mean_token_accuracy": 0.18792783468961716,
      "num_tokens": 30156246.0,
      "step": 11895
    },
    {
      "entropy": 6.15712604522705,
      "epoch": 1.3733410271206001,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004822589545800081,
      "loss": 5.8274,
      "mean_token_accuracy": 0.1893271267414093,
      "num_tokens": 30169349.0,
      "step": 11900
    },
    {
      "entropy": 6.228726768493653,
      "epoch": 1.3739180611656088,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004822428907775784,
      "loss": 5.8571,
      "mean_token_accuracy": 0.18766702711582184,
      "num_tokens": 30181821.0,
      "step": 11905
    },
    {
      "entropy": 6.265668153762817,
      "epoch": 1.3744950952106174,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048222682000463704,
      "loss": 5.8337,
      "mean_token_accuracy": 0.18967443853616714,
      "num_tokens": 30194278.0,
      "step": 11910
    },
    {
      "entropy": 6.2226804256439205,
      "epoch": 1.375072129255626,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004822107422617245,
      "loss": 5.8665,
      "mean_token_accuracy": 0.1866344466805458,
      "num_tokens": 30206706.0,
      "step": 11915
    },
    {
      "entropy": 6.196132373809815,
      "epoch": 1.3756491633006347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048219465754938156,
      "loss": 5.8492,
      "mean_token_accuracy": 0.18530506491661072,
      "num_tokens": 30220356.0,
      "step": 11920
    },
    {
      "entropy": 6.252609586715698,
      "epoch": 1.3762261973456433,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048217856586814926,
      "loss": 5.8863,
      "mean_token_accuracy": 0.18785286396741868,
      "num_tokens": 30232821.0,
      "step": 11925
    },
    {
      "entropy": 6.09111762046814,
      "epoch": 1.376803231390652,
      "grad_norm": 0.875,
      "learning_rate": 0.00048216246721856875,
      "loss": 5.7848,
      "mean_token_accuracy": 0.1939162567257881,
      "num_tokens": 30246094.0,
      "step": 11930
    },
    {
      "entropy": 6.241026115417481,
      "epoch": 1.3773802654356606,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048214636160118164,
      "loss": 5.856,
      "mean_token_accuracy": 0.19184708297252656,
      "num_tokens": 30259313.0,
      "step": 11935
    },
    {
      "entropy": 6.294254970550537,
      "epoch": 1.3779572994806695,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004821302490165295,
      "loss": 5.9085,
      "mean_token_accuracy": 0.18225693851709365,
      "num_tokens": 30274045.0,
      "step": 11940
    },
    {
      "entropy": 6.212713193893433,
      "epoch": 1.378534333525678,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004821141294651544,
      "loss": 5.8568,
      "mean_token_accuracy": 0.185067056119442,
      "num_tokens": 30286250.0,
      "step": 11945
    },
    {
      "entropy": 6.201496553421021,
      "epoch": 1.3791113675706868,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048209800294759836,
      "loss": 5.8883,
      "mean_token_accuracy": 0.18415758460760118,
      "num_tokens": 30299417.0,
      "step": 11950
    },
    {
      "entropy": 6.248281717300415,
      "epoch": 1.3796884016156954,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004820818694644039,
      "loss": 5.852,
      "mean_token_accuracy": 0.18222470283508302,
      "num_tokens": 30311465.0,
      "step": 11955
    },
    {
      "entropy": 6.121262741088867,
      "epoch": 1.380265435660704,
      "grad_norm": 1.25,
      "learning_rate": 0.00048206572901611364,
      "loss": 5.6793,
      "mean_token_accuracy": 0.1987251579761505,
      "num_tokens": 30325941.0,
      "step": 11960
    },
    {
      "entropy": 6.2388745784759525,
      "epoch": 1.3808424697057127,
      "grad_norm": 0.875,
      "learning_rate": 0.00048204958160327034,
      "loss": 5.9908,
      "mean_token_accuracy": 0.17763442546129227,
      "num_tokens": 30339238.0,
      "step": 11965
    },
    {
      "entropy": 6.240598392486572,
      "epoch": 1.3814195037507213,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048203342722641726,
      "loss": 5.7886,
      "mean_token_accuracy": 0.19409446269273758,
      "num_tokens": 30351082.0,
      "step": 11970
    },
    {
      "entropy": 6.2306938648223875,
      "epoch": 1.38199653779573,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048201726588609776,
      "loss": 5.8305,
      "mean_token_accuracy": 0.18730638772249222,
      "num_tokens": 30364472.0,
      "step": 11975
    },
    {
      "entropy": 6.2045042514801025,
      "epoch": 1.3825735718407386,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048200109758285534,
      "loss": 5.7911,
      "mean_token_accuracy": 0.18847112208604813,
      "num_tokens": 30377113.0,
      "step": 11980
    },
    {
      "entropy": 6.269130611419678,
      "epoch": 1.3831506058857472,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004819849223172337,
      "loss": 5.8809,
      "mean_token_accuracy": 0.1845165580511093,
      "num_tokens": 30389313.0,
      "step": 11985
    },
    {
      "entropy": 6.244358777999878,
      "epoch": 1.3837276399307559,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048196874008977716,
      "loss": 5.9731,
      "mean_token_accuracy": 0.1754239618778229,
      "num_tokens": 30402227.0,
      "step": 11990
    },
    {
      "entropy": 6.274998521804809,
      "epoch": 1.3843046739757645,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004819525509010298,
      "loss": 5.8561,
      "mean_token_accuracy": 0.18385921269655228,
      "num_tokens": 30414719.0,
      "step": 11995
    },
    {
      "entropy": 6.237063407897949,
      "epoch": 1.3848817080207732,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004819363547515361,
      "loss": 5.8991,
      "mean_token_accuracy": 0.18092207461595536,
      "num_tokens": 30427989.0,
      "step": 12000
    },
    {
      "epoch": 1.3848817080207732,
      "eval_entropy": 6.126666735748625,
      "eval_loss": 5.926904201507568,
      "eval_mean_token_accuracy": 0.19014542959556344,
      "eval_num_tokens": 30427989.0,
      "eval_runtime": 17.7416,
      "eval_samples_per_second": 1585.88,
      "eval_steps_per_second": 198.235,
      "step": 12000
    },
    {
      "entropy": 6.249981641769409,
      "epoch": 1.385458742065782,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000481920151641841,
      "loss": 5.8135,
      "mean_token_accuracy": 0.1846992164850235,
      "num_tokens": 30441097.0,
      "step": 12005
    },
    {
      "entropy": 6.104136753082275,
      "epoch": 1.3860357761107904,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048190394157248935,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1966317355632782,
      "num_tokens": 30453535.0,
      "step": 12010
    },
    {
      "entropy": 6.222112226486206,
      "epoch": 1.3866128101557993,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004818877245440264,
      "loss": 5.9097,
      "mean_token_accuracy": 0.1850866824388504,
      "num_tokens": 30466350.0,
      "step": 12015
    },
    {
      "entropy": 6.316772365570069,
      "epoch": 1.3871898442008077,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048187150055699763,
      "loss": 5.9185,
      "mean_token_accuracy": 0.18044357895851135,
      "num_tokens": 30479183.0,
      "step": 12020
    },
    {
      "entropy": 6.24091362953186,
      "epoch": 1.3877668782458166,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004818552696119487,
      "loss": 5.8202,
      "mean_token_accuracy": 0.18946371227502823,
      "num_tokens": 30491903.0,
      "step": 12025
    },
    {
      "entropy": 6.213522291183471,
      "epoch": 1.3883439122908252,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004818390317094255,
      "loss": 5.8122,
      "mean_token_accuracy": 0.1901582047343254,
      "num_tokens": 30504424.0,
      "step": 12030
    },
    {
      "entropy": 6.228830480575562,
      "epoch": 1.3889209463358339,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004818227868499742,
      "loss": 5.837,
      "mean_token_accuracy": 0.18693964928388596,
      "num_tokens": 30517339.0,
      "step": 12035
    },
    {
      "entropy": 6.190657043457032,
      "epoch": 1.3894979803808425,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004818065350341412,
      "loss": 5.8226,
      "mean_token_accuracy": 0.18817414343357086,
      "num_tokens": 30529218.0,
      "step": 12040
    },
    {
      "entropy": 6.23997745513916,
      "epoch": 1.3900750144258511,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048179027626247325,
      "loss": 5.7923,
      "mean_token_accuracy": 0.1895095095038414,
      "num_tokens": 30540601.0,
      "step": 12045
    },
    {
      "entropy": 6.087723684310913,
      "epoch": 1.3906520484708598,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004817740105355169,
      "loss": 5.7443,
      "mean_token_accuracy": 0.194292052090168,
      "num_tokens": 30552834.0,
      "step": 12050
    },
    {
      "entropy": 6.1866191864013675,
      "epoch": 1.3912290825158684,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048175773785381947,
      "loss": 5.87,
      "mean_token_accuracy": 0.18968655467033385,
      "num_tokens": 30563799.0,
      "step": 12055
    },
    {
      "entropy": 6.223745107650757,
      "epoch": 1.391806116560877,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048174145821792833,
      "loss": 5.814,
      "mean_token_accuracy": 0.19449094235897063,
      "num_tokens": 30576637.0,
      "step": 12060
    },
    {
      "entropy": 6.216922807693481,
      "epoch": 1.3923831506058857,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004817251716283908,
      "loss": 5.9382,
      "mean_token_accuracy": 0.1813180476427078,
      "num_tokens": 30589720.0,
      "step": 12065
    },
    {
      "entropy": 6.258935022354126,
      "epoch": 1.3929601846508943,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004817088780857548,
      "loss": 5.9224,
      "mean_token_accuracy": 0.18712227046489716,
      "num_tokens": 30602895.0,
      "step": 12070
    },
    {
      "entropy": 6.169496631622314,
      "epoch": 1.393537218695903,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048169257759056845,
      "loss": 5.8149,
      "mean_token_accuracy": 0.1958261877298355,
      "num_tokens": 30615411.0,
      "step": 12075
    },
    {
      "entropy": 6.236493921279907,
      "epoch": 1.3941142527409118,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048167627014337994,
      "loss": 5.8406,
      "mean_token_accuracy": 0.18758523911237718,
      "num_tokens": 30627682.0,
      "step": 12080
    },
    {
      "entropy": 6.276213598251343,
      "epoch": 1.3946912867859202,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00048165995574473764,
      "loss": 5.9761,
      "mean_token_accuracy": 0.17848981022834778,
      "num_tokens": 30641723.0,
      "step": 12085
    },
    {
      "entropy": 6.289062118530273,
      "epoch": 1.395268320830929,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048164363439519043,
      "loss": 5.9096,
      "mean_token_accuracy": 0.18233100771903993,
      "num_tokens": 30654926.0,
      "step": 12090
    },
    {
      "entropy": 6.261133480072021,
      "epoch": 1.3958453548759377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004816273060952873,
      "loss": 5.8746,
      "mean_token_accuracy": 0.1852752983570099,
      "num_tokens": 30667641.0,
      "step": 12095
    },
    {
      "entropy": 6.2364496231079105,
      "epoch": 1.3964223889209464,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048161097084557726,
      "loss": 5.8724,
      "mean_token_accuracy": 0.18757182359695435,
      "num_tokens": 30680841.0,
      "step": 12100
    },
    {
      "entropy": 6.260488986968994,
      "epoch": 1.396999422965955,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048159462864660993,
      "loss": 5.8099,
      "mean_token_accuracy": 0.19341374784708024,
      "num_tokens": 30695829.0,
      "step": 12105
    },
    {
      "entropy": 6.2558189868927006,
      "epoch": 1.3975764570109637,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004815782794989348,
      "loss": 5.7993,
      "mean_token_accuracy": 0.18306184262037278,
      "num_tokens": 30707853.0,
      "step": 12110
    },
    {
      "entropy": 6.154651355743408,
      "epoch": 1.3981534910559723,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004815619234031019,
      "loss": 5.7858,
      "mean_token_accuracy": 0.18576117902994155,
      "num_tokens": 30720568.0,
      "step": 12115
    },
    {
      "entropy": 6.116623640060425,
      "epoch": 1.398730525100981,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004815455603596613,
      "loss": 5.822,
      "mean_token_accuracy": 0.19020836502313615,
      "num_tokens": 30734282.0,
      "step": 12120
    },
    {
      "entropy": 6.257430219650269,
      "epoch": 1.3993075591459896,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004815291903691634,
      "loss": 5.8166,
      "mean_token_accuracy": 0.1919144421815872,
      "num_tokens": 30746186.0,
      "step": 12125
    },
    {
      "entropy": 6.205156087875366,
      "epoch": 1.3998845931909982,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048151281343215873,
      "loss": 5.865,
      "mean_token_accuracy": 0.18733261972665788,
      "num_tokens": 30757697.0,
      "step": 12130
    },
    {
      "entropy": 6.241019439697266,
      "epoch": 1.4004616272360069,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004814964295491982,
      "loss": 5.8178,
      "mean_token_accuracy": 0.19326186925172806,
      "num_tokens": 30770068.0,
      "step": 12135
    },
    {
      "entropy": 6.223239803314209,
      "epoch": 1.4010386612810155,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048148003872083286,
      "loss": 5.789,
      "mean_token_accuracy": 0.19010363519191742,
      "num_tokens": 30782135.0,
      "step": 12140
    },
    {
      "entropy": 6.2338135719299315,
      "epoch": 1.4016156953260244,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048146364094761384,
      "loss": 5.8936,
      "mean_token_accuracy": 0.18535079509019853,
      "num_tokens": 30793448.0,
      "step": 12145
    },
    {
      "entropy": 6.2299798965454105,
      "epoch": 1.4021927293710328,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048144723623009297,
      "loss": 5.8069,
      "mean_token_accuracy": 0.18635861873626708,
      "num_tokens": 30806280.0,
      "step": 12150
    },
    {
      "entropy": 6.202839279174805,
      "epoch": 1.4027697634160416,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004814308245688218,
      "loss": 5.814,
      "mean_token_accuracy": 0.18964794278144836,
      "num_tokens": 30820394.0,
      "step": 12155
    },
    {
      "entropy": 6.2753565311431885,
      "epoch": 1.40334679746105,
      "grad_norm": 0.875,
      "learning_rate": 0.00048141440596435235,
      "loss": 5.8476,
      "mean_token_accuracy": 0.18839685916900634,
      "num_tokens": 30832343.0,
      "step": 12160
    },
    {
      "entropy": 6.226115322113037,
      "epoch": 1.403923831506059,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004813979804172369,
      "loss": 5.8705,
      "mean_token_accuracy": 0.18870767652988435,
      "num_tokens": 30844833.0,
      "step": 12165
    },
    {
      "entropy": 6.1677062034606935,
      "epoch": 1.4045008655510676,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048138154792802795,
      "loss": 5.7662,
      "mean_token_accuracy": 0.19425424933433533,
      "num_tokens": 30856899.0,
      "step": 12170
    },
    {
      "entropy": 6.232876825332641,
      "epoch": 1.4050778995960762,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004813651084972781,
      "loss": 5.8463,
      "mean_token_accuracy": 0.18824744820594788,
      "num_tokens": 30869172.0,
      "step": 12175
    },
    {
      "entropy": 6.265374660491943,
      "epoch": 1.4056549336410848,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048134866212554036,
      "loss": 5.8308,
      "mean_token_accuracy": 0.1878613129258156,
      "num_tokens": 30880751.0,
      "step": 12180
    },
    {
      "entropy": 6.224615955352784,
      "epoch": 1.4062319676860935,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004813322088133679,
      "loss": 5.8103,
      "mean_token_accuracy": 0.1838347613811493,
      "num_tokens": 30893583.0,
      "step": 12185
    },
    {
      "entropy": 6.26621150970459,
      "epoch": 1.4068090017311021,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048131574856131407,
      "loss": 5.9156,
      "mean_token_accuracy": 0.18395811766386033,
      "num_tokens": 30906150.0,
      "step": 12190
    },
    {
      "entropy": 6.233859920501709,
      "epoch": 1.4073860357761108,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004812992813699324,
      "loss": 5.8568,
      "mean_token_accuracy": 0.18411531895399094,
      "num_tokens": 30918505.0,
      "step": 12195
    },
    {
      "entropy": 6.180645418167114,
      "epoch": 1.4079630698211194,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000481282807239777,
      "loss": 5.8581,
      "mean_token_accuracy": 0.1881232365965843,
      "num_tokens": 30932519.0,
      "step": 12200
    },
    {
      "entropy": 6.212860107421875,
      "epoch": 1.408540103866128,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004812663261714019,
      "loss": 5.8929,
      "mean_token_accuracy": 0.18428244590759277,
      "num_tokens": 30944482.0,
      "step": 12205
    },
    {
      "entropy": 6.225508975982666,
      "epoch": 1.4091171379111367,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004812498381653613,
      "loss": 5.7873,
      "mean_token_accuracy": 0.19120807498693465,
      "num_tokens": 30956726.0,
      "step": 12210
    },
    {
      "entropy": 6.200875377655029,
      "epoch": 1.4096941719561453,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004812333432222098,
      "loss": 5.8096,
      "mean_token_accuracy": 0.18800670504570008,
      "num_tokens": 30968557.0,
      "step": 12215
    },
    {
      "entropy": 6.246923923492432,
      "epoch": 1.4102712060011542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004812168413425023,
      "loss": 5.8499,
      "mean_token_accuracy": 0.19499286115169526,
      "num_tokens": 30981442.0,
      "step": 12220
    },
    {
      "entropy": 6.194499492645264,
      "epoch": 1.4108482400461626,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048120033252679374,
      "loss": 5.7828,
      "mean_token_accuracy": 0.18518147319555284,
      "num_tokens": 30994126.0,
      "step": 12225
    },
    {
      "entropy": 6.262293672561645,
      "epoch": 1.4114252740911715,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048118381677563946,
      "loss": 5.8434,
      "mean_token_accuracy": 0.19208305925130845,
      "num_tokens": 31006688.0,
      "step": 12230
    },
    {
      "entropy": 6.267688131332397,
      "epoch": 1.41200230813618,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004811672940895949,
      "loss": 5.9424,
      "mean_token_accuracy": 0.1738465279340744,
      "num_tokens": 31019922.0,
      "step": 12235
    },
    {
      "entropy": 6.255361557006836,
      "epoch": 1.4125793421811887,
      "grad_norm": 0.875,
      "learning_rate": 0.0004811507644692158,
      "loss": 5.9409,
      "mean_token_accuracy": 0.18228746950626373,
      "num_tokens": 31032068.0,
      "step": 12240
    },
    {
      "entropy": 6.203035163879394,
      "epoch": 1.4131563762261974,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004811342279150581,
      "loss": 5.7743,
      "mean_token_accuracy": 0.19129448682069777,
      "num_tokens": 31045547.0,
      "step": 12245
    },
    {
      "entropy": 6.208262205123901,
      "epoch": 1.413733410271206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004811176844276781,
      "loss": 5.8593,
      "mean_token_accuracy": 0.1870012179017067,
      "num_tokens": 31059219.0,
      "step": 12250
    },
    {
      "entropy": 6.277154541015625,
      "epoch": 1.4143104443162147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004811011340076322,
      "loss": 5.8936,
      "mean_token_accuracy": 0.18521436005830766,
      "num_tokens": 31072472.0,
      "step": 12255
    },
    {
      "entropy": 6.284340143203735,
      "epoch": 1.4148874783612233,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048108457665547695,
      "loss": 5.9143,
      "mean_token_accuracy": 0.18017235845327378,
      "num_tokens": 31085394.0,
      "step": 12260
    },
    {
      "entropy": 6.3134997367858885,
      "epoch": 1.415464512406232,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004810680123717694,
      "loss": 5.9067,
      "mean_token_accuracy": 0.18589280545711517,
      "num_tokens": 31098651.0,
      "step": 12265
    },
    {
      "entropy": 6.234366035461425,
      "epoch": 1.4160415464512406,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004810514411570666,
      "loss": 5.8641,
      "mean_token_accuracy": 0.1886117711663246,
      "num_tokens": 31111297.0,
      "step": 12270
    },
    {
      "entropy": 6.187476396560669,
      "epoch": 1.4166185804962492,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004810348630119259,
      "loss": 5.857,
      "mean_token_accuracy": 0.18639783710241317,
      "num_tokens": 31124247.0,
      "step": 12275
    },
    {
      "entropy": 6.227819156646729,
      "epoch": 1.4171956145412579,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048101827793690493,
      "loss": 5.8226,
      "mean_token_accuracy": 0.1915529578924179,
      "num_tokens": 31136106.0,
      "step": 12280
    },
    {
      "entropy": 6.147597074508667,
      "epoch": 1.4177726485862667,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004810016859325615,
      "loss": 5.7261,
      "mean_token_accuracy": 0.19978685677051544,
      "num_tokens": 31148561.0,
      "step": 12285
    },
    {
      "entropy": 6.303668689727783,
      "epoch": 1.4183496826312751,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004809850869994537,
      "loss": 5.854,
      "mean_token_accuracy": 0.18568157851696016,
      "num_tokens": 31162289.0,
      "step": 12290
    },
    {
      "entropy": 6.279684114456177,
      "epoch": 1.418926716676284,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004809684811381398,
      "loss": 5.8186,
      "mean_token_accuracy": 0.18668742179870607,
      "num_tokens": 31174013.0,
      "step": 12295
    },
    {
      "entropy": 6.121853256225586,
      "epoch": 1.4195037507212924,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004809518683491785,
      "loss": 5.7486,
      "mean_token_accuracy": 0.19865711331367492,
      "num_tokens": 31186299.0,
      "step": 12300
    },
    {
      "entropy": 6.2222252368927,
      "epoch": 1.4200807847663013,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048093524863312823,
      "loss": 5.8619,
      "mean_token_accuracy": 0.1875106304883957,
      "num_tokens": 31198458.0,
      "step": 12305
    },
    {
      "entropy": 6.255124759674072,
      "epoch": 1.42065781881131,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004809186219905482,
      "loss": 5.8924,
      "mean_token_accuracy": 0.1819571554660797,
      "num_tokens": 31210712.0,
      "step": 12310
    },
    {
      "entropy": 6.230963945388794,
      "epoch": 1.4212348528563186,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004809019884219976,
      "loss": 5.8255,
      "mean_token_accuracy": 0.1896081894636154,
      "num_tokens": 31223460.0,
      "step": 12315
    },
    {
      "entropy": 6.1729991912841795,
      "epoch": 1.4218118869013272,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048088534792803595,
      "loss": 5.8498,
      "mean_token_accuracy": 0.18703863620758057,
      "num_tokens": 31235298.0,
      "step": 12320
    },
    {
      "entropy": 6.265325021743775,
      "epoch": 1.4223889209463358,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048086870050922286,
      "loss": 5.8776,
      "mean_token_accuracy": 0.18006298840045928,
      "num_tokens": 31247164.0,
      "step": 12325
    },
    {
      "entropy": 6.278746604919434,
      "epoch": 1.4229659549913445,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048085204616611833,
      "loss": 5.8669,
      "mean_token_accuracy": 0.18791476041078567,
      "num_tokens": 31259963.0,
      "step": 12330
    },
    {
      "entropy": 6.192272567749024,
      "epoch": 1.4235429890363531,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048083538489928246,
      "loss": 5.8664,
      "mean_token_accuracy": 0.18336281180381775,
      "num_tokens": 31272646.0,
      "step": 12335
    },
    {
      "entropy": 6.1753192901611325,
      "epoch": 1.4241200230813618,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004808187167092757,
      "loss": 5.7767,
      "mean_token_accuracy": 0.19271085113286973,
      "num_tokens": 31286474.0,
      "step": 12340
    },
    {
      "entropy": 6.158377742767334,
      "epoch": 1.4246970571263704,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004808020415966586,
      "loss": 5.7278,
      "mean_token_accuracy": 0.19635030031204223,
      "num_tokens": 31300086.0,
      "step": 12345
    },
    {
      "entropy": 6.223275995254516,
      "epoch": 1.425274091171379,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000480785359561992,
      "loss": 5.7374,
      "mean_token_accuracy": 0.19766900092363357,
      "num_tokens": 31312562.0,
      "step": 12350
    },
    {
      "entropy": 6.241062307357788,
      "epoch": 1.4258511252163877,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048076867060583704,
      "loss": 5.8846,
      "mean_token_accuracy": 0.18585018813610077,
      "num_tokens": 31325200.0,
      "step": 12355
    },
    {
      "entropy": 6.212400579452515,
      "epoch": 1.4264281592613965,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004807519747287551,
      "loss": 5.822,
      "mean_token_accuracy": 0.19216081649065017,
      "num_tokens": 31337361.0,
      "step": 12360
    },
    {
      "entropy": 6.221188545227051,
      "epoch": 1.427005193306405,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004807352719313078,
      "loss": 5.7934,
      "mean_token_accuracy": 0.19212548136711122,
      "num_tokens": 31351411.0,
      "step": 12365
    },
    {
      "entropy": 6.1886148929595945,
      "epoch": 1.4275822273514138,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004807185622140567,
      "loss": 5.7916,
      "mean_token_accuracy": 0.18944623470306396,
      "num_tokens": 31364616.0,
      "step": 12370
    },
    {
      "entropy": 6.171735334396362,
      "epoch": 1.4281592613964225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048070184557756396,
      "loss": 5.7954,
      "mean_token_accuracy": 0.19159631431102753,
      "num_tokens": 31376515.0,
      "step": 12375
    },
    {
      "entropy": 6.272715759277344,
      "epoch": 1.428736295441431,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048068512202239177,
      "loss": 5.8501,
      "mean_token_accuracy": 0.1893751472234726,
      "num_tokens": 31388408.0,
      "step": 12380
    },
    {
      "entropy": 6.193832731246948,
      "epoch": 1.4293133294864397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004806683915491028,
      "loss": 5.8202,
      "mean_token_accuracy": 0.18815591633319856,
      "num_tokens": 31401001.0,
      "step": 12385
    },
    {
      "entropy": 6.212103605270386,
      "epoch": 1.4298903635314484,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004806516541582596,
      "loss": 5.7747,
      "mean_token_accuracy": 0.1877763420343399,
      "num_tokens": 31413259.0,
      "step": 12390
    },
    {
      "entropy": 6.308931684494018,
      "epoch": 1.430467397576457,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048063490985042506,
      "loss": 5.9743,
      "mean_token_accuracy": 0.17446880042552948,
      "num_tokens": 31425997.0,
      "step": 12395
    },
    {
      "entropy": 6.220090436935425,
      "epoch": 1.4310444316214657,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004806181586261626,
      "loss": 5.8762,
      "mean_token_accuracy": 0.18121187388896942,
      "num_tokens": 31439967.0,
      "step": 12400
    },
    {
      "entropy": 6.257448291778564,
      "epoch": 1.4316214656664743,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048060140048603544,
      "loss": 5.8227,
      "mean_token_accuracy": 0.1857739120721817,
      "num_tokens": 31453122.0,
      "step": 12405
    },
    {
      "entropy": 6.278946018218994,
      "epoch": 1.432198499711483,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004805846354306073,
      "loss": 5.8743,
      "mean_token_accuracy": 0.18489379435777664,
      "num_tokens": 31465043.0,
      "step": 12410
    },
    {
      "entropy": 6.178117418289185,
      "epoch": 1.4327755337564916,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048056786346044214,
      "loss": 5.7707,
      "mean_token_accuracy": 0.19426106810569763,
      "num_tokens": 31477003.0,
      "step": 12415
    },
    {
      "entropy": 6.175166749954224,
      "epoch": 1.4333525678015002,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048055108457610395,
      "loss": 5.8609,
      "mean_token_accuracy": 0.18873385190963746,
      "num_tokens": 31491612.0,
      "step": 12420
    },
    {
      "entropy": 6.229504108428955,
      "epoch": 1.4339296018465089,
      "grad_norm": 0.875,
      "learning_rate": 0.0004805342987781571,
      "loss": 5.7568,
      "mean_token_accuracy": 0.1903178572654724,
      "num_tokens": 31504488.0,
      "step": 12425
    },
    {
      "entropy": 6.232200813293457,
      "epoch": 1.4345066358915175,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004805175060671663,
      "loss": 5.8604,
      "mean_token_accuracy": 0.1884032294154167,
      "num_tokens": 31517770.0,
      "step": 12430
    },
    {
      "entropy": 6.134601402282715,
      "epoch": 1.4350836699365264,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004805007064436962,
      "loss": 5.6456,
      "mean_token_accuracy": 0.19610061049461364,
      "num_tokens": 31530262.0,
      "step": 12435
    },
    {
      "entropy": 6.187317371368408,
      "epoch": 1.4356607039815348,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004804838999083119,
      "loss": 5.8257,
      "mean_token_accuracy": 0.17793625593185425,
      "num_tokens": 31543988.0,
      "step": 12440
    },
    {
      "entropy": 6.180283069610596,
      "epoch": 1.4362377380265436,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004804670864615787,
      "loss": 5.7861,
      "mean_token_accuracy": 0.18876124918460846,
      "num_tokens": 31555690.0,
      "step": 12445
    },
    {
      "entropy": 6.200000762939453,
      "epoch": 1.4368147720715523,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048045026610406223,
      "loss": 5.8376,
      "mean_token_accuracy": 0.1839052200317383,
      "num_tokens": 31568542.0,
      "step": 12450
    },
    {
      "entropy": 6.206508731842041,
      "epoch": 1.437391806116561,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000480433438836328,
      "loss": 5.884,
      "mean_token_accuracy": 0.18490970134735107,
      "num_tokens": 31580538.0,
      "step": 12455
    },
    {
      "entropy": 6.2382430076599125,
      "epoch": 1.4379688401615696,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048041660465894206,
      "loss": 5.757,
      "mean_token_accuracy": 0.19315117597579956,
      "num_tokens": 31593414.0,
      "step": 12460
    },
    {
      "entropy": 6.181693744659424,
      "epoch": 1.4385458742065782,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004803997635724707,
      "loss": 5.7937,
      "mean_token_accuracy": 0.19153431951999664,
      "num_tokens": 31605931.0,
      "step": 12465
    },
    {
      "entropy": 6.252446460723877,
      "epoch": 1.4391229082515868,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004803829155774804,
      "loss": 5.8006,
      "mean_token_accuracy": 0.18253548294305802,
      "num_tokens": 31617949.0,
      "step": 12470
    },
    {
      "entropy": 6.241463136672974,
      "epoch": 1.4396999422965955,
      "grad_norm": 0.875,
      "learning_rate": 0.0004803660606745377,
      "loss": 5.852,
      "mean_token_accuracy": 0.18622518032789231,
      "num_tokens": 31630429.0,
      "step": 12475
    },
    {
      "entropy": 6.2249165058135985,
      "epoch": 1.4402769763416041,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048034919886420953,
      "loss": 5.8341,
      "mean_token_accuracy": 0.1859687700867653,
      "num_tokens": 31641430.0,
      "step": 12480
    },
    {
      "entropy": 6.296145296096801,
      "epoch": 1.4408540103866128,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048033233014706304,
      "loss": 5.9435,
      "mean_token_accuracy": 0.1808713048696518,
      "num_tokens": 31653465.0,
      "step": 12485
    },
    {
      "entropy": 6.2770781993865965,
      "epoch": 1.4414310444316214,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048031545452366565,
      "loss": 5.8362,
      "mean_token_accuracy": 0.18749186843633653,
      "num_tokens": 31665550.0,
      "step": 12490
    },
    {
      "entropy": 6.137077569961548,
      "epoch": 1.44200807847663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048029857199458493,
      "loss": 5.809,
      "mean_token_accuracy": 0.19560253471136094,
      "num_tokens": 31677991.0,
      "step": 12495
    },
    {
      "entropy": 6.284508991241455,
      "epoch": 1.442585112521639,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048028168256038873,
      "loss": 5.8613,
      "mean_token_accuracy": 0.18919273018836974,
      "num_tokens": 31690148.0,
      "step": 12500
    },
    {
      "entropy": 6.291284656524658,
      "epoch": 1.4431621465666473,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048026478622164513,
      "loss": 5.879,
      "mean_token_accuracy": 0.18976499438285827,
      "num_tokens": 31703632.0,
      "step": 12505
    },
    {
      "entropy": 6.220984172821045,
      "epoch": 1.4437391806116562,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048024788297892234,
      "loss": 5.8614,
      "mean_token_accuracy": 0.18823379874229432,
      "num_tokens": 31714722.0,
      "step": 12510
    },
    {
      "entropy": 6.261355590820313,
      "epoch": 1.4443162146566648,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000480230972832789,
      "loss": 5.7962,
      "mean_token_accuracy": 0.19225970655679703,
      "num_tokens": 31726554.0,
      "step": 12515
    },
    {
      "entropy": 6.314726829528809,
      "epoch": 1.4448932487016735,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048021405578381384,
      "loss": 5.9398,
      "mean_token_accuracy": 0.18269460648298264,
      "num_tokens": 31738604.0,
      "step": 12520
    },
    {
      "entropy": 6.14175329208374,
      "epoch": 1.445470282746682,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004801971318325659,
      "loss": 5.7879,
      "mean_token_accuracy": 0.19555617719888688,
      "num_tokens": 31750084.0,
      "step": 12525
    },
    {
      "entropy": 6.198133325576782,
      "epoch": 1.4460473167916907,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004801802009796142,
      "loss": 5.8344,
      "mean_token_accuracy": 0.1847309097647667,
      "num_tokens": 31764068.0,
      "step": 12530
    },
    {
      "entropy": 6.2189037799835205,
      "epoch": 1.4466243508366994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004801632632255285,
      "loss": 5.7757,
      "mean_token_accuracy": 0.18733726143836976,
      "num_tokens": 31777425.0,
      "step": 12535
    },
    {
      "entropy": 6.205999326705933,
      "epoch": 1.447201384881708,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004801463185708783,
      "loss": 5.8155,
      "mean_token_accuracy": 0.19239280074834825,
      "num_tokens": 31789767.0,
      "step": 12540
    },
    {
      "entropy": 6.176294898986816,
      "epoch": 1.4477784189267167,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048012936701623363,
      "loss": 5.7835,
      "mean_token_accuracy": 0.19475572258234025,
      "num_tokens": 31801399.0,
      "step": 12545
    },
    {
      "entropy": 6.252738046646118,
      "epoch": 1.4483554529717253,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048011240856216456,
      "loss": 5.83,
      "mean_token_accuracy": 0.19104738682508468,
      "num_tokens": 31814260.0,
      "step": 12550
    },
    {
      "entropy": 6.287702369689941,
      "epoch": 1.448932487016734,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048009544320924154,
      "loss": 5.8275,
      "mean_token_accuracy": 0.19155623614788056,
      "num_tokens": 31826674.0,
      "step": 12555
    },
    {
      "entropy": 6.263993978500366,
      "epoch": 1.4495095210617426,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004800784709580351,
      "loss": 5.9525,
      "mean_token_accuracy": 0.1792732447385788,
      "num_tokens": 31839951.0,
      "step": 12560
    },
    {
      "entropy": 6.229682683944702,
      "epoch": 1.4500865551067512,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004800614918091161,
      "loss": 5.7859,
      "mean_token_accuracy": 0.19295246303081512,
      "num_tokens": 31852344.0,
      "step": 12565
    },
    {
      "entropy": 6.186488246917724,
      "epoch": 1.4506635891517599,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004800445057630558,
      "loss": 5.7094,
      "mean_token_accuracy": 0.1992287218570709,
      "num_tokens": 31865134.0,
      "step": 12570
    },
    {
      "entropy": 6.178214693069458,
      "epoch": 1.4512406231967687,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004800275128204254,
      "loss": 5.7837,
      "mean_token_accuracy": 0.19416512548923492,
      "num_tokens": 31877395.0,
      "step": 12575
    },
    {
      "entropy": 6.240856552124024,
      "epoch": 1.4518176572417771,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048001051298179637,
      "loss": 5.8125,
      "mean_token_accuracy": 0.18945945650339127,
      "num_tokens": 31889421.0,
      "step": 12580
    },
    {
      "entropy": 6.237762498855591,
      "epoch": 1.452394691286786,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004799935062477407,
      "loss": 5.8043,
      "mean_token_accuracy": 0.1950754702091217,
      "num_tokens": 31903004.0,
      "step": 12585
    },
    {
      "entropy": 6.267592191696167,
      "epoch": 1.4529717253317946,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047997649261883013,
      "loss": 5.8477,
      "mean_token_accuracy": 0.19296132475137712,
      "num_tokens": 31914682.0,
      "step": 12590
    },
    {
      "entropy": 6.271247005462646,
      "epoch": 1.4535487593768033,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004799594720956371,
      "loss": 5.8478,
      "mean_token_accuracy": 0.18563375622034073,
      "num_tokens": 31928861.0,
      "step": 12595
    },
    {
      "entropy": 6.192510175704956,
      "epoch": 1.454125793421812,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047994244467873407,
      "loss": 5.8183,
      "mean_token_accuracy": 0.18786570876836778,
      "num_tokens": 31942043.0,
      "step": 12600
    },
    {
      "entropy": 6.281221723556518,
      "epoch": 1.4547028274668206,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047992541036869364,
      "loss": 5.9085,
      "mean_token_accuracy": 0.18552975058555604,
      "num_tokens": 31954453.0,
      "step": 12605
    },
    {
      "entropy": 6.219685840606689,
      "epoch": 1.4552798615118292,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004799083691660889,
      "loss": 5.7916,
      "mean_token_accuracy": 0.19105765372514724,
      "num_tokens": 31967998.0,
      "step": 12610
    },
    {
      "entropy": 6.242797183990478,
      "epoch": 1.4558568955568378,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004798913210714929,
      "loss": 5.8967,
      "mean_token_accuracy": 0.18078264892101287,
      "num_tokens": 31980445.0,
      "step": 12615
    },
    {
      "entropy": 6.158501052856446,
      "epoch": 1.4564339296018465,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047987426608547915,
      "loss": 5.8063,
      "mean_token_accuracy": 0.19550029784440995,
      "num_tokens": 31993706.0,
      "step": 12620
    },
    {
      "entropy": 6.231975746154785,
      "epoch": 1.4570109636468551,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004798572042086212,
      "loss": 5.8039,
      "mean_token_accuracy": 0.1910783588886261,
      "num_tokens": 32005189.0,
      "step": 12625
    },
    {
      "entropy": 6.20102891921997,
      "epoch": 1.4575879976918638,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047984013544149286,
      "loss": 5.8344,
      "mean_token_accuracy": 0.19009873121976853,
      "num_tokens": 32017883.0,
      "step": 12630
    },
    {
      "entropy": 6.1803216457366945,
      "epoch": 1.4581650317368724,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047982305978466836,
      "loss": 5.7878,
      "mean_token_accuracy": 0.19370767027139663,
      "num_tokens": 32030701.0,
      "step": 12635
    },
    {
      "entropy": 6.213293790817261,
      "epoch": 1.4587420657818813,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047980597723872205,
      "loss": 5.8544,
      "mean_token_accuracy": 0.18105095773935317,
      "num_tokens": 32042726.0,
      "step": 12640
    },
    {
      "entropy": 6.271682071685791,
      "epoch": 1.4593190998268897,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004797888878042283,
      "loss": 5.7766,
      "mean_token_accuracy": 0.1917669579386711,
      "num_tokens": 32055158.0,
      "step": 12645
    },
    {
      "entropy": 6.206538152694702,
      "epoch": 1.4598961338718985,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047977179148176217,
      "loss": 5.8247,
      "mean_token_accuracy": 0.1868303820490837,
      "num_tokens": 32068416.0,
      "step": 12650
    },
    {
      "entropy": 6.254722785949707,
      "epoch": 1.4604731679169072,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004797546882718985,
      "loss": 5.8609,
      "mean_token_accuracy": 0.18896269649267197,
      "num_tokens": 32081800.0,
      "step": 12655
    },
    {
      "entropy": 6.2897961139678955,
      "epoch": 1.4610502019619158,
      "grad_norm": 0.8125,
      "learning_rate": 0.00047973757817521256,
      "loss": 5.8956,
      "mean_token_accuracy": 0.18403008133172988,
      "num_tokens": 32095134.0,
      "step": 12660
    },
    {
      "entropy": 6.153244543075561,
      "epoch": 1.4616272360069245,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004797204611922799,
      "loss": 5.8143,
      "mean_token_accuracy": 0.18783867359161377,
      "num_tokens": 32106685.0,
      "step": 12665
    },
    {
      "entropy": 6.213627099990845,
      "epoch": 1.462204270051933,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047970333732367626,
      "loss": 5.8572,
      "mean_token_accuracy": 0.18631642758846284,
      "num_tokens": 32120072.0,
      "step": 12670
    },
    {
      "entropy": 6.256570625305176,
      "epoch": 1.4627813040969417,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047968620656997754,
      "loss": 5.8174,
      "mean_token_accuracy": 0.18835413455963135,
      "num_tokens": 32133205.0,
      "step": 12675
    },
    {
      "entropy": 6.191640043258667,
      "epoch": 1.4633583381419504,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047966906893175994,
      "loss": 5.7534,
      "mean_token_accuracy": 0.20220461189746858,
      "num_tokens": 32147674.0,
      "step": 12680
    },
    {
      "entropy": 6.226243877410889,
      "epoch": 1.463935372186959,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004796519244095998,
      "loss": 5.8194,
      "mean_token_accuracy": 0.19077904373407364,
      "num_tokens": 32159689.0,
      "step": 12685
    },
    {
      "entropy": 6.2036652088165285,
      "epoch": 1.4645124062319677,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047963477300407394,
      "loss": 5.8644,
      "mean_token_accuracy": 0.1852376013994217,
      "num_tokens": 32172641.0,
      "step": 12690
    },
    {
      "entropy": 6.183947038650513,
      "epoch": 1.4650894402769763,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047961761471575903,
      "loss": 5.8521,
      "mean_token_accuracy": 0.18976031988859177,
      "num_tokens": 32184512.0,
      "step": 12695
    },
    {
      "entropy": 6.211766004562378,
      "epoch": 1.465666474321985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047960044954523237,
      "loss": 5.8431,
      "mean_token_accuracy": 0.19121228754520417,
      "num_tokens": 32195908.0,
      "step": 12700
    },
    {
      "entropy": 6.183211755752564,
      "epoch": 1.4662435083669936,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047958327749307117,
      "loss": 5.7938,
      "mean_token_accuracy": 0.1863427296280861,
      "num_tokens": 32208855.0,
      "step": 12705
    },
    {
      "entropy": 6.158882427215576,
      "epoch": 1.4668205424120022,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047956609855985305,
      "loss": 5.7893,
      "mean_token_accuracy": 0.19222267717123032,
      "num_tokens": 32222769.0,
      "step": 12710
    },
    {
      "entropy": 6.317701530456543,
      "epoch": 1.467397576457011,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004795489127461559,
      "loss": 5.864,
      "mean_token_accuracy": 0.18590313494205474,
      "num_tokens": 32235192.0,
      "step": 12715
    },
    {
      "entropy": 6.1869384288787845,
      "epoch": 1.4679746105020195,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047953172005255756,
      "loss": 5.8366,
      "mean_token_accuracy": 0.19284756034612655,
      "num_tokens": 32249196.0,
      "step": 12720
    },
    {
      "entropy": 6.26898603439331,
      "epoch": 1.4685516445470284,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004795145204796365,
      "loss": 5.8915,
      "mean_token_accuracy": 0.18242456912994384,
      "num_tokens": 32261591.0,
      "step": 12725
    },
    {
      "entropy": 6.255330562591553,
      "epoch": 1.469128678592037,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004794973140279711,
      "loss": 5.8981,
      "mean_token_accuracy": 0.1844722867012024,
      "num_tokens": 32275409.0,
      "step": 12730
    },
    {
      "entropy": 6.261513996124267,
      "epoch": 1.4697057126370456,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047948010069814005,
      "loss": 5.7273,
      "mean_token_accuracy": 0.19453997313976287,
      "num_tokens": 32287897.0,
      "step": 12735
    },
    {
      "entropy": 6.236474657058716,
      "epoch": 1.4702827466820543,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004794628804907225,
      "loss": 5.9273,
      "mean_token_accuracy": 0.18126835078001022,
      "num_tokens": 32300595.0,
      "step": 12740
    },
    {
      "entropy": 6.231548547744751,
      "epoch": 1.470859780727063,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047944565340629755,
      "loss": 5.8382,
      "mean_token_accuracy": 0.18922677040100097,
      "num_tokens": 32312374.0,
      "step": 12745
    },
    {
      "entropy": 6.229601764678955,
      "epoch": 1.4714368147720716,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047942841944544453,
      "loss": 5.7582,
      "mean_token_accuracy": 0.1980397716164589,
      "num_tokens": 32326526.0,
      "step": 12750
    },
    {
      "entropy": 6.283019590377807,
      "epoch": 1.4720138488170802,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004794111786087431,
      "loss": 5.8982,
      "mean_token_accuracy": 0.18576376736164094,
      "num_tokens": 32339591.0,
      "step": 12755
    },
    {
      "entropy": 6.25236177444458,
      "epoch": 1.4725908828620888,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004793939308967733,
      "loss": 5.8642,
      "mean_token_accuracy": 0.18350427299737931,
      "num_tokens": 32352802.0,
      "step": 12760
    },
    {
      "entropy": 6.22540864944458,
      "epoch": 1.4731679169070975,
      "grad_norm": 1.0,
      "learning_rate": 0.0004793766763101152,
      "loss": 5.8205,
      "mean_token_accuracy": 0.1837970107793808,
      "num_tokens": 32365951.0,
      "step": 12765
    },
    {
      "entropy": 6.248742294311524,
      "epoch": 1.4737449509521061,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00047935941484934905,
      "loss": 5.8623,
      "mean_token_accuracy": 0.18229759633541107,
      "num_tokens": 32378573.0,
      "step": 12770
    },
    {
      "entropy": 6.204880952835083,
      "epoch": 1.4743219849971148,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00047934214651505547,
      "loss": 5.786,
      "mean_token_accuracy": 0.19564500004053115,
      "num_tokens": 32392135.0,
      "step": 12775
    },
    {
      "entropy": 6.2123212814331055,
      "epoch": 1.4748990190421236,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047932487130781533,
      "loss": 5.8975,
      "mean_token_accuracy": 0.18628474026918412,
      "num_tokens": 32404821.0,
      "step": 12780
    },
    {
      "entropy": 6.182235765457153,
      "epoch": 1.475476053087132,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004793075892282097,
      "loss": 5.7799,
      "mean_token_accuracy": 0.18896986693143844,
      "num_tokens": 32417118.0,
      "step": 12785
    },
    {
      "entropy": 6.239238929748535,
      "epoch": 1.476053087132141,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004792903002768197,
      "loss": 5.7814,
      "mean_token_accuracy": 0.19235165119171144,
      "num_tokens": 32428272.0,
      "step": 12790
    },
    {
      "entropy": 6.19613790512085,
      "epoch": 1.4766301211771495,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047927300445422687,
      "loss": 5.8621,
      "mean_token_accuracy": 0.1876397266983986,
      "num_tokens": 32441038.0,
      "step": 12795
    },
    {
      "entropy": 6.185892534255982,
      "epoch": 1.4772071552221582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004792557017610131,
      "loss": 5.7658,
      "mean_token_accuracy": 0.1827992841601372,
      "num_tokens": 32452959.0,
      "step": 12800
    },
    {
      "entropy": 6.237241458892822,
      "epoch": 1.4777841892671668,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047923839219776027,
      "loss": 5.8685,
      "mean_token_accuracy": 0.1885818913578987,
      "num_tokens": 32465398.0,
      "step": 12805
    },
    {
      "entropy": 6.228723239898682,
      "epoch": 1.4783612233121755,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004792210757650506,
      "loss": 5.7593,
      "mean_token_accuracy": 0.1885174483060837,
      "num_tokens": 32477128.0,
      "step": 12810
    },
    {
      "entropy": 6.212050294876098,
      "epoch": 1.478938257357184,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047920375246346636,
      "loss": 5.8584,
      "mean_token_accuracy": 0.18698422610759735,
      "num_tokens": 32489831.0,
      "step": 12815
    },
    {
      "entropy": 6.261142444610596,
      "epoch": 1.4795152914021927,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047918642229359044,
      "loss": 5.8803,
      "mean_token_accuracy": 0.18028215318918228,
      "num_tokens": 32502239.0,
      "step": 12820
    },
    {
      "entropy": 6.261108922958374,
      "epoch": 1.4800923254472014,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004791690852560056,
      "loss": 5.8281,
      "mean_token_accuracy": 0.1871498078107834,
      "num_tokens": 32515374.0,
      "step": 12825
    },
    {
      "entropy": 6.194988536834717,
      "epoch": 1.48066935949221,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000479151741351295,
      "loss": 5.8494,
      "mean_token_accuracy": 0.18965427428483964,
      "num_tokens": 32527208.0,
      "step": 12830
    },
    {
      "entropy": 6.267339420318604,
      "epoch": 1.4812463935372187,
      "grad_norm": 0.90625,
      "learning_rate": 0.000479134390580042,
      "loss": 5.9118,
      "mean_token_accuracy": 0.18522002995014192,
      "num_tokens": 32540834.0,
      "step": 12835
    },
    {
      "entropy": 6.238912582397461,
      "epoch": 1.4818234275822273,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00047911703294283015,
      "loss": 5.7791,
      "mean_token_accuracy": 0.18987657129764557,
      "num_tokens": 32552833.0,
      "step": 12840
    },
    {
      "entropy": 6.118738174438477,
      "epoch": 1.482400461627236,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047909966844024334,
      "loss": 5.7749,
      "mean_token_accuracy": 0.19416768550872804,
      "num_tokens": 32565484.0,
      "step": 12845
    },
    {
      "entropy": 6.275965976715088,
      "epoch": 1.4829774956722446,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047908229707286547,
      "loss": 5.8788,
      "mean_token_accuracy": 0.18524256944656373,
      "num_tokens": 32579099.0,
      "step": 12850
    },
    {
      "entropy": 6.240973711013794,
      "epoch": 1.4835545297172534,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000479064918841281,
      "loss": 5.7882,
      "mean_token_accuracy": 0.19356197863817215,
      "num_tokens": 32591530.0,
      "step": 12855
    },
    {
      "entropy": 6.2400532245635985,
      "epoch": 1.4841315637622619,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047904753374607427,
      "loss": 5.9264,
      "mean_token_accuracy": 0.1856238454580307,
      "num_tokens": 32603980.0,
      "step": 12860
    },
    {
      "entropy": 6.238358354568481,
      "epoch": 1.4847085978072707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047903014178783015,
      "loss": 5.8094,
      "mean_token_accuracy": 0.18823243230581282,
      "num_tokens": 32617593.0,
      "step": 12865
    },
    {
      "entropy": 6.236946535110474,
      "epoch": 1.4852856318522794,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004790127429671335,
      "loss": 5.8261,
      "mean_token_accuracy": 0.19242363423109055,
      "num_tokens": 32629813.0,
      "step": 12870
    },
    {
      "entropy": 6.249974012374878,
      "epoch": 1.485862665897288,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004789953372845697,
      "loss": 5.8844,
      "mean_token_accuracy": 0.187124340236187,
      "num_tokens": 32640982.0,
      "step": 12875
    },
    {
      "entropy": 6.172384738922119,
      "epoch": 1.4864396999422966,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000478977924740724,
      "loss": 5.7451,
      "mean_token_accuracy": 0.19784511774778366,
      "num_tokens": 32652717.0,
      "step": 12880
    },
    {
      "entropy": 6.22470989227295,
      "epoch": 1.4870167339873053,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004789605053361821,
      "loss": 5.8258,
      "mean_token_accuracy": 0.18996745347976685,
      "num_tokens": 32665531.0,
      "step": 12885
    },
    {
      "entropy": 6.2256263256072994,
      "epoch": 1.487593768032314,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004789430790715299,
      "loss": 5.8119,
      "mean_token_accuracy": 0.1875125303864479,
      "num_tokens": 32679296.0,
      "step": 12890
    },
    {
      "entropy": 6.233219003677368,
      "epoch": 1.4881708020773226,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047892564594735355,
      "loss": 5.8075,
      "mean_token_accuracy": 0.1963459387421608,
      "num_tokens": 32691806.0,
      "step": 12895
    },
    {
      "entropy": 6.129767894744873,
      "epoch": 1.4887478361223312,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047890820596423943,
      "loss": 5.8099,
      "mean_token_accuracy": 0.19216908216476442,
      "num_tokens": 32704280.0,
      "step": 12900
    },
    {
      "entropy": 6.169640016555786,
      "epoch": 1.4893248701673398,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000478890759122774,
      "loss": 5.8079,
      "mean_token_accuracy": 0.18920942395925522,
      "num_tokens": 32716101.0,
      "step": 12905
    },
    {
      "entropy": 6.290735483169556,
      "epoch": 1.4899019042123485,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004788733054235443,
      "loss": 5.8586,
      "mean_token_accuracy": 0.18383937180042267,
      "num_tokens": 32728063.0,
      "step": 12910
    },
    {
      "entropy": 6.211955547332764,
      "epoch": 1.4904789382573571,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047885584486713717,
      "loss": 5.7767,
      "mean_token_accuracy": 0.19220274537801743,
      "num_tokens": 32742084.0,
      "step": 12915
    },
    {
      "entropy": 6.155074882507324,
      "epoch": 1.491055972302366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004788383774541399,
      "loss": 5.8146,
      "mean_token_accuracy": 0.19339046478271485,
      "num_tokens": 32754708.0,
      "step": 12920
    },
    {
      "entropy": 6.219105529785156,
      "epoch": 1.4916330063473744,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004788209031851402,
      "loss": 5.782,
      "mean_token_accuracy": 0.1921439990401268,
      "num_tokens": 32767326.0,
      "step": 12925
    },
    {
      "entropy": 6.27637939453125,
      "epoch": 1.4922100403923833,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004788034220607256,
      "loss": 5.8783,
      "mean_token_accuracy": 0.18821960240602492,
      "num_tokens": 32781334.0,
      "step": 12930
    },
    {
      "entropy": 6.193084621429444,
      "epoch": 1.4927870744373917,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047878593408148405,
      "loss": 5.8933,
      "mean_token_accuracy": 0.196713787317276,
      "num_tokens": 32793795.0,
      "step": 12935
    },
    {
      "entropy": 6.183307027816772,
      "epoch": 1.4933641084824005,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047876843924800393,
      "loss": 5.7468,
      "mean_token_accuracy": 0.19920461028814315,
      "num_tokens": 32807484.0,
      "step": 12940
    },
    {
      "entropy": 6.199143409729004,
      "epoch": 1.4939411425274092,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004787509375608735,
      "loss": 5.8176,
      "mean_token_accuracy": 0.19462572038173676,
      "num_tokens": 32819129.0,
      "step": 12945
    },
    {
      "entropy": 6.191601181030274,
      "epoch": 1.4945181765724178,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047873342902068157,
      "loss": 5.8433,
      "mean_token_accuracy": 0.18896047323942183,
      "num_tokens": 32832077.0,
      "step": 12950
    },
    {
      "entropy": 6.1801121711730955,
      "epoch": 1.4950952106174265,
      "grad_norm": 0.875,
      "learning_rate": 0.00047871591362801694,
      "loss": 5.8339,
      "mean_token_accuracy": 0.19027223736047744,
      "num_tokens": 32845629.0,
      "step": 12955
    },
    {
      "entropy": 6.232339525222779,
      "epoch": 1.495672244662435,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004786983913834687,
      "loss": 5.8007,
      "mean_token_accuracy": 0.19643778055906297,
      "num_tokens": 32857233.0,
      "step": 12960
    },
    {
      "entropy": 6.197593688964844,
      "epoch": 1.4962492787074437,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047868086228762633,
      "loss": 5.8418,
      "mean_token_accuracy": 0.1854689821600914,
      "num_tokens": 32869210.0,
      "step": 12965
    },
    {
      "entropy": 6.292558717727661,
      "epoch": 1.4968263127524524,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047866332634107926,
      "loss": 5.912,
      "mean_token_accuracy": 0.18875516206026077,
      "num_tokens": 32881633.0,
      "step": 12970
    },
    {
      "entropy": 6.291413974761963,
      "epoch": 1.497403346797461,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047864578354441743,
      "loss": 5.77,
      "mean_token_accuracy": 0.19453564286231995,
      "num_tokens": 32893904.0,
      "step": 12975
    },
    {
      "entropy": 6.168428516387939,
      "epoch": 1.4979803808424696,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004786282338982308,
      "loss": 5.8392,
      "mean_token_accuracy": 0.18961111903190614,
      "num_tokens": 32906612.0,
      "step": 12980
    },
    {
      "entropy": 6.2766454219818115,
      "epoch": 1.4985574148874783,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004786106774031096,
      "loss": 5.8731,
      "mean_token_accuracy": 0.19021489918231965,
      "num_tokens": 32919677.0,
      "step": 12985
    },
    {
      "entropy": 6.18550295829773,
      "epoch": 1.499134448932487,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004785931140596445,
      "loss": 5.7661,
      "mean_token_accuracy": 0.19607024639844894,
      "num_tokens": 32932385.0,
      "step": 12990
    },
    {
      "entropy": 6.178236627578736,
      "epoch": 1.4997114829774958,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047857554386842606,
      "loss": 5.8003,
      "mean_token_accuracy": 0.19436246007680893,
      "num_tokens": 32945524.0,
      "step": 12995
    },
    {
      "entropy": 6.21461296081543,
      "epoch": 1.5002885170225042,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047855796683004534,
      "loss": 5.8128,
      "mean_token_accuracy": 0.19227809607982635,
      "num_tokens": 32957871.0,
      "step": 13000
    },
    {
      "entropy": 6.2002583026885985,
      "epoch": 1.500865551067513,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047854038294509356,
      "loss": 5.8308,
      "mean_token_accuracy": 0.19294960498809816,
      "num_tokens": 32971727.0,
      "step": 13005
    },
    {
      "entropy": 6.203684711456299,
      "epoch": 1.5014425851125215,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004785227922141621,
      "loss": 5.7663,
      "mean_token_accuracy": 0.1920443966984749,
      "num_tokens": 32984884.0,
      "step": 13010
    },
    {
      "entropy": 6.205079221725464,
      "epoch": 1.5020196191575304,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047850519463784263,
      "loss": 5.8581,
      "mean_token_accuracy": 0.1898329183459282,
      "num_tokens": 32997475.0,
      "step": 13015
    },
    {
      "entropy": 6.231913757324219,
      "epoch": 1.502596653202539,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047848759021672693,
      "loss": 5.8748,
      "mean_token_accuracy": 0.18672763109207152,
      "num_tokens": 33010466.0,
      "step": 13020
    },
    {
      "entropy": 6.273961353302002,
      "epoch": 1.5031736872475476,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004784699789514073,
      "loss": 5.8542,
      "mean_token_accuracy": 0.18486677706241608,
      "num_tokens": 33023806.0,
      "step": 13025
    },
    {
      "entropy": 6.234267950057983,
      "epoch": 1.5037507212925563,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000478452360842476,
      "loss": 5.7825,
      "mean_token_accuracy": 0.1904935657978058,
      "num_tokens": 33036132.0,
      "step": 13030
    },
    {
      "entropy": 6.233999967575073,
      "epoch": 1.504327755337565,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047843473589052557,
      "loss": 5.886,
      "mean_token_accuracy": 0.18371170461177827,
      "num_tokens": 33049706.0,
      "step": 13035
    },
    {
      "entropy": 6.2749849319458,
      "epoch": 1.5049047893825735,
      "grad_norm": 1.0,
      "learning_rate": 0.00047841710409614893,
      "loss": 5.7804,
      "mean_token_accuracy": 0.1917159840464592,
      "num_tokens": 33061725.0,
      "step": 13040
    },
    {
      "entropy": 6.235005903244018,
      "epoch": 1.5054818234275822,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004783994654599389,
      "loss": 5.8085,
      "mean_token_accuracy": 0.19072041660547256,
      "num_tokens": 33074908.0,
      "step": 13045
    },
    {
      "entropy": 6.201947546005249,
      "epoch": 1.5060588574725908,
      "grad_norm": 0.875,
      "learning_rate": 0.00047838181998248897,
      "loss": 5.8169,
      "mean_token_accuracy": 0.18958678990602493,
      "num_tokens": 33086631.0,
      "step": 13050
    },
    {
      "entropy": 6.255438995361328,
      "epoch": 1.5066358915175995,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004783641676643925,
      "loss": 5.8391,
      "mean_token_accuracy": 0.19087422490119935,
      "num_tokens": 33100109.0,
      "step": 13055
    },
    {
      "entropy": 6.192552185058593,
      "epoch": 1.5072129255626083,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047834650850624334,
      "loss": 5.755,
      "mean_token_accuracy": 0.19182460755109787,
      "num_tokens": 33111953.0,
      "step": 13060
    },
    {
      "entropy": 6.172150611877441,
      "epoch": 1.5077899596076167,
      "grad_norm": 0.875,
      "learning_rate": 0.0004783288425086353,
      "loss": 5.7809,
      "mean_token_accuracy": 0.1961912602186203,
      "num_tokens": 33124563.0,
      "step": 13065
    },
    {
      "entropy": 6.2797609806060795,
      "epoch": 1.5083669936526256,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004783111696721627,
      "loss": 5.8617,
      "mean_token_accuracy": 0.19080158174037934,
      "num_tokens": 33137636.0,
      "step": 13070
    },
    {
      "entropy": 6.243266344070435,
      "epoch": 1.508944027697634,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004782934899974199,
      "loss": 5.9213,
      "mean_token_accuracy": 0.18482270538806916,
      "num_tokens": 33150209.0,
      "step": 13075
    },
    {
      "entropy": 6.112814474105835,
      "epoch": 1.5095210617426429,
      "grad_norm": 0.875,
      "learning_rate": 0.00047827580348500147,
      "loss": 5.7268,
      "mean_token_accuracy": 0.19872756600379943,
      "num_tokens": 33162424.0,
      "step": 13080
    },
    {
      "entropy": 6.214922475814819,
      "epoch": 1.5100980957876513,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047825811013550246,
      "loss": 5.8235,
      "mean_token_accuracy": 0.19208849370479583,
      "num_tokens": 33174923.0,
      "step": 13085
    },
    {
      "entropy": 6.185507202148438,
      "epoch": 1.5106751298326602,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00047824040994951786,
      "loss": 5.729,
      "mean_token_accuracy": 0.1871372252702713,
      "num_tokens": 33188022.0,
      "step": 13090
    },
    {
      "entropy": 6.217069864273071,
      "epoch": 1.5112521638776688,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004782227029276429,
      "loss": 5.8164,
      "mean_token_accuracy": 0.19105395972728728,
      "num_tokens": 33201657.0,
      "step": 13095
    },
    {
      "entropy": 6.217411088943481,
      "epoch": 1.5118291979226774,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047820498907047345,
      "loss": 5.8085,
      "mean_token_accuracy": 0.19143300056457518,
      "num_tokens": 33214692.0,
      "step": 13100
    },
    {
      "entropy": 6.236828088760376,
      "epoch": 1.512406231967686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000478187268378605,
      "loss": 5.8403,
      "mean_token_accuracy": 0.19182227104902266,
      "num_tokens": 33227419.0,
      "step": 13105
    },
    {
      "entropy": 6.2796234607696535,
      "epoch": 1.5129832660126947,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047816954085263375,
      "loss": 5.8486,
      "mean_token_accuracy": 0.18778171092271806,
      "num_tokens": 33239859.0,
      "step": 13110
    },
    {
      "entropy": 6.2108406066894535,
      "epoch": 1.5135603000577034,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004781518064931559,
      "loss": 5.8003,
      "mean_token_accuracy": 0.1949309453368187,
      "num_tokens": 33252872.0,
      "step": 13115
    },
    {
      "entropy": 6.223305511474609,
      "epoch": 1.514137334102712,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000478134065300768,
      "loss": 5.7583,
      "mean_token_accuracy": 0.19917782694101333,
      "num_tokens": 33264220.0,
      "step": 13120
    },
    {
      "entropy": 6.233260774612427,
      "epoch": 1.5147143681477209,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004781163172760667,
      "loss": 5.8974,
      "mean_token_accuracy": 0.17875912487506868,
      "num_tokens": 33276379.0,
      "step": 13125
    },
    {
      "entropy": 6.147059535980224,
      "epoch": 1.5152914021927293,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047809856241964893,
      "loss": 5.769,
      "mean_token_accuracy": 0.19307036995887755,
      "num_tokens": 33290487.0,
      "step": 13130
    },
    {
      "entropy": 6.265978097915649,
      "epoch": 1.5158684362377381,
      "grad_norm": 0.875,
      "learning_rate": 0.0004780808007321119,
      "loss": 5.924,
      "mean_token_accuracy": 0.17739722579717637,
      "num_tokens": 33302560.0,
      "step": 13135
    },
    {
      "entropy": 6.307068920135498,
      "epoch": 1.5164454702827466,
      "grad_norm": 0.875,
      "learning_rate": 0.0004780630322140531,
      "loss": 5.8716,
      "mean_token_accuracy": 0.18288936614990234,
      "num_tokens": 33315077.0,
      "step": 13140
    },
    {
      "entropy": 6.2329202651977536,
      "epoch": 1.5170225043277554,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047804525686607,
      "loss": 5.8688,
      "mean_token_accuracy": 0.189828522503376,
      "num_tokens": 33326769.0,
      "step": 13145
    },
    {
      "entropy": 6.048126268386841,
      "epoch": 1.5175995383727638,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004780274746887606,
      "loss": 5.6639,
      "mean_token_accuracy": 0.2002888947725296,
      "num_tokens": 33340723.0,
      "step": 13150
    },
    {
      "entropy": 6.2914244651794435,
      "epoch": 1.5181765724177727,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047800968568272284,
      "loss": 5.8539,
      "mean_token_accuracy": 0.19260597378015518,
      "num_tokens": 33353894.0,
      "step": 13155
    },
    {
      "entropy": 6.313607263565063,
      "epoch": 1.5187536064627813,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004779918898485551,
      "loss": 5.94,
      "mean_token_accuracy": 0.1844343960285187,
      "num_tokens": 33366134.0,
      "step": 13160
    },
    {
      "entropy": 6.30645718574524,
      "epoch": 1.51933064050779,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047797408718685614,
      "loss": 5.8249,
      "mean_token_accuracy": 0.1886780485510826,
      "num_tokens": 33377604.0,
      "step": 13165
    },
    {
      "entropy": 6.249757289886475,
      "epoch": 1.5199076745527986,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047795627769822447,
      "loss": 5.8812,
      "mean_token_accuracy": 0.18056475669145583,
      "num_tokens": 33389616.0,
      "step": 13170
    },
    {
      "entropy": 6.233608388900757,
      "epoch": 1.5204847085978073,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047793846138325925,
      "loss": 5.8297,
      "mean_token_accuracy": 0.19117956161499022,
      "num_tokens": 33401493.0,
      "step": 13175
    },
    {
      "entropy": 6.200682020187378,
      "epoch": 1.521061742642816,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004779206382425598,
      "loss": 5.7711,
      "mean_token_accuracy": 0.19977571964263915,
      "num_tokens": 33414088.0,
      "step": 13180
    },
    {
      "entropy": 6.275373935699463,
      "epoch": 1.5216387766878245,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004779028082767253,
      "loss": 5.8414,
      "mean_token_accuracy": 0.18880551755428315,
      "num_tokens": 33426481.0,
      "step": 13185
    },
    {
      "entropy": 6.268014669418335,
      "epoch": 1.5222158107328332,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004778849714863557,
      "loss": 5.8176,
      "mean_token_accuracy": 0.18684846311807632,
      "num_tokens": 33437876.0,
      "step": 13190
    },
    {
      "entropy": 6.288102436065674,
      "epoch": 1.5227928447778418,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004778671278720508,
      "loss": 5.8601,
      "mean_token_accuracy": 0.18551743030548096,
      "num_tokens": 33450164.0,
      "step": 13195
    },
    {
      "entropy": 6.171924066543579,
      "epoch": 1.5233698788228507,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004778492774344109,
      "loss": 5.7668,
      "mean_token_accuracy": 0.19991983324289322,
      "num_tokens": 33462583.0,
      "step": 13200
    },
    {
      "entropy": 6.155540800094604,
      "epoch": 1.523946912867859,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004778314201740363,
      "loss": 5.829,
      "mean_token_accuracy": 0.19874422997236252,
      "num_tokens": 33474099.0,
      "step": 13205
    },
    {
      "entropy": 6.2525571346282955,
      "epoch": 1.524523946912868,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047781355609152764,
      "loss": 5.813,
      "mean_token_accuracy": 0.1912405461072922,
      "num_tokens": 33486565.0,
      "step": 13210
    },
    {
      "entropy": 6.24564561843872,
      "epoch": 1.5251009809578764,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004777956851874858,
      "loss": 5.7811,
      "mean_token_accuracy": 0.1879460483789444,
      "num_tokens": 33498703.0,
      "step": 13215
    },
    {
      "entropy": 6.225109434127807,
      "epoch": 1.5256780150028852,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047777780746251176,
      "loss": 5.8536,
      "mean_token_accuracy": 0.18774268627166749,
      "num_tokens": 33511431.0,
      "step": 13220
    },
    {
      "entropy": 6.213768196105957,
      "epoch": 1.5262550490478937,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00047775992291720687,
      "loss": 5.7701,
      "mean_token_accuracy": 0.1881382629275322,
      "num_tokens": 33524771.0,
      "step": 13225
    },
    {
      "entropy": 6.208869457244873,
      "epoch": 1.5268320830929025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004777420315521728,
      "loss": 5.7672,
      "mean_token_accuracy": 0.1989435777068138,
      "num_tokens": 33538362.0,
      "step": 13230
    },
    {
      "entropy": 6.221317911148072,
      "epoch": 1.5274091171379112,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004777241333680111,
      "loss": 5.8593,
      "mean_token_accuracy": 0.18982814997434616,
      "num_tokens": 33550905.0,
      "step": 13235
    },
    {
      "entropy": 6.287773847579956,
      "epoch": 1.5279861511829198,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004777062283653239,
      "loss": 5.8851,
      "mean_token_accuracy": 0.1905108854174614,
      "num_tokens": 33563665.0,
      "step": 13240
    },
    {
      "entropy": 6.213352394104004,
      "epoch": 1.5285631852279284,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00047768831654471333,
      "loss": 5.7839,
      "mean_token_accuracy": 0.19638318419456482,
      "num_tokens": 33576957.0,
      "step": 13245
    },
    {
      "entropy": 6.242417716979981,
      "epoch": 1.529140219272937,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047767039790678204,
      "loss": 5.8294,
      "mean_token_accuracy": 0.18365730494260787,
      "num_tokens": 33589641.0,
      "step": 13250
    },
    {
      "entropy": 6.222768926620484,
      "epoch": 1.5297172533179457,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047765247245213255,
      "loss": 5.7323,
      "mean_token_accuracy": 0.19437995702028274,
      "num_tokens": 33602377.0,
      "step": 13255
    },
    {
      "entropy": 6.233965301513672,
      "epoch": 1.5302942873629544,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004776345401813678,
      "loss": 5.7698,
      "mean_token_accuracy": 0.19175267964601517,
      "num_tokens": 33614796.0,
      "step": 13260
    },
    {
      "entropy": 6.317765617370606,
      "epoch": 1.5308713214079632,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047761660109509095,
      "loss": 5.9284,
      "mean_token_accuracy": 0.18121396601200104,
      "num_tokens": 33626988.0,
      "step": 13265
    },
    {
      "entropy": 6.15206093788147,
      "epoch": 1.5314483554529716,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004775986551939054,
      "loss": 5.798,
      "mean_token_accuracy": 0.19067296236753464,
      "num_tokens": 33641376.0,
      "step": 13270
    },
    {
      "entropy": 6.2937479496002195,
      "epoch": 1.5320253894979805,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047758070247841465,
      "loss": 5.8202,
      "mean_token_accuracy": 0.18697982132434846,
      "num_tokens": 33655155.0,
      "step": 13275
    },
    {
      "entropy": 6.272083187103272,
      "epoch": 1.532602423542989,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047756274294922266,
      "loss": 5.7967,
      "mean_token_accuracy": 0.19397332668304443,
      "num_tokens": 33668480.0,
      "step": 13280
    },
    {
      "entropy": 6.156498289108276,
      "epoch": 1.5331794575879978,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004775447766069334,
      "loss": 5.7287,
      "mean_token_accuracy": 0.1943129852414131,
      "num_tokens": 33680919.0,
      "step": 13285
    },
    {
      "entropy": 6.163669586181641,
      "epoch": 1.5337564916330062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047752680345215115,
      "loss": 5.8106,
      "mean_token_accuracy": 0.1865456521511078,
      "num_tokens": 33693161.0,
      "step": 13290
    },
    {
      "entropy": 6.2926802158355715,
      "epoch": 1.534333525678015,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004775088234854805,
      "loss": 5.8601,
      "mean_token_accuracy": 0.19105806648731233,
      "num_tokens": 33707537.0,
      "step": 13295
    },
    {
      "entropy": 6.242666339874267,
      "epoch": 1.5349105597230237,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004774908367075262,
      "loss": 5.7786,
      "mean_token_accuracy": 0.18832060545682908,
      "num_tokens": 33719501.0,
      "step": 13300
    },
    {
      "entropy": 6.266810750961303,
      "epoch": 1.5354875937680323,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004774728431188931,
      "loss": 5.9159,
      "mean_token_accuracy": 0.1829509034752846,
      "num_tokens": 33731938.0,
      "step": 13305
    },
    {
      "entropy": 6.270457029342651,
      "epoch": 1.536064627813041,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047745484272018664,
      "loss": 5.8622,
      "mean_token_accuracy": 0.19106511473655702,
      "num_tokens": 33745509.0,
      "step": 13310
    },
    {
      "entropy": 6.274699735641479,
      "epoch": 1.5366416618580496,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004774368355120119,
      "loss": 5.7956,
      "mean_token_accuracy": 0.192768432199955,
      "num_tokens": 33757818.0,
      "step": 13315
    },
    {
      "entropy": 6.188769435882568,
      "epoch": 1.5372186959030583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477418821494975,
      "loss": 5.7936,
      "mean_token_accuracy": 0.1924437552690506,
      "num_tokens": 33769840.0,
      "step": 13320
    },
    {
      "entropy": 6.206253099441528,
      "epoch": 1.537795729948067,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004774008006696814,
      "loss": 5.7984,
      "mean_token_accuracy": 0.1900654211640358,
      "num_tokens": 33783713.0,
      "step": 13325
    },
    {
      "entropy": 6.199253225326538,
      "epoch": 1.5383727639930755,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004773827730367375,
      "loss": 5.8025,
      "mean_token_accuracy": 0.19751518219709396,
      "num_tokens": 33796005.0,
      "step": 13330
    },
    {
      "entropy": 6.257584905624389,
      "epoch": 1.5389497980380842,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047736473859674955,
      "loss": 5.7754,
      "mean_token_accuracy": 0.19293928444385527,
      "num_tokens": 33807295.0,
      "step": 13335
    },
    {
      "entropy": 6.2080999374389645,
      "epoch": 1.539526832083093,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047734669735032404,
      "loss": 5.7354,
      "mean_token_accuracy": 0.189118330180645,
      "num_tokens": 33819772.0,
      "step": 13340
    },
    {
      "entropy": 6.178933811187744,
      "epoch": 1.5401038661281015,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047732864929806796,
      "loss": 5.8402,
      "mean_token_accuracy": 0.18972801864147187,
      "num_tokens": 33831820.0,
      "step": 13345
    },
    {
      "entropy": 6.274142456054688,
      "epoch": 1.5406809001731103,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004773105944405883,
      "loss": 5.8941,
      "mean_token_accuracy": 0.18268177211284636,
      "num_tokens": 33845753.0,
      "step": 13350
    },
    {
      "entropy": 6.309782934188843,
      "epoch": 1.5412579342181187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047729253277849226,
      "loss": 5.8064,
      "mean_token_accuracy": 0.19223188608884811,
      "num_tokens": 33859438.0,
      "step": 13355
    },
    {
      "entropy": 6.227335119247437,
      "epoch": 1.5418349682631276,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047727446431238734,
      "loss": 5.8457,
      "mean_token_accuracy": 0.192528435587883,
      "num_tokens": 33872062.0,
      "step": 13360
    },
    {
      "entropy": 6.175086879730225,
      "epoch": 1.542412002308136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004772563890428813,
      "loss": 5.7827,
      "mean_token_accuracy": 0.1981943279504776,
      "num_tokens": 33883840.0,
      "step": 13365
    },
    {
      "entropy": 6.276683330535889,
      "epoch": 1.5429890363531449,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004772383069705821,
      "loss": 5.8721,
      "mean_token_accuracy": 0.18426510393619538,
      "num_tokens": 33896907.0,
      "step": 13370
    },
    {
      "entropy": 6.157901382446289,
      "epoch": 1.5435660703981535,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004772202180960978,
      "loss": 5.8024,
      "mean_token_accuracy": 0.1902434542775154,
      "num_tokens": 33912135.0,
      "step": 13375
    },
    {
      "entropy": 6.211959362030029,
      "epoch": 1.5441431044431622,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047720212242003703,
      "loss": 5.765,
      "mean_token_accuracy": 0.19536373466253282,
      "num_tokens": 33925014.0,
      "step": 13380
    },
    {
      "entropy": 6.333413028717041,
      "epoch": 1.5447201384881708,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047718401994300825,
      "loss": 5.8917,
      "mean_token_accuracy": 0.18318775445222854,
      "num_tokens": 33937429.0,
      "step": 13385
    },
    {
      "entropy": 6.246161365509034,
      "epoch": 1.5452971725331794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047716591066562043,
      "loss": 5.8121,
      "mean_token_accuracy": 0.1933966115117073,
      "num_tokens": 33951030.0,
      "step": 13390
    },
    {
      "entropy": 6.27585883140564,
      "epoch": 1.545874206578188,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047714779458848255,
      "loss": 5.8985,
      "mean_token_accuracy": 0.1845153197646141,
      "num_tokens": 33962930.0,
      "step": 13395
    },
    {
      "entropy": 6.228730773925781,
      "epoch": 1.5464512406231967,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004771296717122041,
      "loss": 5.789,
      "mean_token_accuracy": 0.1970171809196472,
      "num_tokens": 33977113.0,
      "step": 13400
    },
    {
      "entropy": 6.0996129512786865,
      "epoch": 1.5470282746682056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004771115420373945,
      "loss": 5.5435,
      "mean_token_accuracy": 0.2081379696726799,
      "num_tokens": 33989091.0,
      "step": 13405
    },
    {
      "entropy": 6.2149005889892575,
      "epoch": 1.547605308713214,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047709340556466366,
      "loss": 5.9253,
      "mean_token_accuracy": 0.18596762120723725,
      "num_tokens": 34002051.0,
      "step": 13410
    },
    {
      "entropy": 6.290606355667114,
      "epoch": 1.5481823427582229,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047707526229462144,
      "loss": 5.8423,
      "mean_token_accuracy": 0.19136263728141784,
      "num_tokens": 34014482.0,
      "step": 13415
    },
    {
      "entropy": 6.237525987625122,
      "epoch": 1.5487593768032313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047705711222787816,
      "loss": 5.8351,
      "mean_token_accuracy": 0.19333430975675583,
      "num_tokens": 34025665.0,
      "step": 13420
    },
    {
      "entropy": 6.227857971191407,
      "epoch": 1.5493364108482401,
      "grad_norm": 0.875,
      "learning_rate": 0.00047703895536504423,
      "loss": 5.7966,
      "mean_token_accuracy": 0.19315598905086517,
      "num_tokens": 34038163.0,
      "step": 13425
    },
    {
      "entropy": 6.222235202789307,
      "epoch": 1.5499134448932486,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004770207917067305,
      "loss": 5.8474,
      "mean_token_accuracy": 0.19160848557949067,
      "num_tokens": 34050774.0,
      "step": 13430
    },
    {
      "entropy": 6.283314561843872,
      "epoch": 1.5504904789382574,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047700262125354765,
      "loss": 5.8752,
      "mean_token_accuracy": 0.1832515552639961,
      "num_tokens": 34062480.0,
      "step": 13435
    },
    {
      "entropy": 6.219641160964966,
      "epoch": 1.551067512983266,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047698444400610707,
      "loss": 5.7992,
      "mean_token_accuracy": 0.1960041493177414,
      "num_tokens": 34075314.0,
      "step": 13440
    },
    {
      "entropy": 6.152193546295166,
      "epoch": 1.5516445470282747,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047696625996502,
      "loss": 5.7307,
      "mean_token_accuracy": 0.19171851575374604,
      "num_tokens": 34087528.0,
      "step": 13445
    },
    {
      "entropy": 6.170001602172851,
      "epoch": 1.5522215810732833,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047694806913089815,
      "loss": 5.7191,
      "mean_token_accuracy": 0.19241029024124146,
      "num_tokens": 34101740.0,
      "step": 13450
    },
    {
      "entropy": 6.293499422073364,
      "epoch": 1.552798615118292,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004769298715043533,
      "loss": 5.8261,
      "mean_token_accuracy": 0.18749757260084152,
      "num_tokens": 34113042.0,
      "step": 13455
    },
    {
      "entropy": 6.198850536346436,
      "epoch": 1.5533756491633006,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004769116670859975,
      "loss": 5.8842,
      "mean_token_accuracy": 0.18741277307271959,
      "num_tokens": 34126801.0,
      "step": 13460
    },
    {
      "entropy": 6.255635404586792,
      "epoch": 1.5539526832083093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047689345587644314,
      "loss": 5.8821,
      "mean_token_accuracy": 0.18848055005073547,
      "num_tokens": 34139469.0,
      "step": 13465
    },
    {
      "entropy": 6.278279447555542,
      "epoch": 1.554529717253318,
      "grad_norm": 0.875,
      "learning_rate": 0.00047687523787630256,
      "loss": 5.818,
      "mean_token_accuracy": 0.19236057996749878,
      "num_tokens": 34151940.0,
      "step": 13470
    },
    {
      "entropy": 6.183226442337036,
      "epoch": 1.5551067512983265,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004768570130861887,
      "loss": 5.7705,
      "mean_token_accuracy": 0.1915515825152397,
      "num_tokens": 34164375.0,
      "step": 13475
    },
    {
      "entropy": 6.236829614639282,
      "epoch": 1.5556837853433354,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004768387815067144,
      "loss": 5.8512,
      "mean_token_accuracy": 0.18451953530311585,
      "num_tokens": 34176585.0,
      "step": 13480
    },
    {
      "entropy": 6.277500629425049,
      "epoch": 1.5562608193883438,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047682054313849304,
      "loss": 5.835,
      "mean_token_accuracy": 0.190140999853611,
      "num_tokens": 34188518.0,
      "step": 13485
    },
    {
      "entropy": 6.195441627502442,
      "epoch": 1.5568378534333527,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004768022979821379,
      "loss": 5.7377,
      "mean_token_accuracy": 0.19830055981874467,
      "num_tokens": 34200865.0,
      "step": 13490
    },
    {
      "entropy": 6.2230743885040285,
      "epoch": 1.557414887478361,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004767840460382628,
      "loss": 5.8052,
      "mean_token_accuracy": 0.19546588659286498,
      "num_tokens": 34212766.0,
      "step": 13495
    },
    {
      "entropy": 6.233663177490234,
      "epoch": 1.55799192152337,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004767657873074815,
      "loss": 5.8255,
      "mean_token_accuracy": 0.18949540704488754,
      "num_tokens": 34225401.0,
      "step": 13500
    },
    {
      "entropy": 6.216677951812744,
      "epoch": 1.5585689555683784,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004767475217904081,
      "loss": 5.7761,
      "mean_token_accuracy": 0.19593921899795533,
      "num_tokens": 34237474.0,
      "step": 13505
    },
    {
      "entropy": 6.242624187469483,
      "epoch": 1.5591459896133872,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047672924948765705,
      "loss": 5.8777,
      "mean_token_accuracy": 0.18572050034999849,
      "num_tokens": 34249598.0,
      "step": 13510
    },
    {
      "entropy": 6.201598834991455,
      "epoch": 1.5597230236583959,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047671097039984293,
      "loss": 5.7717,
      "mean_token_accuracy": 0.19565215557813645,
      "num_tokens": 34261773.0,
      "step": 13515
    },
    {
      "entropy": 6.233489322662353,
      "epoch": 1.5603000577034045,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047669268452758053,
      "loss": 5.8435,
      "mean_token_accuracy": 0.19406123608350753,
      "num_tokens": 34275398.0,
      "step": 13520
    },
    {
      "entropy": 6.212414884567261,
      "epoch": 1.5608770917484132,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047667439187148476,
      "loss": 5.7721,
      "mean_token_accuracy": 0.19103527963161468,
      "num_tokens": 34288934.0,
      "step": 13525
    },
    {
      "entropy": 6.206125164031983,
      "epoch": 1.5614541257934218,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004766560924321711,
      "loss": 5.8099,
      "mean_token_accuracy": 0.18824598640203477,
      "num_tokens": 34302187.0,
      "step": 13530
    },
    {
      "entropy": 6.222528457641602,
      "epoch": 1.5620311598384304,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047663778621025496,
      "loss": 5.811,
      "mean_token_accuracy": 0.18573870360851288,
      "num_tokens": 34314745.0,
      "step": 13535
    },
    {
      "entropy": 6.242219161987305,
      "epoch": 1.562608193883439,
      "grad_norm": 0.875,
      "learning_rate": 0.0004766194732063519,
      "loss": 5.7729,
      "mean_token_accuracy": 0.19454296082258224,
      "num_tokens": 34326771.0,
      "step": 13540
    },
    {
      "entropy": 6.192319583892822,
      "epoch": 1.563185227928448,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047660115342107814,
      "loss": 5.7961,
      "mean_token_accuracy": 0.1913350760936737,
      "num_tokens": 34339452.0,
      "step": 13545
    },
    {
      "entropy": 6.252946901321411,
      "epoch": 1.5637622619734564,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047658282685504973,
      "loss": 5.8411,
      "mean_token_accuracy": 0.1796528935432434,
      "num_tokens": 34351872.0,
      "step": 13550
    },
    {
      "entropy": 6.245566129684448,
      "epoch": 1.5643392960184652,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000476564493508883,
      "loss": 5.7962,
      "mean_token_accuracy": 0.19455188810825347,
      "num_tokens": 34363963.0,
      "step": 13555
    },
    {
      "entropy": 6.2495848655700685,
      "epoch": 1.5649163300634736,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047654615338319466,
      "loss": 5.8251,
      "mean_token_accuracy": 0.18784345984458922,
      "num_tokens": 34377245.0,
      "step": 13560
    },
    {
      "entropy": 6.158248710632324,
      "epoch": 1.5654933641084825,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004765278064786016,
      "loss": 5.7005,
      "mean_token_accuracy": 0.19669419080018996,
      "num_tokens": 34391047.0,
      "step": 13565
    },
    {
      "entropy": 6.264983034133911,
      "epoch": 1.566070398153491,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047650945279572085,
      "loss": 5.8083,
      "mean_token_accuracy": 0.1852528780698776,
      "num_tokens": 34404655.0,
      "step": 13570
    },
    {
      "entropy": 6.232136964797974,
      "epoch": 1.5666474321984998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004764910923351698,
      "loss": 5.7447,
      "mean_token_accuracy": 0.19516263604164125,
      "num_tokens": 34417902.0,
      "step": 13575
    },
    {
      "entropy": 6.2776096820831295,
      "epoch": 1.5672244662435084,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047647272509756584,
      "loss": 5.7975,
      "mean_token_accuracy": 0.19093846678733825,
      "num_tokens": 34430630.0,
      "step": 13580
    },
    {
      "entropy": 6.210801124572754,
      "epoch": 1.567801500288517,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004764543510835269,
      "loss": 5.8288,
      "mean_token_accuracy": 0.18807119280099868,
      "num_tokens": 34444320.0,
      "step": 13585
    },
    {
      "entropy": 6.17142686843872,
      "epoch": 1.5683785343335257,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000476435970293671,
      "loss": 5.7107,
      "mean_token_accuracy": 0.20341715961694717,
      "num_tokens": 34457081.0,
      "step": 13590
    },
    {
      "entropy": 6.19086971282959,
      "epoch": 1.5689555683785343,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047641758272861626,
      "loss": 5.761,
      "mean_token_accuracy": 0.19833941459655763,
      "num_tokens": 34469560.0,
      "step": 13595
    },
    {
      "entropy": 6.240159606933593,
      "epoch": 1.569532602423543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004763991883889811,
      "loss": 5.8156,
      "mean_token_accuracy": 0.18774486780166627,
      "num_tokens": 34482036.0,
      "step": 13600
    },
    {
      "entropy": 6.257405138015747,
      "epoch": 1.5701096364685516,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004763807872753843,
      "loss": 5.7771,
      "mean_token_accuracy": 0.19095739275217055,
      "num_tokens": 34494316.0,
      "step": 13605
    },
    {
      "entropy": 6.18026556968689,
      "epoch": 1.5706866705135603,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047636237938844484,
      "loss": 5.7673,
      "mean_token_accuracy": 0.1896099030971527,
      "num_tokens": 34507026.0,
      "step": 13610
    },
    {
      "entropy": 6.208980464935303,
      "epoch": 1.571263704558569,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004763439647287817,
      "loss": 5.7579,
      "mean_token_accuracy": 0.19755264073610307,
      "num_tokens": 34519910.0,
      "step": 13615
    },
    {
      "entropy": 6.20766921043396,
      "epoch": 1.5718407386035778,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004763255432970144,
      "loss": 5.824,
      "mean_token_accuracy": 0.1928902894258499,
      "num_tokens": 34534832.0,
      "step": 13620
    },
    {
      "entropy": 6.233760976791382,
      "epoch": 1.5724177726485862,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00047630711509376235,
      "loss": 5.8214,
      "mean_token_accuracy": 0.18949252963066102,
      "num_tokens": 34548559.0,
      "step": 13625
    },
    {
      "entropy": 6.204715299606323,
      "epoch": 1.572994806693595,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004762886801196455,
      "loss": 5.8505,
      "mean_token_accuracy": 0.19321168810129166,
      "num_tokens": 34561131.0,
      "step": 13630
    },
    {
      "entropy": 6.259569549560547,
      "epoch": 1.5735718407386035,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047627023837528386,
      "loss": 5.8085,
      "mean_token_accuracy": 0.19253587871789932,
      "num_tokens": 34572863.0,
      "step": 13635
    },
    {
      "entropy": 6.189451503753662,
      "epoch": 1.5741488747836123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047625178986129775,
      "loss": 5.648,
      "mean_token_accuracy": 0.20012484341859818,
      "num_tokens": 34586277.0,
      "step": 13640
    },
    {
      "entropy": 6.14692759513855,
      "epoch": 1.5747259088286207,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004762333345783078,
      "loss": 5.7398,
      "mean_token_accuracy": 0.195273657143116,
      "num_tokens": 34598112.0,
      "step": 13645
    },
    {
      "entropy": 6.145579242706299,
      "epoch": 1.5753029428736296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047621487252693436,
      "loss": 5.7552,
      "mean_token_accuracy": 0.19781942367553712,
      "num_tokens": 34610569.0,
      "step": 13650
    },
    {
      "entropy": 6.259181547164917,
      "epoch": 1.5758799769186382,
      "grad_norm": 0.875,
      "learning_rate": 0.00047619640370779876,
      "loss": 5.8231,
      "mean_token_accuracy": 0.18996772468090056,
      "num_tokens": 34623148.0,
      "step": 13655
    },
    {
      "entropy": 6.224069452285766,
      "epoch": 1.5764570109636469,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047617792812152207,
      "loss": 5.801,
      "mean_token_accuracy": 0.1926830917596817,
      "num_tokens": 34635260.0,
      "step": 13660
    },
    {
      "entropy": 6.101304340362549,
      "epoch": 1.5770340450086555,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004761594457687256,
      "loss": 5.7098,
      "mean_token_accuracy": 0.20209331214427947,
      "num_tokens": 34648717.0,
      "step": 13665
    },
    {
      "entropy": 6.251263236999511,
      "epoch": 1.5776110790536642,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004761409566500313,
      "loss": 5.7742,
      "mean_token_accuracy": 0.19245714098215103,
      "num_tokens": 34662442.0,
      "step": 13670
    },
    {
      "entropy": 6.3070228099823,
      "epoch": 1.5781881130986728,
      "grad_norm": 0.875,
      "learning_rate": 0.00047612246076606066,
      "loss": 5.7815,
      "mean_token_accuracy": 0.19415925741195678,
      "num_tokens": 34675213.0,
      "step": 13675
    },
    {
      "entropy": 6.229634523391724,
      "epoch": 1.5787651471436814,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047610395811743594,
      "loss": 5.7687,
      "mean_token_accuracy": 0.1912509709596634,
      "num_tokens": 34687959.0,
      "step": 13680
    },
    {
      "entropy": 6.159662103652954,
      "epoch": 1.5793421811886903,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047608544870477956,
      "loss": 5.7309,
      "mean_token_accuracy": 0.19864692240953447,
      "num_tokens": 34700242.0,
      "step": 13685
    },
    {
      "entropy": 6.176917362213135,
      "epoch": 1.5799192152336987,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047606693252871395,
      "loss": 5.7295,
      "mean_token_accuracy": 0.20046167075634003,
      "num_tokens": 34713245.0,
      "step": 13690
    },
    {
      "entropy": 6.25925726890564,
      "epoch": 1.5804962492787076,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000476048409589862,
      "loss": 5.882,
      "mean_token_accuracy": 0.18273352235555648,
      "num_tokens": 34725248.0,
      "step": 13695
    },
    {
      "entropy": 6.358665800094604,
      "epoch": 1.581073283323716,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004760298798888466,
      "loss": 5.8906,
      "mean_token_accuracy": 0.1849916934967041,
      "num_tokens": 34737801.0,
      "step": 13700
    },
    {
      "entropy": 6.229224681854248,
      "epoch": 1.5816503173687249,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004760113434262911,
      "loss": 5.8509,
      "mean_token_accuracy": 0.1848498120903969,
      "num_tokens": 34749504.0,
      "step": 13705
    },
    {
      "entropy": 6.223398637771607,
      "epoch": 1.5822273514137333,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047599280020281894,
      "loss": 5.8698,
      "mean_token_accuracy": 0.18317728042602538,
      "num_tokens": 34761238.0,
      "step": 13710
    },
    {
      "entropy": 6.308594274520874,
      "epoch": 1.5828043854587421,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047597425021905364,
      "loss": 5.8625,
      "mean_token_accuracy": 0.18460193276405334,
      "num_tokens": 34773662.0,
      "step": 13715
    },
    {
      "entropy": 6.263258838653565,
      "epoch": 1.5833814195037508,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004759556934756194,
      "loss": 5.823,
      "mean_token_accuracy": 0.18266422748565675,
      "num_tokens": 34786774.0,
      "step": 13720
    },
    {
      "entropy": 6.175289869308472,
      "epoch": 1.5839584535487594,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047593712997314017,
      "loss": 5.7764,
      "mean_token_accuracy": 0.19383132159709932,
      "num_tokens": 34798169.0,
      "step": 13725
    },
    {
      "entropy": 6.165880966186523,
      "epoch": 1.584535487593768,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047591855971224035,
      "loss": 5.7435,
      "mean_token_accuracy": 0.19536821842193602,
      "num_tokens": 34809870.0,
      "step": 13730
    },
    {
      "entropy": 6.26498122215271,
      "epoch": 1.5851125216387767,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004758999826935446,
      "loss": 5.8992,
      "mean_token_accuracy": 0.18588720113039017,
      "num_tokens": 34821684.0,
      "step": 13735
    },
    {
      "entropy": 6.2704390525817875,
      "epoch": 1.5856895556837853,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004758813989176778,
      "loss": 5.7928,
      "mean_token_accuracy": 0.19280680119991303,
      "num_tokens": 34834905.0,
      "step": 13740
    },
    {
      "entropy": 6.181961631774902,
      "epoch": 1.586266589728794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047586280838526483,
      "loss": 5.7324,
      "mean_token_accuracy": 0.19703560769557954,
      "num_tokens": 34846797.0,
      "step": 13745
    },
    {
      "entropy": 6.261712551116943,
      "epoch": 1.5868436237738026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004758442110969312,
      "loss": 5.7953,
      "mean_token_accuracy": 0.19001741856336593,
      "num_tokens": 34859261.0,
      "step": 13750
    },
    {
      "entropy": 6.253963470458984,
      "epoch": 1.5874206578188113,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004758256070533022,
      "loss": 5.8478,
      "mean_token_accuracy": 0.18807396292686462,
      "num_tokens": 34871775.0,
      "step": 13755
    },
    {
      "entropy": 6.179570960998535,
      "epoch": 1.5879976918638201,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004758069962550037,
      "loss": 5.7235,
      "mean_token_accuracy": 0.19319581538438796,
      "num_tokens": 34884461.0,
      "step": 13760
    },
    {
      "entropy": 6.273061609268188,
      "epoch": 1.5885747259088285,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047578837870266156,
      "loss": 5.8374,
      "mean_token_accuracy": 0.1873654305934906,
      "num_tokens": 34896772.0,
      "step": 13765
    },
    {
      "entropy": 6.269413852691651,
      "epoch": 1.5891517599538374,
      "grad_norm": 0.875,
      "learning_rate": 0.00047576975439690206,
      "loss": 5.849,
      "mean_token_accuracy": 0.17993571162223815,
      "num_tokens": 34909914.0,
      "step": 13770
    },
    {
      "entropy": 6.258509254455566,
      "epoch": 1.5897287939988458,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047575112333835164,
      "loss": 5.8203,
      "mean_token_accuracy": 0.19258391559123994,
      "num_tokens": 34922298.0,
      "step": 13775
    },
    {
      "entropy": 6.336678504943848,
      "epoch": 1.5903058280438547,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047573248552763684,
      "loss": 5.9481,
      "mean_token_accuracy": 0.18731440156698226,
      "num_tokens": 34935159.0,
      "step": 13780
    },
    {
      "entropy": 6.151774740219116,
      "epoch": 1.590882862088863,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047571384096538474,
      "loss": 5.6668,
      "mean_token_accuracy": 0.19896821528673173,
      "num_tokens": 34948347.0,
      "step": 13785
    },
    {
      "entropy": 6.178550863265992,
      "epoch": 1.591459896133872,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004756951896522222,
      "loss": 5.8317,
      "mean_token_accuracy": 0.1906275674700737,
      "num_tokens": 34961489.0,
      "step": 13790
    },
    {
      "entropy": 6.360392379760742,
      "epoch": 1.5920369301788806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004756765315887766,
      "loss": 5.9569,
      "mean_token_accuracy": 0.17901170253753662,
      "num_tokens": 34974458.0,
      "step": 13795
    },
    {
      "entropy": 6.283846282958985,
      "epoch": 1.5926139642238892,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004756578667756756,
      "loss": 5.7833,
      "mean_token_accuracy": 0.19735266715288163,
      "num_tokens": 34988498.0,
      "step": 13800
    },
    {
      "entropy": 6.029751968383789,
      "epoch": 1.5931909982688979,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004756391952135469,
      "loss": 5.6074,
      "mean_token_accuracy": 0.21699870824813844,
      "num_tokens": 35002101.0,
      "step": 13805
    },
    {
      "entropy": 6.2392970561981205,
      "epoch": 1.5937680323139065,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047562051690301855,
      "loss": 5.9164,
      "mean_token_accuracy": 0.18663058280944825,
      "num_tokens": 35014079.0,
      "step": 13810
    },
    {
      "entropy": 6.235805511474609,
      "epoch": 1.5943450663589152,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047560183184471873,
      "loss": 5.7331,
      "mean_token_accuracy": 0.19860985577106477,
      "num_tokens": 35027455.0,
      "step": 13815
    },
    {
      "entropy": 6.166205787658692,
      "epoch": 1.5949221004039238,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000475583140039276,
      "loss": 5.7424,
      "mean_token_accuracy": 0.1939515545964241,
      "num_tokens": 35040699.0,
      "step": 13820
    },
    {
      "entropy": 6.320330476760864,
      "epoch": 1.5954991344489324,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047556444148731897,
      "loss": 5.8712,
      "mean_token_accuracy": 0.18556371182203293,
      "num_tokens": 35052758.0,
      "step": 13825
    },
    {
      "entropy": 6.224108362197876,
      "epoch": 1.596076168493941,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004755457361894766,
      "loss": 5.8373,
      "mean_token_accuracy": 0.1898344099521637,
      "num_tokens": 35065393.0,
      "step": 13830
    },
    {
      "entropy": 6.266162538528443,
      "epoch": 1.59665320253895,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047552702414637793,
      "loss": 5.8457,
      "mean_token_accuracy": 0.18466853648424147,
      "num_tokens": 35077708.0,
      "step": 13835
    },
    {
      "entropy": 6.2740621089935305,
      "epoch": 1.5972302365839584,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047550830535865245,
      "loss": 5.8637,
      "mean_token_accuracy": 0.18276742249727249,
      "num_tokens": 35088870.0,
      "step": 13840
    },
    {
      "entropy": 6.220741319656372,
      "epoch": 1.5978072706289672,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047548957982692974,
      "loss": 5.7422,
      "mean_token_accuracy": 0.19338622391223909,
      "num_tokens": 35100128.0,
      "step": 13845
    },
    {
      "entropy": 6.250493669509888,
      "epoch": 1.5983843046739756,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004754708475518396,
      "loss": 5.911,
      "mean_token_accuracy": 0.18991688191890715,
      "num_tokens": 35113539.0,
      "step": 13850
    },
    {
      "entropy": 6.31403341293335,
      "epoch": 1.5989613387189845,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047545210853401214,
      "loss": 5.8575,
      "mean_token_accuracy": 0.18729247748851777,
      "num_tokens": 35126897.0,
      "step": 13855
    },
    {
      "entropy": 6.271669244766235,
      "epoch": 1.599538372763993,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047543336277407753,
      "loss": 5.8195,
      "mean_token_accuracy": 0.19559807628393172,
      "num_tokens": 35140197.0,
      "step": 13860
    },
    {
      "entropy": 6.254930686950684,
      "epoch": 1.6001154068090018,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047541461027266624,
      "loss": 5.7888,
      "mean_token_accuracy": 0.1918448105454445,
      "num_tokens": 35151928.0,
      "step": 13865
    },
    {
      "entropy": 6.275524997711182,
      "epoch": 1.6006924408540104,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004753958510304091,
      "loss": 5.8563,
      "mean_token_accuracy": 0.18478392213582992,
      "num_tokens": 35165045.0,
      "step": 13870
    },
    {
      "entropy": 6.219670915603638,
      "epoch": 1.601269474899019,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047537708504793714,
      "loss": 5.7466,
      "mean_token_accuracy": 0.1994813472032547,
      "num_tokens": 35177631.0,
      "step": 13875
    },
    {
      "entropy": 6.153238964080811,
      "epoch": 1.6018465089440277,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047535831232588146,
      "loss": 5.7987,
      "mean_token_accuracy": 0.19119410812854767,
      "num_tokens": 35190166.0,
      "step": 13880
    },
    {
      "entropy": 6.278003787994384,
      "epoch": 1.6024235429890363,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047533953286487345,
      "loss": 5.8794,
      "mean_token_accuracy": 0.1896279901266098,
      "num_tokens": 35202604.0,
      "step": 13885
    },
    {
      "entropy": 6.1833991527557375,
      "epoch": 1.603000577034045,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004753207466655447,
      "loss": 5.6741,
      "mean_token_accuracy": 0.19832319468259813,
      "num_tokens": 35214686.0,
      "step": 13890
    },
    {
      "entropy": 6.272806215286255,
      "epoch": 1.6035776110790536,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004753019537285273,
      "loss": 5.8667,
      "mean_token_accuracy": 0.18936332762241365,
      "num_tokens": 35226230.0,
      "step": 13895
    },
    {
      "entropy": 6.261243581771851,
      "epoch": 1.6041546451240625,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047528315405445296,
      "loss": 5.8112,
      "mean_token_accuracy": 0.18637109994888307,
      "num_tokens": 35240273.0,
      "step": 13900
    },
    {
      "entropy": 6.264919900894165,
      "epoch": 1.604731679169071,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047526434764395435,
      "loss": 5.8702,
      "mean_token_accuracy": 0.18387931287288667,
      "num_tokens": 35253095.0,
      "step": 13905
    },
    {
      "entropy": 6.263901472091675,
      "epoch": 1.6053087132140798,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047524553449766386,
      "loss": 5.7556,
      "mean_token_accuracy": 0.19650789201259614,
      "num_tokens": 35265770.0,
      "step": 13910
    },
    {
      "entropy": 6.1615309715271,
      "epoch": 1.6058857472590882,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047522671461621433,
      "loss": 5.818,
      "mean_token_accuracy": 0.19293110817670822,
      "num_tokens": 35278800.0,
      "step": 13915
    },
    {
      "entropy": 6.26313362121582,
      "epoch": 1.606462781304097,
      "grad_norm": 0.875,
      "learning_rate": 0.0004752078880002386,
      "loss": 5.782,
      "mean_token_accuracy": 0.19223702996969222,
      "num_tokens": 35290423.0,
      "step": 13920
    },
    {
      "entropy": 6.253077173233033,
      "epoch": 1.6070398153491054,
      "grad_norm": 0.78515625,
      "learning_rate": 0.00047518905465037005,
      "loss": 5.8292,
      "mean_token_accuracy": 0.19669198393821716,
      "num_tokens": 35303679.0,
      "step": 13925
    },
    {
      "entropy": 6.296618890762329,
      "epoch": 1.6076168493941143,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047517021456724214,
      "loss": 5.8957,
      "mean_token_accuracy": 0.18560034036636353,
      "num_tokens": 35316900.0,
      "step": 13930
    },
    {
      "entropy": 6.210533380508423,
      "epoch": 1.608193883439123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047515136775148843,
      "loss": 5.807,
      "mean_token_accuracy": 0.19146231114864348,
      "num_tokens": 35329952.0,
      "step": 13935
    },
    {
      "entropy": 6.2653978824615475,
      "epoch": 1.6087709174841316,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004751325142037429,
      "loss": 5.8618,
      "mean_token_accuracy": 0.1843246877193451,
      "num_tokens": 35342107.0,
      "step": 13940
    },
    {
      "entropy": 6.266818904876709,
      "epoch": 1.6093479515291402,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047511365392463974,
      "loss": 5.7742,
      "mean_token_accuracy": 0.19125829190015792,
      "num_tokens": 35354996.0,
      "step": 13945
    },
    {
      "entropy": 6.199180603027344,
      "epoch": 1.6099249855741489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047509478691481317,
      "loss": 5.8318,
      "mean_token_accuracy": 0.18893272280693055,
      "num_tokens": 35367332.0,
      "step": 13950
    },
    {
      "entropy": 6.247616195678711,
      "epoch": 1.6105020196191575,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047507591317489783,
      "loss": 5.7603,
      "mean_token_accuracy": 0.1962567612528801,
      "num_tokens": 35379469.0,
      "step": 13955
    },
    {
      "entropy": 6.228953123092651,
      "epoch": 1.6110790536641661,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004750570327055286,
      "loss": 5.7137,
      "mean_token_accuracy": 0.20355527400970458,
      "num_tokens": 35392579.0,
      "step": 13960
    },
    {
      "entropy": 6.222973489761353,
      "epoch": 1.6116560877091748,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047503814550734034,
      "loss": 5.8288,
      "mean_token_accuracy": 0.18460685461759568,
      "num_tokens": 35405421.0,
      "step": 13965
    },
    {
      "entropy": 6.2718836784362795,
      "epoch": 1.6122331217541834,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004750192515809685,
      "loss": 5.8672,
      "mean_token_accuracy": 0.1915691927075386,
      "num_tokens": 35418046.0,
      "step": 13970
    },
    {
      "entropy": 6.303271913528443,
      "epoch": 1.6128101557991923,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047500035092704843,
      "loss": 5.7897,
      "mean_token_accuracy": 0.19402606189250945,
      "num_tokens": 35430346.0,
      "step": 13975
    },
    {
      "entropy": 6.2423196792602536,
      "epoch": 1.6133871898442007,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004749814435462159,
      "loss": 5.8105,
      "mean_token_accuracy": 0.18759704977273942,
      "num_tokens": 35443607.0,
      "step": 13980
    },
    {
      "entropy": 6.248625755310059,
      "epoch": 1.6139642238892096,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004749625294391069,
      "loss": 5.7845,
      "mean_token_accuracy": 0.18895418345928192,
      "num_tokens": 35455890.0,
      "step": 13985
    },
    {
      "entropy": 6.206034708023071,
      "epoch": 1.614541257934218,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047494360860635755,
      "loss": 5.8132,
      "mean_token_accuracy": 0.19799791723489762,
      "num_tokens": 35470340.0,
      "step": 13990
    },
    {
      "entropy": 6.24536771774292,
      "epoch": 1.6151182919792268,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004749246810486042,
      "loss": 5.7588,
      "mean_token_accuracy": 0.19139318764209748,
      "num_tokens": 35482827.0,
      "step": 13995
    },
    {
      "entropy": 6.3113116264343265,
      "epoch": 1.6156953260242353,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004749057467664836,
      "loss": 5.8804,
      "mean_token_accuracy": 0.1861642614006996,
      "num_tokens": 35494289.0,
      "step": 14000
    },
    {
      "entropy": 6.177021598815918,
      "epoch": 1.6162723600692441,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047488680576063247,
      "loss": 5.7333,
      "mean_token_accuracy": 0.19147407859563828,
      "num_tokens": 35506541.0,
      "step": 14005
    },
    {
      "entropy": 6.135279417037964,
      "epoch": 1.6168493941142528,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004748678580316878,
      "loss": 5.6993,
      "mean_token_accuracy": 0.19809169322252274,
      "num_tokens": 35519628.0,
      "step": 14010
    },
    {
      "entropy": 6.250536823272705,
      "epoch": 1.6174264281592614,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047484890358028714,
      "loss": 5.7114,
      "mean_token_accuracy": 0.19773522019386292,
      "num_tokens": 35532525.0,
      "step": 14015
    },
    {
      "entropy": 6.261101293563843,
      "epoch": 1.61800346220427,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004748299424070678,
      "loss": 5.8457,
      "mean_token_accuracy": 0.18817708939313887,
      "num_tokens": 35544046.0,
      "step": 14020
    },
    {
      "entropy": 6.241388034820557,
      "epoch": 1.6185804962492787,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004748109745126677,
      "loss": 5.8402,
      "mean_token_accuracy": 0.19445383548736572,
      "num_tokens": 35557784.0,
      "step": 14025
    },
    {
      "entropy": 6.326953792572022,
      "epoch": 1.6191575302942873,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047479199989772464,
      "loss": 5.8496,
      "mean_token_accuracy": 0.18643974661827087,
      "num_tokens": 35571472.0,
      "step": 14030
    },
    {
      "entropy": 6.117964601516723,
      "epoch": 1.619734564339296,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000474773018562877,
      "loss": 5.6876,
      "mean_token_accuracy": 0.2141325891017914,
      "num_tokens": 35584992.0,
      "step": 14035
    },
    {
      "entropy": 6.190286445617676,
      "epoch": 1.6203115983843048,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047475403050876297,
      "loss": 5.7406,
      "mean_token_accuracy": 0.19228264689445496,
      "num_tokens": 35596643.0,
      "step": 14040
    },
    {
      "entropy": 6.294740962982178,
      "epoch": 1.6208886324293132,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004747350357360214,
      "loss": 5.8756,
      "mean_token_accuracy": 0.18967580944299697,
      "num_tokens": 35608796.0,
      "step": 14045
    },
    {
      "entropy": 6.214803075790405,
      "epoch": 1.621465666474322,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004747160342452912,
      "loss": 5.759,
      "mean_token_accuracy": 0.1968100354075432,
      "num_tokens": 35620008.0,
      "step": 14050
    },
    {
      "entropy": 6.263499975204468,
      "epoch": 1.6220427005193305,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047469702603721134,
      "loss": 5.7901,
      "mean_token_accuracy": 0.18680422604084015,
      "num_tokens": 35632751.0,
      "step": 14055
    },
    {
      "entropy": 6.180832386016846,
      "epoch": 1.6226197345643394,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004746780111124212,
      "loss": 5.7565,
      "mean_token_accuracy": 0.19398533850908278,
      "num_tokens": 35645031.0,
      "step": 14060
    },
    {
      "entropy": 6.249925374984741,
      "epoch": 1.6231967686093478,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047465898947156033,
      "loss": 5.7778,
      "mean_token_accuracy": 0.1910052165389061,
      "num_tokens": 35656731.0,
      "step": 14065
    },
    {
      "entropy": 6.266978025436401,
      "epoch": 1.6237738026543567,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047463996111526854,
      "loss": 5.7853,
      "mean_token_accuracy": 0.19448653757572174,
      "num_tokens": 35668371.0,
      "step": 14070
    },
    {
      "entropy": 6.255351829528808,
      "epoch": 1.6243508366993653,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047462092604418576,
      "loss": 5.8663,
      "mean_token_accuracy": 0.18993324637413025,
      "num_tokens": 35681339.0,
      "step": 14075
    },
    {
      "entropy": 6.281464242935181,
      "epoch": 1.624927870744374,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047460188425895236,
      "loss": 5.8068,
      "mean_token_accuracy": 0.1934675693511963,
      "num_tokens": 35693775.0,
      "step": 14080
    },
    {
      "entropy": 6.153054094314575,
      "epoch": 1.6255049047893826,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047458283576020874,
      "loss": 5.7072,
      "mean_token_accuracy": 0.2054233267903328,
      "num_tokens": 35705707.0,
      "step": 14085
    },
    {
      "entropy": 6.155913639068603,
      "epoch": 1.6260819388343912,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047456378054859554,
      "loss": 5.7641,
      "mean_token_accuracy": 0.19285307973623275,
      "num_tokens": 35717793.0,
      "step": 14090
    },
    {
      "entropy": 6.216971158981323,
      "epoch": 1.6266589728793999,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047454471862475375,
      "loss": 5.776,
      "mean_token_accuracy": 0.1945430189371109,
      "num_tokens": 35730762.0,
      "step": 14095
    },
    {
      "entropy": 6.256048107147217,
      "epoch": 1.6272360069244085,
      "grad_norm": 1.375,
      "learning_rate": 0.00047452564998932446,
      "loss": 5.75,
      "mean_token_accuracy": 0.19309590607881547,
      "num_tokens": 35742852.0,
      "step": 14100
    },
    {
      "entropy": 6.262771081924439,
      "epoch": 1.6278130409694171,
      "grad_norm": 0.90625,
      "learning_rate": 0.000474506574642949,
      "loss": 5.853,
      "mean_token_accuracy": 0.1850825771689415,
      "num_tokens": 35756416.0,
      "step": 14105
    },
    {
      "entropy": 6.311510276794434,
      "epoch": 1.6283900750144258,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000474487492586269,
      "loss": 5.8708,
      "mean_token_accuracy": 0.18291836231946945,
      "num_tokens": 35768347.0,
      "step": 14110
    },
    {
      "entropy": 6.254725456237793,
      "epoch": 1.6289671090594346,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004744684038199263,
      "loss": 5.7509,
      "mean_token_accuracy": 0.19324211478233339,
      "num_tokens": 35781228.0,
      "step": 14115
    },
    {
      "entropy": 6.233711624145508,
      "epoch": 1.629544143104443,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047444930834456293,
      "loss": 5.7468,
      "mean_token_accuracy": 0.19390194565057756,
      "num_tokens": 35793757.0,
      "step": 14120
    },
    {
      "entropy": 6.293483257293701,
      "epoch": 1.630121177149452,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004744302061608212,
      "loss": 5.8463,
      "mean_token_accuracy": 0.18672384172677994,
      "num_tokens": 35805321.0,
      "step": 14125
    },
    {
      "entropy": 6.227254199981689,
      "epoch": 1.6306982111944603,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047441109726934345,
      "loss": 5.7736,
      "mean_token_accuracy": 0.19554600715637208,
      "num_tokens": 35817755.0,
      "step": 14130
    },
    {
      "entropy": 6.249270629882813,
      "epoch": 1.6312752452394692,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00047439198167077256,
      "loss": 5.7777,
      "mean_token_accuracy": 0.19097306579351425,
      "num_tokens": 35831856.0,
      "step": 14135
    },
    {
      "entropy": 6.205933904647827,
      "epoch": 1.6318522792844776,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004743728593657514,
      "loss": 5.8502,
      "mean_token_accuracy": 0.19486518800258637,
      "num_tokens": 35844613.0,
      "step": 14140
    },
    {
      "entropy": 6.185780811309814,
      "epoch": 1.6324293133294865,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004743537303549231,
      "loss": 5.7286,
      "mean_token_accuracy": 0.19566426575183868,
      "num_tokens": 35856928.0,
      "step": 14145
    },
    {
      "entropy": 6.2138697624206545,
      "epoch": 1.6330063473744951,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004743345946389311,
      "loss": 5.7688,
      "mean_token_accuracy": 0.18796713054180145,
      "num_tokens": 35870161.0,
      "step": 14150
    },
    {
      "entropy": 6.257949161529541,
      "epoch": 1.6335833814195038,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047431545221841907,
      "loss": 5.803,
      "mean_token_accuracy": 0.18929531574249267,
      "num_tokens": 35883176.0,
      "step": 14155
    },
    {
      "entropy": 6.215028476715088,
      "epoch": 1.6341604154645124,
      "grad_norm": 0.875,
      "learning_rate": 0.00047429630309403086,
      "loss": 5.7156,
      "mean_token_accuracy": 0.19241938591003419,
      "num_tokens": 35895741.0,
      "step": 14160
    },
    {
      "entropy": 6.257434463500976,
      "epoch": 1.634737449509521,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047427714726641044,
      "loss": 5.8106,
      "mean_token_accuracy": 0.19251890629529952,
      "num_tokens": 35908921.0,
      "step": 14165
    },
    {
      "entropy": 6.230150032043457,
      "epoch": 1.6353144835545297,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047425798473620215,
      "loss": 5.7675,
      "mean_token_accuracy": 0.20064556002616882,
      "num_tokens": 35921388.0,
      "step": 14170
    },
    {
      "entropy": 6.260846853256226,
      "epoch": 1.6358915175995383,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004742388155040505,
      "loss": 5.8593,
      "mean_token_accuracy": 0.187110435962677,
      "num_tokens": 35933880.0,
      "step": 14175
    },
    {
      "entropy": 6.205022478103638,
      "epoch": 1.6364685516445472,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047421963957060026,
      "loss": 5.7836,
      "mean_token_accuracy": 0.1910621404647827,
      "num_tokens": 35946690.0,
      "step": 14180
    },
    {
      "entropy": 6.250089120864868,
      "epoch": 1.6370455856895556,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004742004569364964,
      "loss": 5.8158,
      "mean_token_accuracy": 0.19081811755895614,
      "num_tokens": 35958437.0,
      "step": 14185
    },
    {
      "entropy": 6.258803415298462,
      "epoch": 1.6376226197345645,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047418126760238416,
      "loss": 5.7919,
      "mean_token_accuracy": 0.19353751093149185,
      "num_tokens": 35969734.0,
      "step": 14190
    },
    {
      "entropy": 6.18122410774231,
      "epoch": 1.6381996537795729,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047416207156890887,
      "loss": 5.7477,
      "mean_token_accuracy": 0.2015211522579193,
      "num_tokens": 35983059.0,
      "step": 14195
    },
    {
      "entropy": 6.284786176681519,
      "epoch": 1.6387766878245817,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004741428688367164,
      "loss": 5.7663,
      "mean_token_accuracy": 0.19465270787477493,
      "num_tokens": 35996207.0,
      "step": 14200
    },
    {
      "entropy": 6.292109537124634,
      "epoch": 1.6393537218695902,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047412365940645225,
      "loss": 5.8441,
      "mean_token_accuracy": 0.18654603213071824,
      "num_tokens": 36008328.0,
      "step": 14205
    },
    {
      "entropy": 6.252365779876709,
      "epoch": 1.639930755914599,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047410444327876286,
      "loss": 5.8962,
      "mean_token_accuracy": 0.18588175624608994,
      "num_tokens": 36021384.0,
      "step": 14210
    },
    {
      "entropy": 6.1903773784637455,
      "epoch": 1.6405077899596077,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047408522045429426,
      "loss": 5.8045,
      "mean_token_accuracy": 0.19350525587797165,
      "num_tokens": 36034101.0,
      "step": 14215
    },
    {
      "entropy": 6.263929557800293,
      "epoch": 1.6410848240046163,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004740659909336933,
      "loss": 5.7364,
      "mean_token_accuracy": 0.1943316340446472,
      "num_tokens": 36046504.0,
      "step": 14220
    },
    {
      "entropy": 6.196165561676025,
      "epoch": 1.641661858049625,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047404675471760655,
      "loss": 5.8198,
      "mean_token_accuracy": 0.19140298962593078,
      "num_tokens": 36058759.0,
      "step": 14225
    },
    {
      "entropy": 6.200320816040039,
      "epoch": 1.6422388920946336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004740275118066811,
      "loss": 5.7793,
      "mean_token_accuracy": 0.19991466999053956,
      "num_tokens": 36070238.0,
      "step": 14230
    },
    {
      "entropy": 6.241323566436767,
      "epoch": 1.6428159261396422,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047400826220156416,
      "loss": 5.8276,
      "mean_token_accuracy": 0.19229099601507188,
      "num_tokens": 36083543.0,
      "step": 14235
    },
    {
      "entropy": 6.219793796539307,
      "epoch": 1.6433929601846509,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047398900590290313,
      "loss": 5.6807,
      "mean_token_accuracy": 0.20024410188198088,
      "num_tokens": 36095250.0,
      "step": 14240
    },
    {
      "entropy": 6.142616605758667,
      "epoch": 1.6439699942296595,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047396974291134575,
      "loss": 5.8011,
      "mean_token_accuracy": 0.1989526629447937,
      "num_tokens": 36107919.0,
      "step": 14245
    },
    {
      "entropy": 6.286284017562866,
      "epoch": 1.6445470282746681,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047395047322754,
      "loss": 5.8255,
      "mean_token_accuracy": 0.19084158539772034,
      "num_tokens": 36120284.0,
      "step": 14250
    },
    {
      "entropy": 6.306645107269287,
      "epoch": 1.645124062319677,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047393119685213374,
      "loss": 5.859,
      "mean_token_accuracy": 0.19061685651540755,
      "num_tokens": 36133710.0,
      "step": 14255
    },
    {
      "entropy": 6.1821150302886965,
      "epoch": 1.6457010963646854,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004739119137857756,
      "loss": 5.7511,
      "mean_token_accuracy": 0.19963260143995284,
      "num_tokens": 36147042.0,
      "step": 14260
    },
    {
      "entropy": 6.284886217117309,
      "epoch": 1.6462781304096943,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047389262402911404,
      "loss": 5.844,
      "mean_token_accuracy": 0.1859625145792961,
      "num_tokens": 36159053.0,
      "step": 14265
    },
    {
      "entropy": 6.158206939697266,
      "epoch": 1.6468551644547027,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047387332758279784,
      "loss": 5.7184,
      "mean_token_accuracy": 0.19666843265295028,
      "num_tokens": 36171321.0,
      "step": 14270
    },
    {
      "entropy": 6.1448578357696535,
      "epoch": 1.6474321984997116,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047385402444747606,
      "loss": 5.6704,
      "mean_token_accuracy": 0.20474444776773454,
      "num_tokens": 36183341.0,
      "step": 14275
    },
    {
      "entropy": 6.197768259048462,
      "epoch": 1.64800923254472,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047383471462379803,
      "loss": 5.7896,
      "mean_token_accuracy": 0.1886861339211464,
      "num_tokens": 36195407.0,
      "step": 14280
    },
    {
      "entropy": 6.28682508468628,
      "epoch": 1.6485862665897288,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000473815398112413,
      "loss": 5.8836,
      "mean_token_accuracy": 0.18015870302915574,
      "num_tokens": 36208736.0,
      "step": 14285
    },
    {
      "entropy": 6.259079837799073,
      "epoch": 1.6491633006347375,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004737960749139708,
      "loss": 5.799,
      "mean_token_accuracy": 0.1944161206483841,
      "num_tokens": 36220022.0,
      "step": 14290
    },
    {
      "entropy": 6.213712310791015,
      "epoch": 1.6497403346797461,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004737767450291215,
      "loss": 5.6912,
      "mean_token_accuracy": 0.19665347933769226,
      "num_tokens": 36232105.0,
      "step": 14295
    },
    {
      "entropy": 6.231172180175781,
      "epoch": 1.6503173687247548,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047375740845851506,
      "loss": 5.7934,
      "mean_token_accuracy": 0.1925998568534851,
      "num_tokens": 36243892.0,
      "step": 14300
    },
    {
      "entropy": 6.232110834121704,
      "epoch": 1.6508944027697634,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004737380652028019,
      "loss": 5.8781,
      "mean_token_accuracy": 0.18462008237838745,
      "num_tokens": 36255365.0,
      "step": 14305
    },
    {
      "entropy": 6.291271638870239,
      "epoch": 1.651471436814772,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047371871526263263,
      "loss": 5.8498,
      "mean_token_accuracy": 0.1917511910200119,
      "num_tokens": 36268238.0,
      "step": 14310
    },
    {
      "entropy": 6.2812707901000975,
      "epoch": 1.6520484708597807,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004736993586386581,
      "loss": 5.8409,
      "mean_token_accuracy": 0.19064221084117888,
      "num_tokens": 36280294.0,
      "step": 14315
    },
    {
      "entropy": 6.187256526947022,
      "epoch": 1.6526255049047895,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047367999533152934,
      "loss": 5.741,
      "mean_token_accuracy": 0.1949601501226425,
      "num_tokens": 36293159.0,
      "step": 14320
    },
    {
      "entropy": 6.314623737335205,
      "epoch": 1.653202538949798,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047366062534189756,
      "loss": 5.8619,
      "mean_token_accuracy": 0.18096422851085664,
      "num_tokens": 36306492.0,
      "step": 14325
    },
    {
      "entropy": 6.247358179092407,
      "epoch": 1.6537795729948068,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047364124867041446,
      "loss": 5.815,
      "mean_token_accuracy": 0.18566266745328902,
      "num_tokens": 36319020.0,
      "step": 14330
    },
    {
      "entropy": 6.237202596664429,
      "epoch": 1.6543566070398152,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047362186531773156,
      "loss": 5.8568,
      "mean_token_accuracy": 0.1839733526110649,
      "num_tokens": 36332163.0,
      "step": 14335
    },
    {
      "entropy": 6.248868703842163,
      "epoch": 1.654933641084824,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004736024752845008,
      "loss": 5.77,
      "mean_token_accuracy": 0.2030853345990181,
      "num_tokens": 36345880.0,
      "step": 14340
    },
    {
      "entropy": 6.205056476593017,
      "epoch": 1.6555106751298325,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004735830785713746,
      "loss": 5.7786,
      "mean_token_accuracy": 0.19691661596298218,
      "num_tokens": 36358927.0,
      "step": 14345
    },
    {
      "entropy": 6.2739174365997314,
      "epoch": 1.6560877091748414,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004735636751790051,
      "loss": 5.8394,
      "mean_token_accuracy": 0.187654946744442,
      "num_tokens": 36371009.0,
      "step": 14350
    },
    {
      "entropy": 6.291009950637817,
      "epoch": 1.65666474321985,
      "grad_norm": 0.76171875,
      "learning_rate": 0.000473544265108045,
      "loss": 5.8426,
      "mean_token_accuracy": 0.1900906816124916,
      "num_tokens": 36385538.0,
      "step": 14355
    },
    {
      "entropy": 6.183791637420654,
      "epoch": 1.6572417772648587,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047352484835914716,
      "loss": 5.7343,
      "mean_token_accuracy": 0.1970362976193428,
      "num_tokens": 36398265.0,
      "step": 14360
    },
    {
      "entropy": 6.2556334972381595,
      "epoch": 1.6578188113098673,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004735054249329647,
      "loss": 5.7943,
      "mean_token_accuracy": 0.1944418415427208,
      "num_tokens": 36410691.0,
      "step": 14365
    },
    {
      "entropy": 6.248236465454101,
      "epoch": 1.658395845354876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047348599483015087,
      "loss": 5.7271,
      "mean_token_accuracy": 0.19309451282024384,
      "num_tokens": 36423531.0,
      "step": 14370
    },
    {
      "entropy": 6.201855945587158,
      "epoch": 1.6589728793998846,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047346655805135916,
      "loss": 5.813,
      "mean_token_accuracy": 0.19417500793933867,
      "num_tokens": 36437429.0,
      "step": 14375
    },
    {
      "entropy": 6.182714605331421,
      "epoch": 1.6595499134448932,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004734471145972434,
      "loss": 5.7006,
      "mean_token_accuracy": 0.20003162175416947,
      "num_tokens": 36449264.0,
      "step": 14380
    },
    {
      "entropy": 6.233363485336303,
      "epoch": 1.6601269474899019,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047342766446845753,
      "loss": 5.7297,
      "mean_token_accuracy": 0.1986885040998459,
      "num_tokens": 36461132.0,
      "step": 14385
    },
    {
      "entropy": 6.249863719940185,
      "epoch": 1.6607039815349105,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004734082076656557,
      "loss": 5.8411,
      "mean_token_accuracy": 0.1922901079058647,
      "num_tokens": 36474195.0,
      "step": 14390
    },
    {
      "entropy": 6.263658142089843,
      "epoch": 1.6612810155799194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047338874418949235,
      "loss": 5.9184,
      "mean_token_accuracy": 0.1843574196100235,
      "num_tokens": 36486197.0,
      "step": 14395
    },
    {
      "entropy": 6.196670818328857,
      "epoch": 1.6618580496249278,
      "grad_norm": 0.875,
      "learning_rate": 0.00047336927404062213,
      "loss": 5.733,
      "mean_token_accuracy": 0.20222598910331727,
      "num_tokens": 36499215.0,
      "step": 14400
    },
    {
      "entropy": 6.24823317527771,
      "epoch": 1.6624350836699366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047334979721969995,
      "loss": 5.8686,
      "mean_token_accuracy": 0.1838680624961853,
      "num_tokens": 36511872.0,
      "step": 14405
    },
    {
      "entropy": 6.34404444694519,
      "epoch": 1.663012117714945,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004733303137273808,
      "loss": 5.8509,
      "mean_token_accuracy": 0.19573148638010024,
      "num_tokens": 36524313.0,
      "step": 14410
    },
    {
      "entropy": 6.178117370605468,
      "epoch": 1.663589151759954,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047331082356432015,
      "loss": 5.755,
      "mean_token_accuracy": 0.19641987532377242,
      "num_tokens": 36538532.0,
      "step": 14415
    },
    {
      "entropy": 6.253229999542237,
      "epoch": 1.6641661858049623,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004732913267311734,
      "loss": 5.8271,
      "mean_token_accuracy": 0.1890140399336815,
      "num_tokens": 36551117.0,
      "step": 14420
    },
    {
      "entropy": 6.239301013946533,
      "epoch": 1.6647432198499712,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004732718232285964,
      "loss": 5.7015,
      "mean_token_accuracy": 0.19407555013895034,
      "num_tokens": 36563704.0,
      "step": 14425
    },
    {
      "entropy": 6.189207983016968,
      "epoch": 1.6653202538949798,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047325231305724507,
      "loss": 5.8731,
      "mean_token_accuracy": 0.18626768440008162,
      "num_tokens": 36575952.0,
      "step": 14430
    },
    {
      "entropy": 6.204950571060181,
      "epoch": 1.6658972879399885,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004732327962177757,
      "loss": 5.7014,
      "mean_token_accuracy": 0.2006660595536232,
      "num_tokens": 36590181.0,
      "step": 14435
    },
    {
      "entropy": 6.224772262573242,
      "epoch": 1.6664743219849971,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004732132727108447,
      "loss": 5.776,
      "mean_token_accuracy": 0.18696050345897675,
      "num_tokens": 36601986.0,
      "step": 14440
    },
    {
      "entropy": 6.259748411178589,
      "epoch": 1.6670513560300058,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047319374253710874,
      "loss": 5.8285,
      "mean_token_accuracy": 0.18592497557401658,
      "num_tokens": 36614135.0,
      "step": 14445
    },
    {
      "entropy": 6.282826519012451,
      "epoch": 1.6676283900750144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004731742056972247,
      "loss": 5.8721,
      "mean_token_accuracy": 0.19084470272064208,
      "num_tokens": 36625952.0,
      "step": 14450
    },
    {
      "entropy": 6.26439905166626,
      "epoch": 1.668205424120023,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004731546621918497,
      "loss": 5.8384,
      "mean_token_accuracy": 0.19177187830209733,
      "num_tokens": 36638945.0,
      "step": 14455
    },
    {
      "entropy": 6.254889392852784,
      "epoch": 1.668782458165032,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000473135112021641,
      "loss": 5.8523,
      "mean_token_accuracy": 0.1898684322834015,
      "num_tokens": 36652550.0,
      "step": 14460
    },
    {
      "entropy": 6.275535440444946,
      "epoch": 1.6693594922100403,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047311555518725617,
      "loss": 5.7917,
      "mean_token_accuracy": 0.19250949174165727,
      "num_tokens": 36665260.0,
      "step": 14465
    },
    {
      "entropy": 6.229044246673584,
      "epoch": 1.6699365262550492,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047309599168935323,
      "loss": 5.7801,
      "mean_token_accuracy": 0.19588902294635774,
      "num_tokens": 36679275.0,
      "step": 14470
    },
    {
      "entropy": 6.150969314575195,
      "epoch": 1.6705135603000576,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047307642152858993,
      "loss": 5.7722,
      "mean_token_accuracy": 0.19715811014175416,
      "num_tokens": 36691849.0,
      "step": 14475
    },
    {
      "entropy": 6.259490823745727,
      "epoch": 1.6710905943450665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047305684470562453,
      "loss": 5.8163,
      "mean_token_accuracy": 0.192976276576519,
      "num_tokens": 36703618.0,
      "step": 14480
    },
    {
      "entropy": 6.217957353591919,
      "epoch": 1.6716676283900749,
      "grad_norm": 0.875,
      "learning_rate": 0.0004730372612211156,
      "loss": 5.7882,
      "mean_token_accuracy": 0.18733999133110046,
      "num_tokens": 36716332.0,
      "step": 14485
    },
    {
      "entropy": 6.1186987400054935,
      "epoch": 1.6722446624350837,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047301767107572174,
      "loss": 5.6935,
      "mean_token_accuracy": 0.20108278840780258,
      "num_tokens": 36729136.0,
      "step": 14490
    },
    {
      "entropy": 6.19894003868103,
      "epoch": 1.6728216964800924,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004729980742701018,
      "loss": 5.7904,
      "mean_token_accuracy": 0.19673506319522857,
      "num_tokens": 36743061.0,
      "step": 14495
    },
    {
      "entropy": 6.314801359176636,
      "epoch": 1.673398730525101,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004729784708049151,
      "loss": 5.8068,
      "mean_token_accuracy": 0.19911820143461229,
      "num_tokens": 36756151.0,
      "step": 14500
    },
    {
      "entropy": 6.255735111236572,
      "epoch": 1.6739757645701097,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004729588606808209,
      "loss": 5.7589,
      "mean_token_accuracy": 0.19361286610364914,
      "num_tokens": 36768008.0,
      "step": 14505
    },
    {
      "entropy": 6.143254613876342,
      "epoch": 1.6745527986151183,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047293924389847864,
      "loss": 5.6557,
      "mean_token_accuracy": 0.20423528254032136,
      "num_tokens": 36780694.0,
      "step": 14510
    },
    {
      "entropy": 6.1240214824676515,
      "epoch": 1.675129832660127,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004729196204585483,
      "loss": 5.7663,
      "mean_token_accuracy": 0.19527502059936525,
      "num_tokens": 36793040.0,
      "step": 14515
    },
    {
      "entropy": 6.255216217041015,
      "epoch": 1.6757068667051356,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047289999036168983,
      "loss": 5.8305,
      "mean_token_accuracy": 0.19100727438926696,
      "num_tokens": 36806293.0,
      "step": 14520
    },
    {
      "entropy": 6.247564792633057,
      "epoch": 1.6762839007501442,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004728803536085634,
      "loss": 5.7633,
      "mean_token_accuracy": 0.19272203743457794,
      "num_tokens": 36818717.0,
      "step": 14525
    },
    {
      "entropy": 6.2713957786560055,
      "epoch": 1.6768609347951529,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047286071019982974,
      "loss": 5.7538,
      "mean_token_accuracy": 0.19024229496717454,
      "num_tokens": 36832984.0,
      "step": 14530
    },
    {
      "entropy": 6.214433574676514,
      "epoch": 1.6774379688401617,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047284106013614926,
      "loss": 5.7562,
      "mean_token_accuracy": 0.19173718243837357,
      "num_tokens": 36844088.0,
      "step": 14535
    },
    {
      "entropy": 6.2930127620697025,
      "epoch": 1.6780150028851701,
      "grad_norm": 0.875,
      "learning_rate": 0.000472821403418183,
      "loss": 5.8839,
      "mean_token_accuracy": 0.1864546552300453,
      "num_tokens": 36856846.0,
      "step": 14540
    },
    {
      "entropy": 6.249903535842895,
      "epoch": 1.678592036930179,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004728017400465921,
      "loss": 5.6968,
      "mean_token_accuracy": 0.20189250856637955,
      "num_tokens": 36870023.0,
      "step": 14545
    },
    {
      "entropy": 6.234823894500733,
      "epoch": 1.6791690709751874,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047278207002203796,
      "loss": 5.8186,
      "mean_token_accuracy": 0.19094231426715852,
      "num_tokens": 36883298.0,
      "step": 14550
    },
    {
      "entropy": 6.256910514831543,
      "epoch": 1.6797461050201963,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047276239334518216,
      "loss": 5.8089,
      "mean_token_accuracy": 0.19194251149892808,
      "num_tokens": 36894714.0,
      "step": 14555
    },
    {
      "entropy": 6.1421058654785154,
      "epoch": 1.6803231390652047,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047274271001668646,
      "loss": 5.6306,
      "mean_token_accuracy": 0.21167417466640473,
      "num_tokens": 36908332.0,
      "step": 14560
    },
    {
      "entropy": 6.215278387069702,
      "epoch": 1.6809001731102136,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047272302003721286,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1925485461950302,
      "num_tokens": 36921235.0,
      "step": 14565
    },
    {
      "entropy": 6.25142126083374,
      "epoch": 1.6814772071552222,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047270332340742377,
      "loss": 5.846,
      "mean_token_accuracy": 0.1862723633646965,
      "num_tokens": 36934580.0,
      "step": 14570
    },
    {
      "entropy": 6.260641574859619,
      "epoch": 1.6820542412002308,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047268362012798157,
      "loss": 5.8215,
      "mean_token_accuracy": 0.19092659950256347,
      "num_tokens": 36946936.0,
      "step": 14575
    },
    {
      "entropy": 6.132468318939209,
      "epoch": 1.6826312752452395,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004726639101995491,
      "loss": 5.701,
      "mean_token_accuracy": 0.1978867843747139,
      "num_tokens": 36960179.0,
      "step": 14580
    },
    {
      "entropy": 6.304263353347778,
      "epoch": 1.6832083092902481,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047264419362278906,
      "loss": 5.8166,
      "mean_token_accuracy": 0.19583916366100312,
      "num_tokens": 36972739.0,
      "step": 14585
    },
    {
      "entropy": 6.145615005493164,
      "epoch": 1.6837853433352568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047262447039836484,
      "loss": 5.7097,
      "mean_token_accuracy": 0.19732235372066498,
      "num_tokens": 36987314.0,
      "step": 14590
    },
    {
      "entropy": 6.23006911277771,
      "epoch": 1.6843623773802654,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047260474052693963,
      "loss": 5.7962,
      "mean_token_accuracy": 0.1936758741736412,
      "num_tokens": 36999050.0,
      "step": 14595
    },
    {
      "entropy": 6.253016710281372,
      "epoch": 1.684939411425274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047258500400917724,
      "loss": 5.7495,
      "mean_token_accuracy": 0.19344482570886612,
      "num_tokens": 37011905.0,
      "step": 14600
    },
    {
      "entropy": 6.306393814086914,
      "epoch": 1.6855164454702827,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047256526084574137,
      "loss": 5.8317,
      "mean_token_accuracy": 0.19267585277557372,
      "num_tokens": 37023926.0,
      "step": 14605
    },
    {
      "entropy": 6.232395505905151,
      "epoch": 1.6860934795152915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047254551103729597,
      "loss": 5.8272,
      "mean_token_accuracy": 0.18683594912290574,
      "num_tokens": 37037734.0,
      "step": 14610
    },
    {
      "entropy": 6.299908828735352,
      "epoch": 1.6866705135603,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004725257545845055,
      "loss": 5.8603,
      "mean_token_accuracy": 0.19340324103832246,
      "num_tokens": 37049689.0,
      "step": 14615
    },
    {
      "entropy": 6.322448635101319,
      "epoch": 1.6872475476053088,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047250599148803443,
      "loss": 5.8693,
      "mean_token_accuracy": 0.18939805328845977,
      "num_tokens": 37063999.0,
      "step": 14620
    },
    {
      "entropy": 6.137071800231934,
      "epoch": 1.6878245816503172,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047248622174854746,
      "loss": 5.6977,
      "mean_token_accuracy": 0.20105212181806564,
      "num_tokens": 37076855.0,
      "step": 14625
    },
    {
      "entropy": 6.294154930114746,
      "epoch": 1.688401615695326,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004724664453667094,
      "loss": 5.8168,
      "mean_token_accuracy": 0.1961251378059387,
      "num_tokens": 37089235.0,
      "step": 14630
    },
    {
      "entropy": 6.206166553497314,
      "epoch": 1.6889786497403345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004724466623431857,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18961945325136184,
      "num_tokens": 37102011.0,
      "step": 14635
    },
    {
      "entropy": 6.172718667984009,
      "epoch": 1.6895556837853434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004724268726786415,
      "loss": 5.7155,
      "mean_token_accuracy": 0.2011133074760437,
      "num_tokens": 37115246.0,
      "step": 14640
    },
    {
      "entropy": 6.308906745910645,
      "epoch": 1.690132717830352,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004724070763737424,
      "loss": 5.9142,
      "mean_token_accuracy": 0.18521232157945633,
      "num_tokens": 37127705.0,
      "step": 14645
    },
    {
      "entropy": 6.259133911132812,
      "epoch": 1.6907097518753607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004723872734291545,
      "loss": 5.7358,
      "mean_token_accuracy": 0.1956348106265068,
      "num_tokens": 37141136.0,
      "step": 14650
    },
    {
      "entropy": 6.195092678070068,
      "epoch": 1.6912867859203693,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047236746384554364,
      "loss": 5.7482,
      "mean_token_accuracy": 0.19372663646936417,
      "num_tokens": 37154242.0,
      "step": 14655
    },
    {
      "entropy": 6.181477403640747,
      "epoch": 1.691863819965378,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004723476476235762,
      "loss": 5.7876,
      "mean_token_accuracy": 0.19548340886831284,
      "num_tokens": 37166254.0,
      "step": 14660
    },
    {
      "entropy": 6.293699312210083,
      "epoch": 1.6924408540103866,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004723278247639186,
      "loss": 5.825,
      "mean_token_accuracy": 0.19488025903701783,
      "num_tokens": 37179579.0,
      "step": 14665
    },
    {
      "entropy": 6.172354745864868,
      "epoch": 1.6930178880553952,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004723079952672377,
      "loss": 5.7373,
      "mean_token_accuracy": 0.20032483637332915,
      "num_tokens": 37191478.0,
      "step": 14670
    },
    {
      "entropy": 6.1569726943969725,
      "epoch": 1.693594922100404,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047228815913420035,
      "loss": 5.6341,
      "mean_token_accuracy": 0.20368757843971252,
      "num_tokens": 37203785.0,
      "step": 14675
    },
    {
      "entropy": 6.216964149475098,
      "epoch": 1.6941719561454125,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004722683163654738,
      "loss": 5.6538,
      "mean_token_accuracy": 0.21186617612838746,
      "num_tokens": 37216298.0,
      "step": 14680
    },
    {
      "entropy": 6.237495279312133,
      "epoch": 1.6947489901904214,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004722484669617254,
      "loss": 5.8176,
      "mean_token_accuracy": 0.1979591429233551,
      "num_tokens": 37228659.0,
      "step": 14685
    },
    {
      "entropy": 6.2111368656158445,
      "epoch": 1.6953260242354298,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047222861092362277,
      "loss": 5.7739,
      "mean_token_accuracy": 0.19570069313049315,
      "num_tokens": 37240763.0,
      "step": 14690
    },
    {
      "entropy": 6.197583436965942,
      "epoch": 1.6959030582804386,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047220874825183387,
      "loss": 5.7111,
      "mean_token_accuracy": 0.19744271636009217,
      "num_tokens": 37252238.0,
      "step": 14695
    },
    {
      "entropy": 6.2674706935882565,
      "epoch": 1.696480092325447,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047218887894702656,
      "loss": 5.8422,
      "mean_token_accuracy": 0.18785254061222076,
      "num_tokens": 37264475.0,
      "step": 14700
    },
    {
      "entropy": 6.2490815162658695,
      "epoch": 1.697057126370456,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004721690030098693,
      "loss": 5.7649,
      "mean_token_accuracy": 0.1941169485449791,
      "num_tokens": 37277338.0,
      "step": 14705
    },
    {
      "entropy": 6.293146705627441,
      "epoch": 1.6976341604154646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004721491204410305,
      "loss": 5.8133,
      "mean_token_accuracy": 0.19054362624883653,
      "num_tokens": 37289985.0,
      "step": 14710
    },
    {
      "entropy": 6.299290037155151,
      "epoch": 1.6982111944604732,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047212923124117915,
      "loss": 5.8455,
      "mean_token_accuracy": 0.19336792528629304,
      "num_tokens": 37303177.0,
      "step": 14715
    },
    {
      "entropy": 6.249830102920532,
      "epoch": 1.6987882285054818,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004721093354109839,
      "loss": 5.7776,
      "mean_token_accuracy": 0.19173800200223923,
      "num_tokens": 37316662.0,
      "step": 14720
    },
    {
      "entropy": 6.212270212173462,
      "epoch": 1.6993652625504905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047208943295111406,
      "loss": 5.8113,
      "mean_token_accuracy": 0.1935466155409813,
      "num_tokens": 37327782.0,
      "step": 14725
    },
    {
      "entropy": 6.20684404373169,
      "epoch": 1.6999422965954991,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047206952386223905,
      "loss": 5.7118,
      "mean_token_accuracy": 0.20027389824390412,
      "num_tokens": 37340650.0,
      "step": 14730
    },
    {
      "entropy": 6.218963146209717,
      "epoch": 1.7005193306405078,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004720496081450285,
      "loss": 5.8575,
      "mean_token_accuracy": 0.18615148067474366,
      "num_tokens": 37354330.0,
      "step": 14735
    },
    {
      "entropy": 6.254994487762451,
      "epoch": 1.7010963646855164,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047202968580015224,
      "loss": 5.8109,
      "mean_token_accuracy": 0.18774326294660568,
      "num_tokens": 37367447.0,
      "step": 14740
    },
    {
      "entropy": 6.21269965171814,
      "epoch": 1.701673398730525,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047200975682828045,
      "loss": 5.7083,
      "mean_token_accuracy": 0.20425475686788558,
      "num_tokens": 37380171.0,
      "step": 14745
    },
    {
      "entropy": 6.241073322296143,
      "epoch": 1.702250432775534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004719898212300832,
      "loss": 5.7646,
      "mean_token_accuracy": 0.1975268229842186,
      "num_tokens": 37393042.0,
      "step": 14750
    },
    {
      "entropy": 6.260925817489624,
      "epoch": 1.7028274668205423,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047196987900623134,
      "loss": 5.8335,
      "mean_token_accuracy": 0.19406626522541046,
      "num_tokens": 37406854.0,
      "step": 14755
    },
    {
      "entropy": 6.278013086318969,
      "epoch": 1.7034045008655512,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047194993015739527,
      "loss": 5.8416,
      "mean_token_accuracy": 0.18444779068231582,
      "num_tokens": 37419765.0,
      "step": 14760
    },
    {
      "entropy": 6.276658296585083,
      "epoch": 1.7039815349105596,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047192997468424624,
      "loss": 5.7598,
      "mean_token_accuracy": 0.1970341309905052,
      "num_tokens": 37432280.0,
      "step": 14765
    },
    {
      "entropy": 6.287650012969971,
      "epoch": 1.7045585689555685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719100125874553,
      "loss": 5.88,
      "mean_token_accuracy": 0.18567380905151368,
      "num_tokens": 37444315.0,
      "step": 14770
    },
    {
      "entropy": 6.311885118484497,
      "epoch": 1.7051356030005769,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004718900438676939,
      "loss": 5.8598,
      "mean_token_accuracy": 0.18312131017446517,
      "num_tokens": 37457588.0,
      "step": 14775
    },
    {
      "entropy": 6.271833992004394,
      "epoch": 1.7057126370455857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004718700685256337,
      "loss": 5.7828,
      "mean_token_accuracy": 0.188807712495327,
      "num_tokens": 37471740.0,
      "step": 14780
    },
    {
      "entropy": 6.241542768478394,
      "epoch": 1.7062896710905944,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004718500865619465,
      "loss": 5.7624,
      "mean_token_accuracy": 0.19013513773679733,
      "num_tokens": 37484835.0,
      "step": 14785
    },
    {
      "entropy": 6.222132205963135,
      "epoch": 1.706866705135603,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004718300979773044,
      "loss": 5.7822,
      "mean_token_accuracy": 0.19364413172006606,
      "num_tokens": 37497987.0,
      "step": 14790
    },
    {
      "entropy": 6.233245754241944,
      "epoch": 1.7074437391806117,
      "grad_norm": 0.875,
      "learning_rate": 0.00047181010277237977,
      "loss": 5.7965,
      "mean_token_accuracy": 0.19197079092264174,
      "num_tokens": 37510454.0,
      "step": 14795
    },
    {
      "entropy": 6.223056221008301,
      "epoch": 1.7080207732256203,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004717901009478451,
      "loss": 5.7408,
      "mean_token_accuracy": 0.19609536081552506,
      "num_tokens": 37523959.0,
      "step": 14800
    },
    {
      "entropy": 6.237800121307373,
      "epoch": 1.708597807270629,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047177009250437313,
      "loss": 5.8545,
      "mean_token_accuracy": 0.19167944490909578,
      "num_tokens": 37538116.0,
      "step": 14805
    },
    {
      "entropy": 6.248590517044067,
      "epoch": 1.7091748413156376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047175007744263683,
      "loss": 5.7887,
      "mean_token_accuracy": 0.18856527507305146,
      "num_tokens": 37549940.0,
      "step": 14810
    },
    {
      "entropy": 6.235120105743408,
      "epoch": 1.7097518753606464,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047173005576330935,
      "loss": 5.7989,
      "mean_token_accuracy": 0.1889104962348938,
      "num_tokens": 37562854.0,
      "step": 14815
    },
    {
      "entropy": 6.158454179763794,
      "epoch": 1.7103289094056549,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047171002746706424,
      "loss": 5.6782,
      "mean_token_accuracy": 0.19953847974538802,
      "num_tokens": 37575639.0,
      "step": 14820
    },
    {
      "entropy": 6.150306129455567,
      "epoch": 1.7109059434506637,
      "grad_norm": 0.796875,
      "learning_rate": 0.00047168999255457506,
      "loss": 5.7059,
      "mean_token_accuracy": 0.20052077621221542,
      "num_tokens": 37588555.0,
      "step": 14825
    },
    {
      "entropy": 6.270140981674194,
      "epoch": 1.7114829774956721,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047166995102651565,
      "loss": 5.7766,
      "mean_token_accuracy": 0.19444984793663025,
      "num_tokens": 37602636.0,
      "step": 14830
    },
    {
      "entropy": 6.176152229309082,
      "epoch": 1.712060011540681,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004716499028835601,
      "loss": 5.7555,
      "mean_token_accuracy": 0.1888708472251892,
      "num_tokens": 37616769.0,
      "step": 14835
    },
    {
      "entropy": 6.289465618133545,
      "epoch": 1.7126370455856894,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004716298481263828,
      "loss": 5.8139,
      "mean_token_accuracy": 0.19653743058443068,
      "num_tokens": 37628173.0,
      "step": 14840
    },
    {
      "entropy": 6.259516859054566,
      "epoch": 1.7132140796306983,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004716097867556583,
      "loss": 5.7219,
      "mean_token_accuracy": 0.20036691427230835,
      "num_tokens": 37640953.0,
      "step": 14845
    },
    {
      "entropy": 6.275711297988892,
      "epoch": 1.713791113675707,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047158971877206114,
      "loss": 5.784,
      "mean_token_accuracy": 0.194672891497612,
      "num_tokens": 37655091.0,
      "step": 14850
    },
    {
      "entropy": 6.267841482162476,
      "epoch": 1.7143681477207156,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004715696441762665,
      "loss": 5.7952,
      "mean_token_accuracy": 0.1917007863521576,
      "num_tokens": 37668120.0,
      "step": 14855
    },
    {
      "entropy": 6.255063152313232,
      "epoch": 1.7149451817657242,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047154956296894954,
      "loss": 5.6805,
      "mean_token_accuracy": 0.19967543482780456,
      "num_tokens": 37679844.0,
      "step": 14860
    },
    {
      "entropy": 6.151014900207519,
      "epoch": 1.7155222158107328,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004715294751507856,
      "loss": 5.7848,
      "mean_token_accuracy": 0.19969408810138703,
      "num_tokens": 37692756.0,
      "step": 14865
    },
    {
      "entropy": 6.228969764709473,
      "epoch": 1.7160992498557415,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004715093807224505,
      "loss": 5.841,
      "mean_token_accuracy": 0.19111501574516296,
      "num_tokens": 37704525.0,
      "step": 14870
    },
    {
      "entropy": 6.268322706222534,
      "epoch": 1.71667628390075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047148927968462,
      "loss": 5.8243,
      "mean_token_accuracy": 0.19542557150125503,
      "num_tokens": 37715980.0,
      "step": 14875
    },
    {
      "entropy": 6.16850962638855,
      "epoch": 1.7172533179457588,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047146917203797,
      "loss": 5.6816,
      "mean_token_accuracy": 0.19586372673511504,
      "num_tokens": 37728448.0,
      "step": 14880
    },
    {
      "entropy": 6.24118013381958,
      "epoch": 1.7178303519907674,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004714490577831771,
      "loss": 5.735,
      "mean_token_accuracy": 0.19564872086048127,
      "num_tokens": 37741798.0,
      "step": 14885
    },
    {
      "entropy": 6.127769565582275,
      "epoch": 1.7184073860357763,
      "grad_norm": 0.875,
      "learning_rate": 0.0004714289369209177,
      "loss": 5.7005,
      "mean_token_accuracy": 0.20417662113904952,
      "num_tokens": 37755056.0,
      "step": 14890
    },
    {
      "entropy": 6.258713865280152,
      "epoch": 1.7189844200807847,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047140880945186863,
      "loss": 5.8498,
      "mean_token_accuracy": 0.18246588259935378,
      "num_tokens": 37767554.0,
      "step": 14895
    },
    {
      "entropy": 6.265377187728882,
      "epoch": 1.7195614541257935,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047138867537670676,
      "loss": 5.8117,
      "mean_token_accuracy": 0.19259098470211028,
      "num_tokens": 37781430.0,
      "step": 14900
    },
    {
      "entropy": 6.142825984954834,
      "epoch": 1.720138488170802,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047136853469610933,
      "loss": 5.6698,
      "mean_token_accuracy": 0.19435940086841583,
      "num_tokens": 37793138.0,
      "step": 14905
    },
    {
      "entropy": 6.261962604522705,
      "epoch": 1.7207155222158108,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047134838741075383,
      "loss": 5.8484,
      "mean_token_accuracy": 0.18697416484355928,
      "num_tokens": 37806685.0,
      "step": 14910
    },
    {
      "entropy": 6.145495653152466,
      "epoch": 1.7212925562608192,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047132823352131787,
      "loss": 5.6826,
      "mean_token_accuracy": 0.20596079975366594,
      "num_tokens": 37818571.0,
      "step": 14915
    },
    {
      "entropy": 6.245194911956787,
      "epoch": 1.721869590305828,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004713080730284793,
      "loss": 5.8005,
      "mean_token_accuracy": 0.18864455223083496,
      "num_tokens": 37831245.0,
      "step": 14920
    },
    {
      "entropy": 6.28859658241272,
      "epoch": 1.7224466243508367,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047128790593291617,
      "loss": 5.857,
      "mean_token_accuracy": 0.1906914085149765,
      "num_tokens": 37844140.0,
      "step": 14925
    },
    {
      "entropy": 6.262699174880981,
      "epoch": 1.7230236583958454,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047126773223530677,
      "loss": 5.8359,
      "mean_token_accuracy": 0.18864661306142808,
      "num_tokens": 37856783.0,
      "step": 14930
    },
    {
      "entropy": 6.243388843536377,
      "epoch": 1.723600692440854,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047124755193632975,
      "loss": 5.8196,
      "mean_token_accuracy": 0.19158670604228972,
      "num_tokens": 37871219.0,
      "step": 14935
    },
    {
      "entropy": 6.235396671295166,
      "epoch": 1.7241777264858626,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047122736503666377,
      "loss": 5.7965,
      "mean_token_accuracy": 0.20036006718873978,
      "num_tokens": 37882510.0,
      "step": 14940
    },
    {
      "entropy": 6.265579509735107,
      "epoch": 1.7247547605308713,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004712071715369878,
      "loss": 5.8222,
      "mean_token_accuracy": 0.19382917284965515,
      "num_tokens": 37894785.0,
      "step": 14945
    },
    {
      "entropy": 6.183929061889648,
      "epoch": 1.72533179457588,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004711869714379812,
      "loss": 5.7432,
      "mean_token_accuracy": 0.19326967149972915,
      "num_tokens": 37907598.0,
      "step": 14950
    },
    {
      "entropy": 6.19739556312561,
      "epoch": 1.7259088286208888,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004711667647403232,
      "loss": 5.7405,
      "mean_token_accuracy": 0.19208179116249086,
      "num_tokens": 37920399.0,
      "step": 14955
    },
    {
      "entropy": 6.287768983840943,
      "epoch": 1.7264858626658972,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047114655144469354,
      "loss": 5.8563,
      "mean_token_accuracy": 0.191630719602108,
      "num_tokens": 37934157.0,
      "step": 14960
    },
    {
      "entropy": 6.234230899810791,
      "epoch": 1.727062896710906,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047112633155177203,
      "loss": 5.7158,
      "mean_token_accuracy": 0.19391874819993973,
      "num_tokens": 37946058.0,
      "step": 14965
    },
    {
      "entropy": 6.2441102981567385,
      "epoch": 1.7276399307559145,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004711061050622388,
      "loss": 5.8032,
      "mean_token_accuracy": 0.19097260385751724,
      "num_tokens": 37958549.0,
      "step": 14970
    },
    {
      "entropy": 6.2596588134765625,
      "epoch": 1.7282169648009233,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047108587197677404,
      "loss": 5.7802,
      "mean_token_accuracy": 0.1936912551522255,
      "num_tokens": 37971633.0,
      "step": 14975
    },
    {
      "entropy": 6.218993473052978,
      "epoch": 1.7287939988459318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047106563229605845,
      "loss": 5.6964,
      "mean_token_accuracy": 0.1986278250813484,
      "num_tokens": 37984048.0,
      "step": 14980
    },
    {
      "entropy": 6.243938875198364,
      "epoch": 1.7293710328909406,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047104538602077276,
      "loss": 5.8083,
      "mean_token_accuracy": 0.19199582785367966,
      "num_tokens": 37996092.0,
      "step": 14985
    },
    {
      "entropy": 6.165258264541626,
      "epoch": 1.7299480669359493,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004710251331515978,
      "loss": 5.6962,
      "mean_token_accuracy": 0.20166029036045074,
      "num_tokens": 38008769.0,
      "step": 14990
    },
    {
      "entropy": 6.196075010299682,
      "epoch": 1.730525100980958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047100487368921485,
      "loss": 5.7088,
      "mean_token_accuracy": 0.201326784491539,
      "num_tokens": 38021344.0,
      "step": 14995
    },
    {
      "entropy": 6.264497995376587,
      "epoch": 1.7311021350259665,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004709846076343055,
      "loss": 5.7748,
      "mean_token_accuracy": 0.19463672935962678,
      "num_tokens": 38033839.0,
      "step": 15000
    },
    {
      "epoch": 1.7311021350259665,
      "eval_entropy": 6.064682703110384,
      "eval_loss": 5.845510005950928,
      "eval_mean_token_accuracy": 0.19813954393133915,
      "eval_num_tokens": 38033839.0,
      "eval_runtime": 17.7091,
      "eval_samples_per_second": 1588.788,
      "eval_steps_per_second": 198.598,
      "step": 15000
    },
    {
      "entropy": 6.29112286567688,
      "epoch": 1.7316791690709752,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00047096433498755103,
      "loss": 5.8686,
      "mean_token_accuracy": 0.1880417764186859,
      "num_tokens": 38047026.0,
      "step": 15005
    },
    {
      "entropy": 6.211699533462524,
      "epoch": 1.7322562031159838,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047094405574963364,
      "loss": 5.7485,
      "mean_token_accuracy": 0.19679904282093047,
      "num_tokens": 38059089.0,
      "step": 15010
    },
    {
      "entropy": 6.191127967834473,
      "epoch": 1.7328332371609925,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047092376992123516,
      "loss": 5.7486,
      "mean_token_accuracy": 0.1922982946038246,
      "num_tokens": 38072925.0,
      "step": 15015
    },
    {
      "entropy": 6.232037830352783,
      "epoch": 1.733410271206001,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047090347750303803,
      "loss": 5.7665,
      "mean_token_accuracy": 0.19748477041721343,
      "num_tokens": 38084532.0,
      "step": 15020
    },
    {
      "entropy": 6.237213897705078,
      "epoch": 1.7339873052510097,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004708831784957247,
      "loss": 5.7789,
      "mean_token_accuracy": 0.19841754734516143,
      "num_tokens": 38098024.0,
      "step": 15025
    },
    {
      "entropy": 6.216627788543701,
      "epoch": 1.7345643392960186,
      "grad_norm": 1.0,
      "learning_rate": 0.0004708628728999781,
      "loss": 5.7706,
      "mean_token_accuracy": 0.19481493383646012,
      "num_tokens": 38110718.0,
      "step": 15030
    },
    {
      "entropy": 6.1884454727172855,
      "epoch": 1.735141373341027,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004708425607164809,
      "loss": 5.7988,
      "mean_token_accuracy": 0.19548208713531495,
      "num_tokens": 38124807.0,
      "step": 15035
    },
    {
      "entropy": 6.245121669769287,
      "epoch": 1.7357184073860359,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004708222419459165,
      "loss": 5.7852,
      "mean_token_accuracy": 0.18792637288570405,
      "num_tokens": 38136660.0,
      "step": 15040
    },
    {
      "entropy": 6.21821608543396,
      "epoch": 1.7362954414310443,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004708019165889683,
      "loss": 5.8043,
      "mean_token_accuracy": 0.1919826164841652,
      "num_tokens": 38148154.0,
      "step": 15045
    },
    {
      "entropy": 6.218349599838257,
      "epoch": 1.7368724754760532,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004707815846463199,
      "loss": 5.7078,
      "mean_token_accuracy": 0.19730029553174971,
      "num_tokens": 38161092.0,
      "step": 15050
    },
    {
      "entropy": 6.233605480194091,
      "epoch": 1.7374495095210616,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004707612461186552,
      "loss": 5.8269,
      "mean_token_accuracy": 0.19417399764060975,
      "num_tokens": 38172349.0,
      "step": 15055
    },
    {
      "entropy": 6.204860830307007,
      "epoch": 1.7380265435660704,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047074090100665805,
      "loss": 5.8383,
      "mean_token_accuracy": 0.18733077496290207,
      "num_tokens": 38184266.0,
      "step": 15060
    },
    {
      "entropy": 6.124334621429443,
      "epoch": 1.738603577611079,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047072054931101306,
      "loss": 5.6539,
      "mean_token_accuracy": 0.2053797036409378,
      "num_tokens": 38196865.0,
      "step": 15065
    },
    {
      "entropy": 6.2508533000946045,
      "epoch": 1.7391806116560877,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004707001910324046,
      "loss": 5.7363,
      "mean_token_accuracy": 0.19580460339784622,
      "num_tokens": 38208416.0,
      "step": 15070
    },
    {
      "entropy": 6.199764919281006,
      "epoch": 1.7397576457010964,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047067982617151737,
      "loss": 5.7844,
      "mean_token_accuracy": 0.19084140956401824,
      "num_tokens": 38220649.0,
      "step": 15075
    },
    {
      "entropy": 6.234386920928955,
      "epoch": 1.740334679746105,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004706594547290365,
      "loss": 5.7862,
      "mean_token_accuracy": 0.19425734132528305,
      "num_tokens": 38233538.0,
      "step": 15080
    },
    {
      "entropy": 6.191281175613403,
      "epoch": 1.7409117137911136,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047063907670564695,
      "loss": 5.6324,
      "mean_token_accuracy": 0.20612397342920302,
      "num_tokens": 38245603.0,
      "step": 15085
    },
    {
      "entropy": 6.224277448654175,
      "epoch": 1.7414887478361223,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004706186921020342,
      "loss": 5.8384,
      "mean_token_accuracy": 0.18908633291721344,
      "num_tokens": 38258201.0,
      "step": 15090
    },
    {
      "entropy": 6.191832828521728,
      "epoch": 1.7420657818811311,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047059830091888407,
      "loss": 5.6889,
      "mean_token_accuracy": 0.20990267097949983,
      "num_tokens": 38270659.0,
      "step": 15095
    },
    {
      "entropy": 6.259460687637329,
      "epoch": 1.7426428159261396,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004705779031568821,
      "loss": 5.7893,
      "mean_token_accuracy": 0.189946049451828,
      "num_tokens": 38283101.0,
      "step": 15100
    },
    {
      "entropy": 6.256536769866943,
      "epoch": 1.7432198499711484,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047055749881671463,
      "loss": 5.6879,
      "mean_token_accuracy": 0.20892496258020402,
      "num_tokens": 38296855.0,
      "step": 15105
    },
    {
      "entropy": 6.155543470382691,
      "epoch": 1.7437968840161568,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004705370878990677,
      "loss": 5.6618,
      "mean_token_accuracy": 0.20107071697711945,
      "num_tokens": 38308354.0,
      "step": 15110
    },
    {
      "entropy": 6.210873031616211,
      "epoch": 1.7443739180611657,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047051667040462806,
      "loss": 5.8233,
      "mean_token_accuracy": 0.18743733167648316,
      "num_tokens": 38321239.0,
      "step": 15115
    },
    {
      "entropy": 6.311090040206909,
      "epoch": 1.7449509521061741,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047049624633408224,
      "loss": 5.8469,
      "mean_token_accuracy": 0.18604665100574494,
      "num_tokens": 38333672.0,
      "step": 15120
    },
    {
      "entropy": 6.221187925338745,
      "epoch": 1.745527986151183,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047047581568811724,
      "loss": 5.741,
      "mean_token_accuracy": 0.19838133305311204,
      "num_tokens": 38346809.0,
      "step": 15125
    },
    {
      "entropy": 6.241187000274659,
      "epoch": 1.7461050201961916,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047045537846742043,
      "loss": 5.7785,
      "mean_token_accuracy": 0.18868411779403688,
      "num_tokens": 38359129.0,
      "step": 15130
    },
    {
      "entropy": 6.305779933929443,
      "epoch": 1.7466820542412003,
      "grad_norm": 0.9375,
      "learning_rate": 0.000470434934672679,
      "loss": 5.8807,
      "mean_token_accuracy": 0.1896311268210411,
      "num_tokens": 38371621.0,
      "step": 15135
    },
    {
      "entropy": 6.273411130905151,
      "epoch": 1.747259088286209,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047041448430458054,
      "loss": 5.8061,
      "mean_token_accuracy": 0.19316441267728807,
      "num_tokens": 38384579.0,
      "step": 15140
    },
    {
      "entropy": 6.329048538208008,
      "epoch": 1.7478361223312175,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047039402736381305,
      "loss": 5.8802,
      "mean_token_accuracy": 0.18634075820446014,
      "num_tokens": 38396765.0,
      "step": 15145
    },
    {
      "entropy": 6.227564573287964,
      "epoch": 1.7484131563762262,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004703735638510645,
      "loss": 5.763,
      "mean_token_accuracy": 0.19446907788515091,
      "num_tokens": 38409778.0,
      "step": 15150
    },
    {
      "entropy": 6.242182016372681,
      "epoch": 1.7489901904212348,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004703530937670232,
      "loss": 5.7848,
      "mean_token_accuracy": 0.19276127964258194,
      "num_tokens": 38421729.0,
      "step": 15155
    },
    {
      "entropy": 6.267483186721802,
      "epoch": 1.7495672244662435,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004703326171123776,
      "loss": 5.7312,
      "mean_token_accuracy": 0.1934648722410202,
      "num_tokens": 38435136.0,
      "step": 15160
    },
    {
      "entropy": 6.279381942749024,
      "epoch": 1.750144258511252,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004703121338878164,
      "loss": 5.7818,
      "mean_token_accuracy": 0.1895918920636177,
      "num_tokens": 38447362.0,
      "step": 15165
    },
    {
      "entropy": 6.18130578994751,
      "epoch": 1.750721292556261,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004702916440940286,
      "loss": 5.6628,
      "mean_token_accuracy": 0.2025582104921341,
      "num_tokens": 38459796.0,
      "step": 15170
    },
    {
      "entropy": 6.320867919921875,
      "epoch": 1.7512983266012694,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004702711477317034,
      "loss": 5.9161,
      "mean_token_accuracy": 0.18067701905965805,
      "num_tokens": 38472536.0,
      "step": 15175
    },
    {
      "entropy": 6.281063508987427,
      "epoch": 1.7518753606462782,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004702506448015301,
      "loss": 5.7954,
      "mean_token_accuracy": 0.1926846131682396,
      "num_tokens": 38485107.0,
      "step": 15180
    },
    {
      "entropy": 6.268717288970947,
      "epoch": 1.7524523946912867,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047023013530419843,
      "loss": 5.8607,
      "mean_token_accuracy": 0.1821786180138588,
      "num_tokens": 38498560.0,
      "step": 15185
    },
    {
      "entropy": 6.186358451843262,
      "epoch": 1.7530294287362955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004702096192403981,
      "loss": 5.5788,
      "mean_token_accuracy": 0.2046142637729645,
      "num_tokens": 38510553.0,
      "step": 15190
    },
    {
      "entropy": 6.1883259296417235,
      "epoch": 1.753606462781304,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004701890966108192,
      "loss": 5.7672,
      "mean_token_accuracy": 0.1968217372894287,
      "num_tokens": 38524191.0,
      "step": 15195
    },
    {
      "entropy": 6.243796110153198,
      "epoch": 1.7541834968263128,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000470168567416152,
      "loss": 5.8068,
      "mean_token_accuracy": 0.19425098448991776,
      "num_tokens": 38535654.0,
      "step": 15200
    },
    {
      "entropy": 6.3139420509338375,
      "epoch": 1.7547605308713214,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004701480316570869,
      "loss": 5.8602,
      "mean_token_accuracy": 0.18602106273174285,
      "num_tokens": 38548596.0,
      "step": 15205
    },
    {
      "entropy": 6.1814368724822994,
      "epoch": 1.75533756491633,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004701274893343147,
      "loss": 5.7583,
      "mean_token_accuracy": 0.1969182476401329,
      "num_tokens": 38559852.0,
      "step": 15210
    },
    {
      "entropy": 6.126230001449585,
      "epoch": 1.7559145989613387,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047010694044852643,
      "loss": 5.7338,
      "mean_token_accuracy": 0.1970731124281883,
      "num_tokens": 38571773.0,
      "step": 15215
    },
    {
      "entropy": 6.29340877532959,
      "epoch": 1.7564916330063474,
      "grad_norm": 0.921875,
      "learning_rate": 0.000470086385000413,
      "loss": 5.7394,
      "mean_token_accuracy": 0.20051656514406205,
      "num_tokens": 38583898.0,
      "step": 15220
    },
    {
      "entropy": 6.236847352981568,
      "epoch": 1.757068667051356,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004700658229906661,
      "loss": 5.8017,
      "mean_token_accuracy": 0.19162568002939223,
      "num_tokens": 38596573.0,
      "step": 15225
    },
    {
      "entropy": 6.249780750274658,
      "epoch": 1.7576457010963646,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047004525441997694,
      "loss": 5.8146,
      "mean_token_accuracy": 0.19034143835306166,
      "num_tokens": 38610083.0,
      "step": 15230
    },
    {
      "entropy": 6.272627973556519,
      "epoch": 1.7582227351413735,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004700246792890376,
      "loss": 5.7902,
      "mean_token_accuracy": 0.191970856487751,
      "num_tokens": 38622710.0,
      "step": 15235
    },
    {
      "entropy": 6.223754262924194,
      "epoch": 1.758799769186382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004700040975985401,
      "loss": 5.7051,
      "mean_token_accuracy": 0.19880941957235337,
      "num_tokens": 38634633.0,
      "step": 15240
    },
    {
      "entropy": 6.242884159088135,
      "epoch": 1.7593768032313908,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046998350934917654,
      "loss": 5.8169,
      "mean_token_accuracy": 0.19429637044668197,
      "num_tokens": 38646836.0,
      "step": 15245
    },
    {
      "entropy": 6.238810396194458,
      "epoch": 1.7599538372763992,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046996291454163956,
      "loss": 5.7553,
      "mean_token_accuracy": 0.19605858772993087,
      "num_tokens": 38659251.0,
      "step": 15250
    },
    {
      "entropy": 6.263228511810302,
      "epoch": 1.760530871321408,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004699423131766218,
      "loss": 5.7813,
      "mean_token_accuracy": 0.19510549753904344,
      "num_tokens": 38672464.0,
      "step": 15255
    },
    {
      "entropy": 6.222991991043091,
      "epoch": 1.7611079053664165,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004699217052548161,
      "loss": 5.793,
      "mean_token_accuracy": 0.19427470564842225,
      "num_tokens": 38684989.0,
      "step": 15260
    },
    {
      "entropy": 6.271829605102539,
      "epoch": 1.7616849394114253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046990109077691577,
      "loss": 5.8337,
      "mean_token_accuracy": 0.19096603989601135,
      "num_tokens": 38698528.0,
      "step": 15265
    },
    {
      "entropy": 6.2350969314575195,
      "epoch": 1.762261973456434,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046988046974361406,
      "loss": 5.742,
      "mean_token_accuracy": 0.19062920957803725,
      "num_tokens": 38710525.0,
      "step": 15270
    },
    {
      "entropy": 6.231028127670288,
      "epoch": 1.7628390075014426,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004698598421556045,
      "loss": 5.7666,
      "mean_token_accuracy": 0.19077669978141784,
      "num_tokens": 38723668.0,
      "step": 15275
    },
    {
      "entropy": 6.286417007446289,
      "epoch": 1.7634160415464513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004698392080135809,
      "loss": 5.8025,
      "mean_token_accuracy": 0.1893696218729019,
      "num_tokens": 38735136.0,
      "step": 15280
    },
    {
      "entropy": 6.2255021095275875,
      "epoch": 1.76399307559146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004698185673182374,
      "loss": 5.7849,
      "mean_token_accuracy": 0.1927477926015854,
      "num_tokens": 38748118.0,
      "step": 15285
    },
    {
      "entropy": 6.163179683685303,
      "epoch": 1.7645701096364685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046979792007026817,
      "loss": 5.7293,
      "mean_token_accuracy": 0.19634648859500886,
      "num_tokens": 38759685.0,
      "step": 15290
    },
    {
      "entropy": 6.204593086242676,
      "epoch": 1.7651471436814772,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004697772662703676,
      "loss": 5.6773,
      "mean_token_accuracy": 0.20042047500610352,
      "num_tokens": 38772541.0,
      "step": 15295
    },
    {
      "entropy": 6.290904188156128,
      "epoch": 1.7657241777264858,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046975660591923047,
      "loss": 5.827,
      "mean_token_accuracy": 0.18909635841846467,
      "num_tokens": 38785319.0,
      "step": 15300
    },
    {
      "entropy": 6.208968019485473,
      "epoch": 1.7663012117714945,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004697359390175516,
      "loss": 5.7904,
      "mean_token_accuracy": 0.18779968321323395,
      "num_tokens": 38797979.0,
      "step": 15305
    },
    {
      "entropy": 6.164976501464844,
      "epoch": 1.7668782458165033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046971526556602625,
      "loss": 5.6631,
      "mean_token_accuracy": 0.20508671551942825,
      "num_tokens": 38810797.0,
      "step": 15310
    },
    {
      "entropy": 6.192159700393677,
      "epoch": 1.7674552798615117,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004696945855653495,
      "loss": 5.7544,
      "mean_token_accuracy": 0.19130376875400543,
      "num_tokens": 38822409.0,
      "step": 15315
    },
    {
      "entropy": 6.285590791702271,
      "epoch": 1.7680323139065206,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004696738990162172,
      "loss": 5.833,
      "mean_token_accuracy": 0.18697229623794556,
      "num_tokens": 38835952.0,
      "step": 15320
    },
    {
      "entropy": 6.291071033477783,
      "epoch": 1.768609347951529,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000469653205919325,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1975365847349167,
      "num_tokens": 38848778.0,
      "step": 15325
    },
    {
      "entropy": 6.162906837463379,
      "epoch": 1.7691863819965379,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046963250627536884,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1997571513056755,
      "num_tokens": 38860885.0,
      "step": 15330
    },
    {
      "entropy": 6.224546575546265,
      "epoch": 1.7697634160415463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004696118000850451,
      "loss": 5.7075,
      "mean_token_accuracy": 0.1974983334541321,
      "num_tokens": 38873939.0,
      "step": 15335
    },
    {
      "entropy": 6.3215491771698,
      "epoch": 1.7703404500865552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046959108734905003,
      "loss": 5.8154,
      "mean_token_accuracy": 0.18865651190280913,
      "num_tokens": 38887648.0,
      "step": 15340
    },
    {
      "entropy": 6.265390920639038,
      "epoch": 1.7709174841315638,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046957036806808045,
      "loss": 5.8561,
      "mean_token_accuracy": 0.18631444573402406,
      "num_tokens": 38899908.0,
      "step": 15345
    },
    {
      "entropy": 6.327470111846924,
      "epoch": 1.7714945181765724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004695496422428332,
      "loss": 5.8563,
      "mean_token_accuracy": 0.18313935697078704,
      "num_tokens": 38910918.0,
      "step": 15350
    },
    {
      "entropy": 6.2637574672698975,
      "epoch": 1.772071552221581,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004695289098740054,
      "loss": 5.8142,
      "mean_token_accuracy": 0.19002840518951417,
      "num_tokens": 38923427.0,
      "step": 15355
    },
    {
      "entropy": 6.162798023223877,
      "epoch": 1.7726485862665897,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004695081709622943,
      "loss": 5.7252,
      "mean_token_accuracy": 0.20075140595436097,
      "num_tokens": 38935308.0,
      "step": 15360
    },
    {
      "entropy": 6.314521074295044,
      "epoch": 1.7732256203115984,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046948742550839744,
      "loss": 5.8748,
      "mean_token_accuracy": 0.19232352823019028,
      "num_tokens": 38948444.0,
      "step": 15365
    },
    {
      "entropy": 6.251675796508789,
      "epoch": 1.773802654356607,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004694666735130127,
      "loss": 5.7561,
      "mean_token_accuracy": 0.19015007466077805,
      "num_tokens": 38959888.0,
      "step": 15370
    },
    {
      "entropy": 6.112936019897461,
      "epoch": 1.7743796884016156,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000469445914976838,
      "loss": 5.6794,
      "mean_token_accuracy": 0.2034168228507042,
      "num_tokens": 38971769.0,
      "step": 15375
    },
    {
      "entropy": 6.232662105560303,
      "epoch": 1.7749567224466243,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004694251499005715,
      "loss": 5.7328,
      "mean_token_accuracy": 0.19551930129528045,
      "num_tokens": 38984718.0,
      "step": 15380
    },
    {
      "entropy": 6.236335563659668,
      "epoch": 1.7755337564916331,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004694043782849116,
      "loss": 5.766,
      "mean_token_accuracy": 0.19136644005775452,
      "num_tokens": 38996670.0,
      "step": 15385
    },
    {
      "entropy": 6.195712375640869,
      "epoch": 1.7761107905366416,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046938360013055715,
      "loss": 5.7279,
      "mean_token_accuracy": 0.20199797600507735,
      "num_tokens": 39009270.0,
      "step": 15390
    },
    {
      "entropy": 6.277393770217896,
      "epoch": 1.7766878245816504,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00046936281543820673,
      "loss": 5.8051,
      "mean_token_accuracy": 0.18696814477443696,
      "num_tokens": 39022332.0,
      "step": 15395
    },
    {
      "entropy": 6.2402856826782225,
      "epoch": 1.7772648586266588,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046934202420855967,
      "loss": 5.7086,
      "mean_token_accuracy": 0.19538037180900575,
      "num_tokens": 39034964.0,
      "step": 15400
    },
    {
      "entropy": 6.219359350204468,
      "epoch": 1.7778418926716677,
      "grad_norm": 3.75,
      "learning_rate": 0.00046932122644231507,
      "loss": 5.791,
      "mean_token_accuracy": 0.19419807344675064,
      "num_tokens": 39048705.0,
      "step": 15405
    },
    {
      "entropy": 6.235350275039673,
      "epoch": 1.7784189267166761,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046930042214017256,
      "loss": 5.8028,
      "mean_token_accuracy": 0.19112778902053834,
      "num_tokens": 39061994.0,
      "step": 15410
    },
    {
      "entropy": 6.27709002494812,
      "epoch": 1.778995960761685,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004692796113028319,
      "loss": 5.7709,
      "mean_token_accuracy": 0.19875227361917497,
      "num_tokens": 39075279.0,
      "step": 15415
    },
    {
      "entropy": 6.295723247528076,
      "epoch": 1.7795729948066936,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000469258793930993,
      "loss": 5.807,
      "mean_token_accuracy": 0.19260800778865814,
      "num_tokens": 39087914.0,
      "step": 15420
    },
    {
      "entropy": 6.251241636276245,
      "epoch": 1.7801500288517023,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046923797002535605,
      "loss": 5.7558,
      "mean_token_accuracy": 0.20084340125322342,
      "num_tokens": 39100302.0,
      "step": 15425
    },
    {
      "entropy": 6.180109071731567,
      "epoch": 1.780727062896711,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004692171395866214,
      "loss": 5.7828,
      "mean_token_accuracy": 0.19683018028736116,
      "num_tokens": 39114073.0,
      "step": 15430
    },
    {
      "entropy": 6.222915935516357,
      "epoch": 1.7813040969417195,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046919630261548986,
      "loss": 5.7866,
      "mean_token_accuracy": 0.19572454541921616,
      "num_tokens": 39126213.0,
      "step": 15435
    },
    {
      "entropy": 6.256944942474365,
      "epoch": 1.7818811309867282,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046917545911266207,
      "loss": 5.7819,
      "mean_token_accuracy": 0.1914262130856514,
      "num_tokens": 39138600.0,
      "step": 15440
    },
    {
      "entropy": 6.267107343673706,
      "epoch": 1.7824581650317368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046915460907883923,
      "loss": 5.7616,
      "mean_token_accuracy": 0.1944403275847435,
      "num_tokens": 39149356.0,
      "step": 15445
    },
    {
      "entropy": 6.229015064239502,
      "epoch": 1.7830351990767457,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046913375251472246,
      "loss": 5.7307,
      "mean_token_accuracy": 0.19398368299007415,
      "num_tokens": 39162141.0,
      "step": 15450
    },
    {
      "entropy": 6.247556972503662,
      "epoch": 1.783612233121754,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046911288942101345,
      "loss": 5.7947,
      "mean_token_accuracy": 0.18928243815898896,
      "num_tokens": 39173714.0,
      "step": 15455
    },
    {
      "entropy": 6.236412000656128,
      "epoch": 1.784189267166763,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004690920197984138,
      "loss": 5.6856,
      "mean_token_accuracy": 0.19748856127262115,
      "num_tokens": 39186498.0,
      "step": 15460
    },
    {
      "entropy": 6.233781194686889,
      "epoch": 1.7847663012117714,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004690711436476254,
      "loss": 5.7696,
      "mean_token_accuracy": 0.19932501167058944,
      "num_tokens": 39199762.0,
      "step": 15465
    },
    {
      "entropy": 6.135803461074829,
      "epoch": 1.7853433352567802,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046905026096935056,
      "loss": 5.7229,
      "mean_token_accuracy": 0.1936773106455803,
      "num_tokens": 39213236.0,
      "step": 15470
    },
    {
      "entropy": 6.3089663028717045,
      "epoch": 1.7859203693017887,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004690293717642916,
      "loss": 5.8107,
      "mean_token_accuracy": 0.18709516376256943,
      "num_tokens": 39225538.0,
      "step": 15475
    },
    {
      "entropy": 6.28318338394165,
      "epoch": 1.7864974033467975,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000469008476033151,
      "loss": 5.8236,
      "mean_token_accuracy": 0.19107496291399,
      "num_tokens": 39237889.0,
      "step": 15480
    },
    {
      "entropy": 6.1974225521087645,
      "epoch": 1.7870744373918062,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046898757377663176,
      "loss": 5.8032,
      "mean_token_accuracy": 0.19034328907728196,
      "num_tokens": 39250686.0,
      "step": 15485
    },
    {
      "entropy": 6.204254055023194,
      "epoch": 1.7876514714368148,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046896666499543683,
      "loss": 5.7138,
      "mean_token_accuracy": 0.19629829227924347,
      "num_tokens": 39262708.0,
      "step": 15490
    },
    {
      "entropy": 6.068603038787842,
      "epoch": 1.7882285054818234,
      "grad_norm": 1.25,
      "learning_rate": 0.00046894574969026946,
      "loss": 5.5842,
      "mean_token_accuracy": 0.2097725123167038,
      "num_tokens": 39274778.0,
      "step": 15495
    },
    {
      "entropy": 6.245699262619018,
      "epoch": 1.788805539526832,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046892482786183313,
      "loss": 5.816,
      "mean_token_accuracy": 0.18913054168224336,
      "num_tokens": 39288088.0,
      "step": 15500
    },
    {
      "entropy": 6.196365261077881,
      "epoch": 1.7893825735718407,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046890389951083155,
      "loss": 5.6741,
      "mean_token_accuracy": 0.20917547345161439,
      "num_tokens": 39300828.0,
      "step": 15505
    },
    {
      "entropy": 6.2137336254119875,
      "epoch": 1.7899596076168494,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046888296463796857,
      "loss": 5.7199,
      "mean_token_accuracy": 0.2002682641148567,
      "num_tokens": 39313839.0,
      "step": 15510
    },
    {
      "entropy": 6.159156751632691,
      "epoch": 1.790536641661858,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004688620232439485,
      "loss": 5.7999,
      "mean_token_accuracy": 0.195141963660717,
      "num_tokens": 39326115.0,
      "step": 15515
    },
    {
      "entropy": 6.349920082092285,
      "epoch": 1.7911136757068666,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046884107532947547,
      "loss": 5.7657,
      "mean_token_accuracy": 0.19419683665037155,
      "num_tokens": 39338756.0,
      "step": 15520
    },
    {
      "entropy": 6.318724298477173,
      "epoch": 1.7916907097518755,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046882012089525415,
      "loss": 5.8236,
      "mean_token_accuracy": 0.1831202358007431,
      "num_tokens": 39352406.0,
      "step": 15525
    },
    {
      "entropy": 6.212570571899414,
      "epoch": 1.792267743796884,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004687991599419895,
      "loss": 5.7911,
      "mean_token_accuracy": 0.18902456015348434,
      "num_tokens": 39366532.0,
      "step": 15530
    },
    {
      "entropy": 6.268385744094848,
      "epoch": 1.7928447778418928,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046877819247038624,
      "loss": 5.7982,
      "mean_token_accuracy": 0.19253941476345063,
      "num_tokens": 39378730.0,
      "step": 15535
    },
    {
      "entropy": 6.310798978805542,
      "epoch": 1.7934218118869012,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004687572184811497,
      "loss": 5.8285,
      "mean_token_accuracy": 0.18768298327922822,
      "num_tokens": 39391448.0,
      "step": 15540
    },
    {
      "entropy": 6.230965328216553,
      "epoch": 1.79399884593191,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046873623797498545,
      "loss": 5.7513,
      "mean_token_accuracy": 0.20275395810604097,
      "num_tokens": 39405120.0,
      "step": 15545
    },
    {
      "entropy": 6.277521085739136,
      "epoch": 1.7945758799769185,
      "grad_norm": 0.921875,
      "learning_rate": 0.000468715250952599,
      "loss": 5.7963,
      "mean_token_accuracy": 0.1894914209842682,
      "num_tokens": 39417508.0,
      "step": 15550
    },
    {
      "entropy": 6.263333702087403,
      "epoch": 1.7951529140219273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046869425741469635,
      "loss": 5.7715,
      "mean_token_accuracy": 0.19220728129148484,
      "num_tokens": 39430185.0,
      "step": 15555
    },
    {
      "entropy": 6.267724323272705,
      "epoch": 1.795729948066936,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004686732573619835,
      "loss": 5.8464,
      "mean_token_accuracy": 0.20052760988473892,
      "num_tokens": 39442855.0,
      "step": 15560
    },
    {
      "entropy": 6.198214960098267,
      "epoch": 1.7963069821119446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004686522507951669,
      "loss": 5.7302,
      "mean_token_accuracy": 0.19882250130176543,
      "num_tokens": 39454855.0,
      "step": 15565
    },
    {
      "entropy": 6.252347660064697,
      "epoch": 1.7968840161569533,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004686312377149531,
      "loss": 5.7657,
      "mean_token_accuracy": 0.18927662819623947,
      "num_tokens": 39466905.0,
      "step": 15570
    },
    {
      "entropy": 6.263672018051148,
      "epoch": 1.797461050201962,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046861021812204874,
      "loss": 5.756,
      "mean_token_accuracy": 0.1955069601535797,
      "num_tokens": 39479907.0,
      "step": 15575
    },
    {
      "entropy": 6.179773950576783,
      "epoch": 1.7980380842469705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046858919201716085,
      "loss": 5.6914,
      "mean_token_accuracy": 0.19704041182994841,
      "num_tokens": 39491934.0,
      "step": 15580
    },
    {
      "entropy": 6.263770055770874,
      "epoch": 1.7986151182919792,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685681594009966,
      "loss": 5.8154,
      "mean_token_accuracy": 0.19629359096288682,
      "num_tokens": 39503669.0,
      "step": 15585
    },
    {
      "entropy": 6.301582479476929,
      "epoch": 1.799192152336988,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046854712027426357,
      "loss": 5.804,
      "mean_token_accuracy": 0.19538252502679826,
      "num_tokens": 39515555.0,
      "step": 15590
    },
    {
      "entropy": 6.2045458316802975,
      "epoch": 1.7997691863819965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046852607463766923,
      "loss": 5.7162,
      "mean_token_accuracy": 0.19864823520183564,
      "num_tokens": 39528867.0,
      "step": 15595
    },
    {
      "entropy": 6.250128316879272,
      "epoch": 1.8003462204270053,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685050224919215,
      "loss": 5.8093,
      "mean_token_accuracy": 0.1922564446926117,
      "num_tokens": 39541768.0,
      "step": 15600
    },
    {
      "entropy": 6.267262411117554,
      "epoch": 1.8009232544720137,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00046848396383772844,
      "loss": 5.7726,
      "mean_token_accuracy": 0.19399042725563048,
      "num_tokens": 39553578.0,
      "step": 15605
    },
    {
      "entropy": 6.2750020027160645,
      "epoch": 1.8015002885170226,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004684628986757984,
      "loss": 5.7283,
      "mean_token_accuracy": 0.19758292138576508,
      "num_tokens": 39565846.0,
      "step": 15610
    },
    {
      "entropy": 6.274501514434815,
      "epoch": 1.802077322562031,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004684418270068398,
      "loss": 5.7868,
      "mean_token_accuracy": 0.19587256163358688,
      "num_tokens": 39578536.0,
      "step": 15615
    },
    {
      "entropy": 6.217433214187622,
      "epoch": 1.8026543566070399,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004684207488315615,
      "loss": 5.7429,
      "mean_token_accuracy": 0.19850727617740632,
      "num_tokens": 39591233.0,
      "step": 15620
    },
    {
      "entropy": 6.245366144180298,
      "epoch": 1.8032313906520485,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004683996641506724,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1955167680978775,
      "num_tokens": 39602335.0,
      "step": 15625
    },
    {
      "entropy": 6.135843896865845,
      "epoch": 1.8038084246970572,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046837857296488163,
      "loss": 5.6746,
      "mean_token_accuracy": 0.19608980268239976,
      "num_tokens": 39614278.0,
      "step": 15630
    },
    {
      "entropy": 6.186823081970215,
      "epoch": 1.8043854587420658,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046835747527489857,
      "loss": 5.7182,
      "mean_token_accuracy": 0.20170872658491135,
      "num_tokens": 39627218.0,
      "step": 15635
    },
    {
      "entropy": 6.191946983337402,
      "epoch": 1.8049624927870744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000468336371081433,
      "loss": 5.7976,
      "mean_token_accuracy": 0.19195626229047774,
      "num_tokens": 39639607.0,
      "step": 15640
    },
    {
      "entropy": 6.235162591934204,
      "epoch": 1.805539526832083,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046831526038519444,
      "loss": 5.7604,
      "mean_token_accuracy": 0.19048485308885574,
      "num_tokens": 39652748.0,
      "step": 15645
    },
    {
      "entropy": 6.2385009765625,
      "epoch": 1.8061165608770917,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004682941431868933,
      "loss": 5.7316,
      "mean_token_accuracy": 0.1982282429933548,
      "num_tokens": 39665984.0,
      "step": 15650
    },
    {
      "entropy": 6.2360090732574465,
      "epoch": 1.8066935949221004,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00046827301948723963,
      "loss": 5.7195,
      "mean_token_accuracy": 0.19947951138019562,
      "num_tokens": 39680017.0,
      "step": 15655
    },
    {
      "entropy": 6.215149974822998,
      "epoch": 1.807270628967109,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046825188928694393,
      "loss": 5.7431,
      "mean_token_accuracy": 0.1935771480202675,
      "num_tokens": 39694070.0,
      "step": 15660
    },
    {
      "entropy": 6.221736335754395,
      "epoch": 1.8078476630121179,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004682307525867169,
      "loss": 5.7418,
      "mean_token_accuracy": 0.18687294870615007,
      "num_tokens": 39706341.0,
      "step": 15665
    },
    {
      "entropy": 6.200480890274048,
      "epoch": 1.8084246970571263,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004682096093872695,
      "loss": 5.8245,
      "mean_token_accuracy": 0.20415562093257905,
      "num_tokens": 39719552.0,
      "step": 15670
    },
    {
      "entropy": 6.345661163330078,
      "epoch": 1.8090017311021351,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046818845968931284,
      "loss": 5.8691,
      "mean_token_accuracy": 0.1814697802066803,
      "num_tokens": 39732424.0,
      "step": 15675
    },
    {
      "entropy": 6.2327552318573,
      "epoch": 1.8095787651471436,
      "grad_norm": 0.875,
      "learning_rate": 0.00046816730349355843,
      "loss": 5.7719,
      "mean_token_accuracy": 0.19547220766544343,
      "num_tokens": 39745088.0,
      "step": 15680
    },
    {
      "entropy": 6.253853464126587,
      "epoch": 1.8101557991921524,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046814614080071756,
      "loss": 5.7446,
      "mean_token_accuracy": 0.20291758924722672,
      "num_tokens": 39756912.0,
      "step": 15685
    },
    {
      "entropy": 6.252977085113526,
      "epoch": 1.8107328332371608,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046812497161150224,
      "loss": 5.7483,
      "mean_token_accuracy": 0.18975120335817336,
      "num_tokens": 39768224.0,
      "step": 15690
    },
    {
      "entropy": 6.190802097320557,
      "epoch": 1.8113098672821697,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046810379592662445,
      "loss": 5.7399,
      "mean_token_accuracy": 0.19100562632083892,
      "num_tokens": 39780486.0,
      "step": 15695
    },
    {
      "entropy": 6.300315713882446,
      "epoch": 1.8118869013271783,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046808261374679637,
      "loss": 5.7371,
      "mean_token_accuracy": 0.19604469388723372,
      "num_tokens": 39793572.0,
      "step": 15700
    },
    {
      "entropy": 6.229718160629273,
      "epoch": 1.812463935372187,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004680614250727305,
      "loss": 5.7477,
      "mean_token_accuracy": 0.20104105174541473,
      "num_tokens": 39806064.0,
      "step": 15705
    },
    {
      "entropy": 6.15344820022583,
      "epoch": 1.8130409694171956,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004680402299051395,
      "loss": 5.7301,
      "mean_token_accuracy": 0.19799976497888566,
      "num_tokens": 39818638.0,
      "step": 15710
    },
    {
      "entropy": 6.271948337554932,
      "epoch": 1.8136180034622043,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004680190282447363,
      "loss": 5.8174,
      "mean_token_accuracy": 0.1912863403558731,
      "num_tokens": 39831788.0,
      "step": 15715
    },
    {
      "entropy": 6.253791332244873,
      "epoch": 1.814195037507213,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004679978200922339,
      "loss": 5.7456,
      "mean_token_accuracy": 0.1949395641684532,
      "num_tokens": 39844043.0,
      "step": 15720
    },
    {
      "entropy": 6.224784421920776,
      "epoch": 1.8147720715522215,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004679766054483457,
      "loss": 5.8309,
      "mean_token_accuracy": 0.19360052943229675,
      "num_tokens": 39857351.0,
      "step": 15725
    },
    {
      "entropy": 6.193504762649536,
      "epoch": 1.8153491055972304,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004679553843137852,
      "loss": 5.733,
      "mean_token_accuracy": 0.1973591759800911,
      "num_tokens": 39868975.0,
      "step": 15730
    },
    {
      "entropy": 6.188113212585449,
      "epoch": 1.8159261396422388,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046793415668926625,
      "loss": 5.6889,
      "mean_token_accuracy": 0.20426925420761108,
      "num_tokens": 39880995.0,
      "step": 15735
    },
    {
      "entropy": 6.256506490707397,
      "epoch": 1.8165031736872477,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004679129225755028,
      "loss": 5.7023,
      "mean_token_accuracy": 0.1929987534880638,
      "num_tokens": 39893821.0,
      "step": 15740
    },
    {
      "entropy": 6.216187620162964,
      "epoch": 1.817080207732256,
      "grad_norm": 0.9375,
      "learning_rate": 0.000467891681973209,
      "loss": 5.7736,
      "mean_token_accuracy": 0.1991810292005539,
      "num_tokens": 39906938.0,
      "step": 15745
    },
    {
      "entropy": 6.205814266204834,
      "epoch": 1.817657241777265,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046787043488309926,
      "loss": 5.7023,
      "mean_token_accuracy": 0.20117733329534532,
      "num_tokens": 39920647.0,
      "step": 15750
    },
    {
      "entropy": 6.292474889755249,
      "epoch": 1.8182342758222734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004678491813058882,
      "loss": 5.7567,
      "mean_token_accuracy": 0.18879933655261993,
      "num_tokens": 39932925.0,
      "step": 15755
    },
    {
      "entropy": 6.22605447769165,
      "epoch": 1.8188113098672822,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004678279212422908,
      "loss": 5.7648,
      "mean_token_accuracy": 0.1930248484015465,
      "num_tokens": 39945950.0,
      "step": 15760
    },
    {
      "entropy": 6.201946640014649,
      "epoch": 1.8193883439122909,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004678066546930221,
      "loss": 5.8024,
      "mean_token_accuracy": 0.19275868982076644,
      "num_tokens": 39957231.0,
      "step": 15765
    },
    {
      "entropy": 6.2169187545776365,
      "epoch": 1.8199653779572995,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046778538165879725,
      "loss": 5.7043,
      "mean_token_accuracy": 0.1997300773859024,
      "num_tokens": 39969390.0,
      "step": 15770
    },
    {
      "entropy": 6.24988021850586,
      "epoch": 1.8205424120023082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046776410214033185,
      "loss": 5.7842,
      "mean_token_accuracy": 0.19539873003959657,
      "num_tokens": 39982893.0,
      "step": 15775
    },
    {
      "entropy": 6.247120046615601,
      "epoch": 1.8211194460473168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004677428161383417,
      "loss": 5.8012,
      "mean_token_accuracy": 0.19424043893814086,
      "num_tokens": 39996413.0,
      "step": 15780
    },
    {
      "entropy": 6.254956674575806,
      "epoch": 1.8216964800923254,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004677215236535426,
      "loss": 5.8139,
      "mean_token_accuracy": 0.19268016368150712,
      "num_tokens": 40008662.0,
      "step": 15785
    },
    {
      "entropy": 6.223039817810059,
      "epoch": 1.822273514137334,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004677002246866508,
      "loss": 5.7544,
      "mean_token_accuracy": 0.19817463010549546,
      "num_tokens": 40021511.0,
      "step": 15790
    },
    {
      "entropy": 6.194359207153321,
      "epoch": 1.8228505481823427,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046767891923838264,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1944455772638321,
      "num_tokens": 40034205.0,
      "step": 15795
    },
    {
      "entropy": 6.255421161651611,
      "epoch": 1.8234275822273514,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004676576073094548,
      "loss": 5.7875,
      "mean_token_accuracy": 0.1912333220243454,
      "num_tokens": 40046507.0,
      "step": 15800
    },
    {
      "entropy": 6.163492345809937,
      "epoch": 1.8240046162723602,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046763628890058396,
      "loss": 5.6484,
      "mean_token_accuracy": 0.20560259968042374,
      "num_tokens": 40058561.0,
      "step": 15805
    },
    {
      "entropy": 6.227328586578369,
      "epoch": 1.8245816503173686,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046761496401248734,
      "loss": 5.7339,
      "mean_token_accuracy": 0.2057000756263733,
      "num_tokens": 40070351.0,
      "step": 15810
    },
    {
      "entropy": 6.129059457778931,
      "epoch": 1.8251586843623775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000467593632645882,
      "loss": 5.6103,
      "mean_token_accuracy": 0.21389590501785277,
      "num_tokens": 40084057.0,
      "step": 15815
    },
    {
      "entropy": 6.3129823207855225,
      "epoch": 1.825735718407386,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004675722948014855,
      "loss": 5.8632,
      "mean_token_accuracy": 0.18708214312791824,
      "num_tokens": 40097584.0,
      "step": 15820
    },
    {
      "entropy": 6.101898908615112,
      "epoch": 1.8263127524523948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046755095048001556,
      "loss": 5.7465,
      "mean_token_accuracy": 0.20018865019083024,
      "num_tokens": 40110303.0,
      "step": 15825
    },
    {
      "entropy": 6.295633268356323,
      "epoch": 1.8268897864974032,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004675295996821899,
      "loss": 5.8309,
      "mean_token_accuracy": 0.19179421961307525,
      "num_tokens": 40122594.0,
      "step": 15830
    },
    {
      "entropy": 6.270220851898193,
      "epoch": 1.827466820542412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000467508242408727,
      "loss": 5.7189,
      "mean_token_accuracy": 0.19498141705989838,
      "num_tokens": 40133737.0,
      "step": 15835
    },
    {
      "entropy": 6.249630308151245,
      "epoch": 1.8280438545874207,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004674868786603448,
      "loss": 5.7971,
      "mean_token_accuracy": 0.19331606924533845,
      "num_tokens": 40144891.0,
      "step": 15840
    },
    {
      "entropy": 6.252945137023926,
      "epoch": 1.8286208886324293,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004674655084377622,
      "loss": 5.8527,
      "mean_token_accuracy": 0.1925600752234459,
      "num_tokens": 40158217.0,
      "step": 15845
    },
    {
      "entropy": 6.2838582515716555,
      "epoch": 1.829197922677438,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004674441317416978,
      "loss": 5.7463,
      "mean_token_accuracy": 0.1917080193758011,
      "num_tokens": 40170733.0,
      "step": 15850
    },
    {
      "entropy": 6.205789089202881,
      "epoch": 1.8297749567224466,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046742274857287057,
      "loss": 5.7825,
      "mean_token_accuracy": 0.19572699964046478,
      "num_tokens": 40183204.0,
      "step": 15855
    },
    {
      "entropy": 6.280247354507447,
      "epoch": 1.8303519907674553,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004674013589319998,
      "loss": 5.8202,
      "mean_token_accuracy": 0.19004671424627304,
      "num_tokens": 40194816.0,
      "step": 15860
    },
    {
      "entropy": 6.261229610443115,
      "epoch": 1.830929024812464,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004673799628198049,
      "loss": 5.7782,
      "mean_token_accuracy": 0.18980274647474288,
      "num_tokens": 40205796.0,
      "step": 15865
    },
    {
      "entropy": 6.2518385887146,
      "epoch": 1.8315060588574728,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046735856023700546,
      "loss": 5.7203,
      "mean_token_accuracy": 0.18953161984682082,
      "num_tokens": 40218232.0,
      "step": 15870
    },
    {
      "entropy": 6.291367053985596,
      "epoch": 1.8320830929024812,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004673371511843215,
      "loss": 5.8239,
      "mean_token_accuracy": 0.1943132370710373,
      "num_tokens": 40230441.0,
      "step": 15875
    },
    {
      "entropy": 6.274786138534546,
      "epoch": 1.83266012694749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004673157356624729,
      "loss": 5.814,
      "mean_token_accuracy": 0.18727343231439592,
      "num_tokens": 40241858.0,
      "step": 15880
    },
    {
      "entropy": 6.285803937911988,
      "epoch": 1.8332371609924984,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004672943136721801,
      "loss": 5.7139,
      "mean_token_accuracy": 0.19417479634284973,
      "num_tokens": 40254310.0,
      "step": 15885
    },
    {
      "entropy": 6.1876890659332275,
      "epoch": 1.8338141950375073,
      "grad_norm": 0.875,
      "learning_rate": 0.0004672728852141636,
      "loss": 5.7541,
      "mean_token_accuracy": 0.19430211633443834,
      "num_tokens": 40267851.0,
      "step": 15890
    },
    {
      "entropy": 6.3007842063903805,
      "epoch": 1.8343912290825157,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046725145028914404,
      "loss": 5.8036,
      "mean_token_accuracy": 0.1982533484697342,
      "num_tokens": 40279899.0,
      "step": 15895
    },
    {
      "entropy": 6.227988243103027,
      "epoch": 1.8349682631275246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004672300088978425,
      "loss": 5.7977,
      "mean_token_accuracy": 0.1875316709280014,
      "num_tokens": 40291996.0,
      "step": 15900
    },
    {
      "entropy": 6.209213972091675,
      "epoch": 1.8355452971725332,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004672085610409801,
      "loss": 5.6182,
      "mean_token_accuracy": 0.20946350544691086,
      "num_tokens": 40304426.0,
      "step": 15905
    },
    {
      "entropy": 6.21409592628479,
      "epoch": 1.8361223312175419,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046718710671927815,
      "loss": 5.7722,
      "mean_token_accuracy": 0.19383736848831176,
      "num_tokens": 40318287.0,
      "step": 15910
    },
    {
      "entropy": 6.194166040420532,
      "epoch": 1.8366993652625505,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046716564593345843,
      "loss": 5.6838,
      "mean_token_accuracy": 0.1984562784433365,
      "num_tokens": 40330906.0,
      "step": 15915
    },
    {
      "entropy": 6.222714900970459,
      "epoch": 1.8372763993075591,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046714417868424265,
      "loss": 5.75,
      "mean_token_accuracy": 0.19042068868875503,
      "num_tokens": 40342310.0,
      "step": 15920
    },
    {
      "entropy": 6.281884813308716,
      "epoch": 1.8378534333525678,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046712270497235284,
      "loss": 5.7914,
      "mean_token_accuracy": 0.19422318488359452,
      "num_tokens": 40355596.0,
      "step": 15925
    },
    {
      "entropy": 6.293680477142334,
      "epoch": 1.8384304673975764,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004671012247985112,
      "loss": 5.7795,
      "mean_token_accuracy": 0.19289343804121017,
      "num_tokens": 40368841.0,
      "step": 15930
    },
    {
      "entropy": 6.214682102203369,
      "epoch": 1.839007501442585,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046707973816344044,
      "loss": 5.7414,
      "mean_token_accuracy": 0.19698394238948821,
      "num_tokens": 40381496.0,
      "step": 15935
    },
    {
      "entropy": 6.303281211853028,
      "epoch": 1.8395845354875937,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046705824506786304,
      "loss": 5.8184,
      "mean_token_accuracy": 0.1932331934571266,
      "num_tokens": 40393750.0,
      "step": 15940
    },
    {
      "entropy": 6.122893476486206,
      "epoch": 1.8401615695326026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046703674551250193,
      "loss": 5.6058,
      "mean_token_accuracy": 0.21050636619329452,
      "num_tokens": 40406262.0,
      "step": 15945
    },
    {
      "entropy": 6.103861284255982,
      "epoch": 1.840738603577611,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004670152394980803,
      "loss": 5.6044,
      "mean_token_accuracy": 0.20329046696424485,
      "num_tokens": 40418776.0,
      "step": 15950
    },
    {
      "entropy": 6.209722185134888,
      "epoch": 1.8413156376226198,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004669937270253214,
      "loss": 5.8022,
      "mean_token_accuracy": 0.1866112992167473,
      "num_tokens": 40432300.0,
      "step": 15955
    },
    {
      "entropy": 6.221731185913086,
      "epoch": 1.8418926716676283,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000466972208094949,
      "loss": 5.8365,
      "mean_token_accuracy": 0.19025204032659532,
      "num_tokens": 40446215.0,
      "step": 15960
    },
    {
      "entropy": 6.327477598190308,
      "epoch": 1.8424697057126371,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046695068270768665,
      "loss": 5.8,
      "mean_token_accuracy": 0.1913394272327423,
      "num_tokens": 40457932.0,
      "step": 15965
    },
    {
      "entropy": 6.259097099304199,
      "epoch": 1.8430467397576455,
      "grad_norm": 0.875,
      "learning_rate": 0.00046692915086425846,
      "loss": 5.7328,
      "mean_token_accuracy": 0.19911783784627915,
      "num_tokens": 40469897.0,
      "step": 15970
    },
    {
      "entropy": 6.236852693557739,
      "epoch": 1.8436237738026544,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046690761256538857,
      "loss": 5.7791,
      "mean_token_accuracy": 0.1926998108625412,
      "num_tokens": 40483104.0,
      "step": 15975
    },
    {
      "entropy": 6.198714876174927,
      "epoch": 1.844200807847663,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004668860678118015,
      "loss": 5.691,
      "mean_token_accuracy": 0.20391276627779006,
      "num_tokens": 40494844.0,
      "step": 15980
    },
    {
      "entropy": 6.260548400878906,
      "epoch": 1.8447778418926717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046686451660422185,
      "loss": 5.8139,
      "mean_token_accuracy": 0.1951197177171707,
      "num_tokens": 40506972.0,
      "step": 15985
    },
    {
      "entropy": 6.207719135284424,
      "epoch": 1.8453548759376803,
      "grad_norm": 0.828125,
      "learning_rate": 0.00046684295894337455,
      "loss": 5.7693,
      "mean_token_accuracy": 0.1959094986319542,
      "num_tokens": 40519575.0,
      "step": 15990
    },
    {
      "entropy": 6.225582790374756,
      "epoch": 1.845931909982689,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004668213948299845,
      "loss": 5.71,
      "mean_token_accuracy": 0.20351666063070298,
      "num_tokens": 40532895.0,
      "step": 15995
    },
    {
      "entropy": 6.231615304946899,
      "epoch": 1.8465089440276976,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004667998242647772,
      "loss": 5.7344,
      "mean_token_accuracy": 0.19390795528888702,
      "num_tokens": 40544315.0,
      "step": 16000
    },
    {
      "entropy": 6.215937185287475,
      "epoch": 1.8470859780727062,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000466778247248478,
      "loss": 5.7746,
      "mean_token_accuracy": 0.19531953036785127,
      "num_tokens": 40557942.0,
      "step": 16005
    },
    {
      "entropy": 6.295252847671509,
      "epoch": 1.847663012117715,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046675666378181275,
      "loss": 5.8413,
      "mean_token_accuracy": 0.18516817837953567,
      "num_tokens": 40570364.0,
      "step": 16010
    },
    {
      "entropy": 6.286907768249511,
      "epoch": 1.8482400461627235,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004667350738655074,
      "loss": 5.8771,
      "mean_token_accuracy": 0.18463923782110214,
      "num_tokens": 40583259.0,
      "step": 16015
    },
    {
      "entropy": 6.203505611419677,
      "epoch": 1.8488170802077324,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046671347750028806,
      "loss": 5.7028,
      "mean_token_accuracy": 0.1987761825323105,
      "num_tokens": 40595810.0,
      "step": 16020
    },
    {
      "entropy": 6.262021017074585,
      "epoch": 1.8493941142527408,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004666918746868811,
      "loss": 5.8156,
      "mean_token_accuracy": 0.19647061079740524,
      "num_tokens": 40606856.0,
      "step": 16025
    },
    {
      "entropy": 6.281299161911011,
      "epoch": 1.8499711482977497,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004666702654260133,
      "loss": 5.8329,
      "mean_token_accuracy": 0.18882491290569306,
      "num_tokens": 40618764.0,
      "step": 16030
    },
    {
      "entropy": 6.321339464187622,
      "epoch": 1.850548182342758,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046664864971841113,
      "loss": 5.8081,
      "mean_token_accuracy": 0.19403723925352095,
      "num_tokens": 40631833.0,
      "step": 16035
    },
    {
      "entropy": 6.318940353393555,
      "epoch": 1.851125216387767,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046662702756480195,
      "loss": 5.8223,
      "mean_token_accuracy": 0.191567325592041,
      "num_tokens": 40644215.0,
      "step": 16040
    },
    {
      "entropy": 6.246023750305175,
      "epoch": 1.8517022504327756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046660539896591286,
      "loss": 5.7016,
      "mean_token_accuracy": 0.20183758735656737,
      "num_tokens": 40656473.0,
      "step": 16045
    },
    {
      "entropy": 6.2594574928283695,
      "epoch": 1.8522792844777842,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004665837639224713,
      "loss": 5.741,
      "mean_token_accuracy": 0.20216525346040726,
      "num_tokens": 40668949.0,
      "step": 16050
    },
    {
      "entropy": 6.259402275085449,
      "epoch": 1.8528563185227929,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046656212243520505,
      "loss": 5.8044,
      "mean_token_accuracy": 0.19243557900190353,
      "num_tokens": 40681550.0,
      "step": 16055
    },
    {
      "entropy": 6.14393572807312,
      "epoch": 1.8534333525678015,
      "grad_norm": 1.125,
      "learning_rate": 0.00046654047450484193,
      "loss": 5.6616,
      "mean_token_accuracy": 0.1994571343064308,
      "num_tokens": 40695197.0,
      "step": 16060
    },
    {
      "entropy": 6.239932250976563,
      "epoch": 1.8540103866128101,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004665188201321102,
      "loss": 5.7274,
      "mean_token_accuracy": 0.199634413421154,
      "num_tokens": 40707091.0,
      "step": 16065
    },
    {
      "entropy": 6.216319990158081,
      "epoch": 1.8545874206578188,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046649715931773795,
      "loss": 5.6478,
      "mean_token_accuracy": 0.20916593223810195,
      "num_tokens": 40719929.0,
      "step": 16070
    },
    {
      "entropy": 6.195571804046631,
      "epoch": 1.8551644547028274,
      "grad_norm": 0.90625,
      "learning_rate": 0.000466475492062454,
      "loss": 5.7223,
      "mean_token_accuracy": 0.20459873080253602,
      "num_tokens": 40732897.0,
      "step": 16075
    },
    {
      "entropy": 6.13204402923584,
      "epoch": 1.855741488747836,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046645381836698684,
      "loss": 5.6772,
      "mean_token_accuracy": 0.19863571375608444,
      "num_tokens": 40746427.0,
      "step": 16080
    },
    {
      "entropy": 6.257300329208374,
      "epoch": 1.856318522792845,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004664321382320657,
      "loss": 5.7591,
      "mean_token_accuracy": 0.2003425642848015,
      "num_tokens": 40758538.0,
      "step": 16085
    },
    {
      "entropy": 6.287589454650879,
      "epoch": 1.8568955568378533,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046641045165841965,
      "loss": 5.8307,
      "mean_token_accuracy": 0.1933046504855156,
      "num_tokens": 40772008.0,
      "step": 16090
    },
    {
      "entropy": 6.241870450973511,
      "epoch": 1.8574725908828622,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046638875864677814,
      "loss": 5.7942,
      "mean_token_accuracy": 0.18724849075078964,
      "num_tokens": 40785654.0,
      "step": 16095
    },
    {
      "entropy": 6.3359452247619625,
      "epoch": 1.8580496249278706,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004663670591978708,
      "loss": 5.8144,
      "mean_token_accuracy": 0.19239169955253602,
      "num_tokens": 40799394.0,
      "step": 16100
    },
    {
      "entropy": 6.317283344268799,
      "epoch": 1.8586266589728795,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004663453533124275,
      "loss": 5.8393,
      "mean_token_accuracy": 0.18916345238685608,
      "num_tokens": 40812054.0,
      "step": 16105
    },
    {
      "entropy": 6.244693946838379,
      "epoch": 1.859203693017888,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004663236409911783,
      "loss": 5.8867,
      "mean_token_accuracy": 0.1841985195875168,
      "num_tokens": 40825138.0,
      "step": 16110
    },
    {
      "entropy": 6.281439399719238,
      "epoch": 1.8597807270628968,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046630192223485345,
      "loss": 5.8328,
      "mean_token_accuracy": 0.18876519799232483,
      "num_tokens": 40837766.0,
      "step": 16115
    },
    {
      "entropy": 6.283810615539551,
      "epoch": 1.8603577611079054,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046628019704418345,
      "loss": 5.7578,
      "mean_token_accuracy": 0.1991705119609833,
      "num_tokens": 40851042.0,
      "step": 16120
    },
    {
      "entropy": 6.17557053565979,
      "epoch": 1.860934795152914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000466258465419899,
      "loss": 5.643,
      "mean_token_accuracy": 0.2127433329820633,
      "num_tokens": 40864074.0,
      "step": 16125
    },
    {
      "entropy": 6.240874719619751,
      "epoch": 1.8615118291979227,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004662367273627312,
      "loss": 5.7528,
      "mean_token_accuracy": 0.20022245198488237,
      "num_tokens": 40876404.0,
      "step": 16130
    },
    {
      "entropy": 6.272482919692993,
      "epoch": 1.8620888632429313,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000466214982873411,
      "loss": 5.7251,
      "mean_token_accuracy": 0.19963156431913376,
      "num_tokens": 40889684.0,
      "step": 16135
    },
    {
      "entropy": 6.2206511974334715,
      "epoch": 1.86266589728794,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046619323195266975,
      "loss": 5.6892,
      "mean_token_accuracy": 0.20303053110837938,
      "num_tokens": 40902223.0,
      "step": 16140
    },
    {
      "entropy": 6.2893890857696535,
      "epoch": 1.8632429313329486,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004661714746012392,
      "loss": 5.8189,
      "mean_token_accuracy": 0.19742645919322968,
      "num_tokens": 40916251.0,
      "step": 16145
    },
    {
      "entropy": 6.2405060768127445,
      "epoch": 1.8638199653779572,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000466149710819851,
      "loss": 5.8305,
      "mean_token_accuracy": 0.19311869889497757,
      "num_tokens": 40929248.0,
      "step": 16150
    },
    {
      "entropy": 6.244088745117187,
      "epoch": 1.8643969994229659,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004661279406092373,
      "loss": 5.7278,
      "mean_token_accuracy": 0.19801610112190246,
      "num_tokens": 40943575.0,
      "step": 16155
    },
    {
      "entropy": 6.314790391921997,
      "epoch": 1.8649740334679747,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046610616397013027,
      "loss": 5.7805,
      "mean_token_accuracy": 0.19115549176931382,
      "num_tokens": 40956256.0,
      "step": 16160
    },
    {
      "entropy": 6.245018053054809,
      "epoch": 1.8655510675129832,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004660843809032623,
      "loss": 5.8426,
      "mean_token_accuracy": 0.18806178271770477,
      "num_tokens": 40968786.0,
      "step": 16165
    },
    {
      "entropy": 6.2229218006134035,
      "epoch": 1.866128101557992,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004660625914093661,
      "loss": 5.8262,
      "mean_token_accuracy": 0.19061132222414018,
      "num_tokens": 40982221.0,
      "step": 16170
    },
    {
      "entropy": 6.280834817886353,
      "epoch": 1.8667051356030004,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004660407954891745,
      "loss": 5.797,
      "mean_token_accuracy": 0.18846398890018462,
      "num_tokens": 40995226.0,
      "step": 16175
    },
    {
      "entropy": 6.271138334274292,
      "epoch": 1.8672821696480093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004660189931434207,
      "loss": 5.7236,
      "mean_token_accuracy": 0.1999063178896904,
      "num_tokens": 41007251.0,
      "step": 16180
    },
    {
      "entropy": 6.273029899597168,
      "epoch": 1.8678592036930177,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004659971843728379,
      "loss": 5.778,
      "mean_token_accuracy": 0.1993556424975395,
      "num_tokens": 41020218.0,
      "step": 16185
    },
    {
      "entropy": 6.254138040542602,
      "epoch": 1.8684362377380266,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004659753691781597,
      "loss": 5.7543,
      "mean_token_accuracy": 0.19554494321346283,
      "num_tokens": 41032862.0,
      "step": 16190
    },
    {
      "entropy": 6.179647922515869,
      "epoch": 1.8690132717830352,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004659535475601198,
      "loss": 5.6401,
      "mean_token_accuracy": 0.20295362770557404,
      "num_tokens": 41045992.0,
      "step": 16195
    },
    {
      "entropy": 6.16816816329956,
      "epoch": 1.8695903058280439,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046593171951945226,
      "loss": 5.6973,
      "mean_token_accuracy": 0.19748976081609726,
      "num_tokens": 41057902.0,
      "step": 16200
    },
    {
      "entropy": 6.20846962928772,
      "epoch": 1.8701673398730525,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046590988505689114,
      "loss": 5.7063,
      "mean_token_accuracy": 0.20045862048864366,
      "num_tokens": 41070318.0,
      "step": 16205
    },
    {
      "entropy": 6.201190710067749,
      "epoch": 1.8707443739180611,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046588804417317084,
      "loss": 5.784,
      "mean_token_accuracy": 0.19457844495773316,
      "num_tokens": 41083918.0,
      "step": 16210
    },
    {
      "entropy": 6.172313404083252,
      "epoch": 1.8713214079630698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046586619686902597,
      "loss": 5.7244,
      "mean_token_accuracy": 0.20320954769849778,
      "num_tokens": 41097730.0,
      "step": 16215
    },
    {
      "entropy": 6.292459297180176,
      "epoch": 1.8718984420080784,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046584434314519144,
      "loss": 5.8313,
      "mean_token_accuracy": 0.19101618379354476,
      "num_tokens": 41110795.0,
      "step": 16220
    },
    {
      "entropy": 6.246219968795776,
      "epoch": 1.8724754760530873,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004658224830024022,
      "loss": 5.7284,
      "mean_token_accuracy": 0.19725659340620041,
      "num_tokens": 41123216.0,
      "step": 16225
    },
    {
      "entropy": 6.259813928604126,
      "epoch": 1.8730525100980957,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004658006164413935,
      "loss": 5.7768,
      "mean_token_accuracy": 0.1950672298669815,
      "num_tokens": 41135293.0,
      "step": 16230
    },
    {
      "entropy": 6.174546813964843,
      "epoch": 1.8736295441431046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004657787434629009,
      "loss": 5.7653,
      "mean_token_accuracy": 0.20130514055490495,
      "num_tokens": 41149475.0,
      "step": 16235
    },
    {
      "entropy": 6.306607246398926,
      "epoch": 1.874206578188113,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046575686406765993,
      "loss": 5.8288,
      "mean_token_accuracy": 0.19102464318275453,
      "num_tokens": 41162380.0,
      "step": 16240
    },
    {
      "entropy": 6.222295618057251,
      "epoch": 1.8747836122331218,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046573497825640664,
      "loss": 5.6547,
      "mean_token_accuracy": 0.19223218113183976,
      "num_tokens": 41175483.0,
      "step": 16245
    },
    {
      "entropy": 6.086669492721557,
      "epoch": 1.8753606462781303,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004657130860298771,
      "loss": 5.6249,
      "mean_token_accuracy": 0.2080286458134651,
      "num_tokens": 41189342.0,
      "step": 16250
    },
    {
      "entropy": 6.164557933807373,
      "epoch": 1.8759376803231391,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004656911873888077,
      "loss": 5.7228,
      "mean_token_accuracy": 0.20647228062152861,
      "num_tokens": 41202342.0,
      "step": 16255
    },
    {
      "entropy": 6.175295782089234,
      "epoch": 1.8765147143681478,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004656692823339349,
      "loss": 5.6372,
      "mean_token_accuracy": 0.20017538219690323,
      "num_tokens": 41215006.0,
      "step": 16260
    },
    {
      "entropy": 6.281871128082275,
      "epoch": 1.8770917484131564,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004656473708659955,
      "loss": 5.7665,
      "mean_token_accuracy": 0.19565648436546326,
      "num_tokens": 41227735.0,
      "step": 16265
    },
    {
      "entropy": 6.219022750854492,
      "epoch": 1.877668782458165,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00046562545298572646,
      "loss": 5.7405,
      "mean_token_accuracy": 0.19766054600477218,
      "num_tokens": 41240731.0,
      "step": 16270
    },
    {
      "entropy": 6.272245693206787,
      "epoch": 1.8782458165031737,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000465603528693865,
      "loss": 5.7728,
      "mean_token_accuracy": 0.19435095191001892,
      "num_tokens": 41253190.0,
      "step": 16275
    },
    {
      "entropy": 6.273740673065186,
      "epoch": 1.8788228505481823,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046558159799114853,
      "loss": 5.8082,
      "mean_token_accuracy": 0.18764509707689286,
      "num_tokens": 41265224.0,
      "step": 16280
    },
    {
      "entropy": 6.144162893295288,
      "epoch": 1.879399884593191,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004655596608783147,
      "loss": 5.6823,
      "mean_token_accuracy": 0.2048535168170929,
      "num_tokens": 41277865.0,
      "step": 16285
    },
    {
      "entropy": 6.175737810134888,
      "epoch": 1.8799769186381996,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046553771735610136,
      "loss": 5.6981,
      "mean_token_accuracy": 0.19645863473415376,
      "num_tokens": 41291660.0,
      "step": 16290
    },
    {
      "entropy": 6.276864051818848,
      "epoch": 1.8805539526832082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004655157674252465,
      "loss": 5.8442,
      "mean_token_accuracy": 0.18751444518566132,
      "num_tokens": 41304284.0,
      "step": 16295
    },
    {
      "entropy": 6.2097259044647215,
      "epoch": 1.881130986728217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046549381108648843,
      "loss": 5.7218,
      "mean_token_accuracy": 0.20165908485651016,
      "num_tokens": 41316541.0,
      "step": 16300
    },
    {
      "entropy": 6.257354593276977,
      "epoch": 1.8817080207732255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004654718483405656,
      "loss": 5.7424,
      "mean_token_accuracy": 0.2002338483929634,
      "num_tokens": 41329809.0,
      "step": 16305
    },
    {
      "entropy": 6.22561731338501,
      "epoch": 1.8822850548182344,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046544987918821685,
      "loss": 5.7235,
      "mean_token_accuracy": 0.1975063219666481,
      "num_tokens": 41341883.0,
      "step": 16310
    },
    {
      "entropy": 6.199308443069458,
      "epoch": 1.8828620888632428,
      "grad_norm": 0.921875,
      "learning_rate": 0.000465427903630181,
      "loss": 5.6612,
      "mean_token_accuracy": 0.19936496615409852,
      "num_tokens": 41354161.0,
      "step": 16315
    },
    {
      "entropy": 6.178978061676025,
      "epoch": 1.8834391229082517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004654059216671972,
      "loss": 5.6034,
      "mean_token_accuracy": 0.20627519339323044,
      "num_tokens": 41366737.0,
      "step": 16320
    },
    {
      "entropy": 6.259507942199707,
      "epoch": 1.88401615695326,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004653839333000048,
      "loss": 5.7458,
      "mean_token_accuracy": 0.1976940467953682,
      "num_tokens": 41380525.0,
      "step": 16325
    },
    {
      "entropy": 6.216932678222657,
      "epoch": 1.884593190998269,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046536193852934334,
      "loss": 5.6945,
      "mean_token_accuracy": 0.19599267840385437,
      "num_tokens": 41393361.0,
      "step": 16330
    },
    {
      "entropy": 6.230396699905396,
      "epoch": 1.8851702250432776,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046533993735595273,
      "loss": 5.7698,
      "mean_token_accuracy": 0.19981591999530793,
      "num_tokens": 41405919.0,
      "step": 16335
    },
    {
      "entropy": 6.294840908050537,
      "epoch": 1.8857472590882862,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046531792978057277,
      "loss": 5.7687,
      "mean_token_accuracy": 0.19727562516927719,
      "num_tokens": 41417870.0,
      "step": 16340
    },
    {
      "entropy": 6.191200876235962,
      "epoch": 1.8863242931332949,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004652959158039438,
      "loss": 5.7332,
      "mean_token_accuracy": 0.20186103284358978,
      "num_tokens": 41429242.0,
      "step": 16345
    },
    {
      "entropy": 6.271541881561279,
      "epoch": 1.8869013271783035,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004652738954268062,
      "loss": 5.8307,
      "mean_token_accuracy": 0.18908559828996657,
      "num_tokens": 41442073.0,
      "step": 16350
    },
    {
      "entropy": 6.25528883934021,
      "epoch": 1.8874783612233121,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046525186864990073,
      "loss": 5.8332,
      "mean_token_accuracy": 0.18892111182212828,
      "num_tokens": 41453429.0,
      "step": 16355
    },
    {
      "entropy": 6.262342262268066,
      "epoch": 1.8880553952683208,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046522983547396806,
      "loss": 5.7707,
      "mean_token_accuracy": 0.19305487275123595,
      "num_tokens": 41464940.0,
      "step": 16360
    },
    {
      "entropy": 6.248039484024048,
      "epoch": 1.8886324293133296,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004652077958997494,
      "loss": 5.7557,
      "mean_token_accuracy": 0.19191061407327653,
      "num_tokens": 41477562.0,
      "step": 16365
    },
    {
      "entropy": 6.219162654876709,
      "epoch": 1.889209463358338,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046518574992798604,
      "loss": 5.7286,
      "mean_token_accuracy": 0.20486625730991365,
      "num_tokens": 41490299.0,
      "step": 16370
    },
    {
      "entropy": 6.213928127288819,
      "epoch": 1.889786497403347,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046516369755941945,
      "loss": 5.6834,
      "mean_token_accuracy": 0.1972496286034584,
      "num_tokens": 41501305.0,
      "step": 16375
    },
    {
      "entropy": 6.280097913742066,
      "epoch": 1.8903635314483553,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004651416387947914,
      "loss": 5.8167,
      "mean_token_accuracy": 0.19565966725349426,
      "num_tokens": 41515390.0,
      "step": 16380
    },
    {
      "entropy": 6.297719860076905,
      "epoch": 1.8909405654933642,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004651195736348437,
      "loss": 5.8235,
      "mean_token_accuracy": 0.19829970449209214,
      "num_tokens": 41528247.0,
      "step": 16385
    },
    {
      "entropy": 6.231280517578125,
      "epoch": 1.8915175995383726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004650975020803186,
      "loss": 5.6938,
      "mean_token_accuracy": 0.19400931894779205,
      "num_tokens": 41542249.0,
      "step": 16390
    },
    {
      "entropy": 6.2088109970092775,
      "epoch": 1.8920946335833815,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004650754241319584,
      "loss": 5.7039,
      "mean_token_accuracy": 0.1979651540517807,
      "num_tokens": 41554835.0,
      "step": 16395
    },
    {
      "entropy": 6.293802642822266,
      "epoch": 1.8926716676283901,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046505333979050573,
      "loss": 5.8084,
      "mean_token_accuracy": 0.1877210408449173,
      "num_tokens": 41566427.0,
      "step": 16400
    },
    {
      "entropy": 6.232387542724609,
      "epoch": 1.8932487016733988,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004650312490567035,
      "loss": 5.8099,
      "mean_token_accuracy": 0.19262741655111312,
      "num_tokens": 41580193.0,
      "step": 16405
    },
    {
      "entropy": 6.173836040496826,
      "epoch": 1.8938257357184074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004650091519312945,
      "loss": 5.6346,
      "mean_token_accuracy": 0.20523083955049515,
      "num_tokens": 41594048.0,
      "step": 16410
    },
    {
      "entropy": 6.279299545288086,
      "epoch": 1.894402769763416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046498704841502203,
      "loss": 5.7853,
      "mean_token_accuracy": 0.19502951204776764,
      "num_tokens": 41607283.0,
      "step": 16415
    },
    {
      "entropy": 6.185412120819092,
      "epoch": 1.8949798038084247,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004649649385086296,
      "loss": 5.6745,
      "mean_token_accuracy": 0.2010557919740677,
      "num_tokens": 41619844.0,
      "step": 16420
    },
    {
      "entropy": 6.102963018417358,
      "epoch": 1.8955568378534333,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004649428222128608,
      "loss": 5.634,
      "mean_token_accuracy": 0.20188225209712982,
      "num_tokens": 41631764.0,
      "step": 16425
    },
    {
      "entropy": 6.314756202697754,
      "epoch": 1.896133871898442,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046492069952845953,
      "loss": 5.7826,
      "mean_token_accuracy": 0.1914699226617813,
      "num_tokens": 41643703.0,
      "step": 16430
    },
    {
      "entropy": 6.213639163970948,
      "epoch": 1.8967109059434506,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046489857045617,
      "loss": 5.7464,
      "mean_token_accuracy": 0.20060871094465255,
      "num_tokens": 41657057.0,
      "step": 16435
    },
    {
      "entropy": 6.219952583312988,
      "epoch": 1.8972879399884595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046487643499673625,
      "loss": 5.7822,
      "mean_token_accuracy": 0.1977525144815445,
      "num_tokens": 41669870.0,
      "step": 16440
    },
    {
      "entropy": 6.266088628768921,
      "epoch": 1.8978649740334679,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004648542931509029,
      "loss": 5.7313,
      "mean_token_accuracy": 0.1937894716858864,
      "num_tokens": 41681015.0,
      "step": 16445
    },
    {
      "entropy": 6.201165676116943,
      "epoch": 1.8984420080784767,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004648321449194148,
      "loss": 5.735,
      "mean_token_accuracy": 0.20242813378572463,
      "num_tokens": 41694116.0,
      "step": 16450
    },
    {
      "entropy": 6.206316947937012,
      "epoch": 1.8990190421234852,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046480999030301673,
      "loss": 5.6843,
      "mean_token_accuracy": 0.2028385207056999,
      "num_tokens": 41707294.0,
      "step": 16455
    },
    {
      "entropy": 6.288750743865966,
      "epoch": 1.899596076168494,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046478782930245395,
      "loss": 5.8456,
      "mean_token_accuracy": 0.19137783497571945,
      "num_tokens": 41719837.0,
      "step": 16460
    },
    {
      "entropy": 6.271579790115356,
      "epoch": 1.9001731102135024,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046476566191847176,
      "loss": 5.7061,
      "mean_token_accuracy": 0.19760711044073104,
      "num_tokens": 41732308.0,
      "step": 16465
    },
    {
      "entropy": 6.212190818786621,
      "epoch": 1.9007501442585113,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004647434881518159,
      "loss": 5.704,
      "mean_token_accuracy": 0.1995425671339035,
      "num_tokens": 41745051.0,
      "step": 16470
    },
    {
      "entropy": 6.254608392715454,
      "epoch": 1.90132717830352,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046472130800323194,
      "loss": 5.7598,
      "mean_token_accuracy": 0.19814508110284806,
      "num_tokens": 41757545.0,
      "step": 16475
    },
    {
      "entropy": 6.1872954845428465,
      "epoch": 1.9019042123485286,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004646991214734661,
      "loss": 5.7444,
      "mean_token_accuracy": 0.1966193735599518,
      "num_tokens": 41772157.0,
      "step": 16480
    },
    {
      "entropy": 6.265322637557984,
      "epoch": 1.9024812463935372,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004646769285632645,
      "loss": 5.7584,
      "mean_token_accuracy": 0.19746726006269455,
      "num_tokens": 41785404.0,
      "step": 16485
    },
    {
      "entropy": 6.199524545669556,
      "epoch": 1.9030582804385459,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004646547292733737,
      "loss": 5.7119,
      "mean_token_accuracy": 0.20719417333602905,
      "num_tokens": 41797758.0,
      "step": 16490
    },
    {
      "entropy": 6.1841377258300785,
      "epoch": 1.9036353144835545,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004646325236045402,
      "loss": 5.7871,
      "mean_token_accuracy": 0.1993813306093216,
      "num_tokens": 41811162.0,
      "step": 16495
    },
    {
      "entropy": 6.209798955917359,
      "epoch": 1.9042123485285631,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000464610311557511,
      "loss": 5.6761,
      "mean_token_accuracy": 0.2027891531586647,
      "num_tokens": 41823304.0,
      "step": 16500
    },
    {
      "entropy": 6.117609024047852,
      "epoch": 1.904789382573572,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004645880931330331,
      "loss": 5.6481,
      "mean_token_accuracy": 0.19958235770463945,
      "num_tokens": 41837027.0,
      "step": 16505
    },
    {
      "entropy": 6.213974094390869,
      "epoch": 1.9053664166185804,
      "grad_norm": 0.875,
      "learning_rate": 0.0004645658683318539,
      "loss": 5.7499,
      "mean_token_accuracy": 0.19403923898935319,
      "num_tokens": 41851887.0,
      "step": 16510
    },
    {
      "entropy": 6.2310285568237305,
      "epoch": 1.9059434506635893,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004645436371547209,
      "loss": 5.7188,
      "mean_token_accuracy": 0.20376719683408737,
      "num_tokens": 41864908.0,
      "step": 16515
    },
    {
      "entropy": 6.273329210281372,
      "epoch": 1.9065204847085977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046452139960238186,
      "loss": 5.7057,
      "mean_token_accuracy": 0.2070438876748085,
      "num_tokens": 41877604.0,
      "step": 16520
    },
    {
      "entropy": 6.259437704086304,
      "epoch": 1.9070975187536066,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046449915567558467,
      "loss": 5.7682,
      "mean_token_accuracy": 0.18715291023254393,
      "num_tokens": 41890580.0,
      "step": 16525
    },
    {
      "entropy": 6.215609359741211,
      "epoch": 1.907674552798615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004644769053750775,
      "loss": 5.732,
      "mean_token_accuracy": 0.1968011423945427,
      "num_tokens": 41903833.0,
      "step": 16530
    },
    {
      "entropy": 6.245735788345337,
      "epoch": 1.9082515868436238,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004644546487016087,
      "loss": 5.7621,
      "mean_token_accuracy": 0.19157301783561706,
      "num_tokens": 41916094.0,
      "step": 16535
    },
    {
      "entropy": 6.201154613494873,
      "epoch": 1.9088286208886325,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046443238565592687,
      "loss": 5.6802,
      "mean_token_accuracy": 0.20244794189929963,
      "num_tokens": 41929199.0,
      "step": 16540
    },
    {
      "entropy": 6.213739538192749,
      "epoch": 1.9094056549336411,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046441011623878087,
      "loss": 5.7411,
      "mean_token_accuracy": 0.19532374590635299,
      "num_tokens": 41942254.0,
      "step": 16545
    },
    {
      "entropy": 6.2611377239227295,
      "epoch": 1.9099826889786498,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004643878404509197,
      "loss": 5.7613,
      "mean_token_accuracy": 0.1942107930779457,
      "num_tokens": 41954129.0,
      "step": 16550
    },
    {
      "entropy": 6.296992492675781,
      "epoch": 1.9105597230236584,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046436555829309264,
      "loss": 5.7828,
      "mean_token_accuracy": 0.18973256200551986,
      "num_tokens": 41966489.0,
      "step": 16555
    },
    {
      "entropy": 6.155859804153442,
      "epoch": 1.911136757068667,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000464343269766049,
      "loss": 5.7043,
      "mean_token_accuracy": 0.20632706135511397,
      "num_tokens": 41980235.0,
      "step": 16560
    },
    {
      "entropy": 6.232020282745362,
      "epoch": 1.9117137911136757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046432097487053857,
      "loss": 5.7039,
      "mean_token_accuracy": 0.2020736113190651,
      "num_tokens": 41992531.0,
      "step": 16565
    },
    {
      "entropy": 6.202987337112427,
      "epoch": 1.9122908251586843,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046429867360731124,
      "loss": 5.708,
      "mean_token_accuracy": 0.20205148458480834,
      "num_tokens": 42004825.0,
      "step": 16570
    },
    {
      "entropy": 6.22787184715271,
      "epoch": 1.912867859203693,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046427636597711695,
      "loss": 5.6643,
      "mean_token_accuracy": 0.2029886856675148,
      "num_tokens": 42017275.0,
      "step": 16575
    },
    {
      "entropy": 6.226530694961548,
      "epoch": 1.9134448932487018,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046425405198070624,
      "loss": 5.7239,
      "mean_token_accuracy": 0.20072105079889296,
      "num_tokens": 42031102.0,
      "step": 16580
    },
    {
      "entropy": 6.232250833511353,
      "epoch": 1.9140219272937102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046423173161882944,
      "loss": 5.7839,
      "mean_token_accuracy": 0.18810754120349885,
      "num_tokens": 42043276.0,
      "step": 16585
    },
    {
      "entropy": 6.275153255462646,
      "epoch": 1.914598961338719,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046420940489223735,
      "loss": 5.7531,
      "mean_token_accuracy": 0.19817679822444917,
      "num_tokens": 42055634.0,
      "step": 16590
    },
    {
      "entropy": 6.205233240127564,
      "epoch": 1.9151759953837275,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004641870718016809,
      "loss": 5.6734,
      "mean_token_accuracy": 0.19888432025909425,
      "num_tokens": 42068910.0,
      "step": 16595
    },
    {
      "entropy": 6.139797496795654,
      "epoch": 1.9157530294287364,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004641647323479113,
      "loss": 5.6501,
      "mean_token_accuracy": 0.200643490254879,
      "num_tokens": 42081008.0,
      "step": 16600
    },
    {
      "entropy": 6.2025312900543215,
      "epoch": 1.9163300634737448,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004641423865316798,
      "loss": 5.7393,
      "mean_token_accuracy": 0.20066003799438475,
      "num_tokens": 42093982.0,
      "step": 16605
    },
    {
      "entropy": 6.279962968826294,
      "epoch": 1.9169070975187537,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004641200343537381,
      "loss": 5.7564,
      "mean_token_accuracy": 0.19527011513710021,
      "num_tokens": 42105441.0,
      "step": 16610
    },
    {
      "entropy": 6.294757938385009,
      "epoch": 1.9174841315637623,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000464097675814838,
      "loss": 5.7602,
      "mean_token_accuracy": 0.19696361422538758,
      "num_tokens": 42117636.0,
      "step": 16615
    },
    {
      "entropy": 6.238898468017578,
      "epoch": 1.918061165608771,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004640753109157316,
      "loss": 5.7698,
      "mean_token_accuracy": 0.19026945978403093,
      "num_tokens": 42130009.0,
      "step": 16620
    },
    {
      "entropy": 6.233647394180298,
      "epoch": 1.9186381996537796,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046405293965717093,
      "loss": 5.7454,
      "mean_token_accuracy": 0.19405922442674636,
      "num_tokens": 42142181.0,
      "step": 16625
    },
    {
      "entropy": 6.2317156314849855,
      "epoch": 1.9192152336987882,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004640305620399086,
      "loss": 5.7528,
      "mean_token_accuracy": 0.2025137037038803,
      "num_tokens": 42155626.0,
      "step": 16630
    },
    {
      "entropy": 6.188432455062866,
      "epoch": 1.9197922677437969,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004640081780646971,
      "loss": 5.7471,
      "mean_token_accuracy": 0.20006103962659835,
      "num_tokens": 42169309.0,
      "step": 16635
    },
    {
      "entropy": 6.308851099014282,
      "epoch": 1.9203693017888055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046398578773228954,
      "loss": 5.7859,
      "mean_token_accuracy": 0.19979457110166549,
      "num_tokens": 42182372.0,
      "step": 16640
    },
    {
      "entropy": 6.164843130111694,
      "epoch": 1.9209463358338144,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046396339104343886,
      "loss": 5.6279,
      "mean_token_accuracy": 0.20405071824789048,
      "num_tokens": 42194792.0,
      "step": 16645
    },
    {
      "entropy": 6.237176418304443,
      "epoch": 1.9215233698788228,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004639409879988984,
      "loss": 5.7974,
      "mean_token_accuracy": 0.19285742342472076,
      "num_tokens": 42207194.0,
      "step": 16650
    },
    {
      "entropy": 6.1757570743560795,
      "epoch": 1.9221004039238316,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004639185785994216,
      "loss": 5.723,
      "mean_token_accuracy": 0.20376081764698029,
      "num_tokens": 42219058.0,
      "step": 16655
    },
    {
      "entropy": 6.206959056854248,
      "epoch": 1.92267743796884,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046389616284576226,
      "loss": 5.724,
      "mean_token_accuracy": 0.20171431601047515,
      "num_tokens": 42230726.0,
      "step": 16660
    },
    {
      "entropy": 6.078776073455811,
      "epoch": 1.923254472013849,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046387374073867435,
      "loss": 5.5654,
      "mean_token_accuracy": 0.2157515689730644,
      "num_tokens": 42243859.0,
      "step": 16665
    },
    {
      "entropy": 6.237895965576172,
      "epoch": 1.9238315060588573,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046385131227891197,
      "loss": 5.805,
      "mean_token_accuracy": 0.18995663970708848,
      "num_tokens": 42257253.0,
      "step": 16670
    },
    {
      "entropy": 6.257089138031006,
      "epoch": 1.9244085401038662,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004638288774672295,
      "loss": 5.7924,
      "mean_token_accuracy": 0.19080203920602798,
      "num_tokens": 42269202.0,
      "step": 16675
    },
    {
      "entropy": 6.240876007080078,
      "epoch": 1.9249855741488748,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004638064363043816,
      "loss": 5.7595,
      "mean_token_accuracy": 0.19908578991889953,
      "num_tokens": 42282714.0,
      "step": 16680
    },
    {
      "entropy": 6.280086278915405,
      "epoch": 1.9255626081938835,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046378398879112293,
      "loss": 5.7475,
      "mean_token_accuracy": 0.19422405064105988,
      "num_tokens": 42297272.0,
      "step": 16685
    },
    {
      "entropy": 6.283558082580567,
      "epoch": 1.9261396422388921,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004637615349282086,
      "loss": 5.8069,
      "mean_token_accuracy": 0.19420734345912932,
      "num_tokens": 42308938.0,
      "step": 16690
    },
    {
      "entropy": 6.2312945365905765,
      "epoch": 1.9267166762839008,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004637390747163938,
      "loss": 5.7394,
      "mean_token_accuracy": 0.19510370045900344,
      "num_tokens": 42321310.0,
      "step": 16695
    },
    {
      "entropy": 6.113681268692017,
      "epoch": 1.9272937103289094,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004637166081564339,
      "loss": 5.6763,
      "mean_token_accuracy": 0.2055169388651848,
      "num_tokens": 42332760.0,
      "step": 16700
    },
    {
      "entropy": 6.224835157394409,
      "epoch": 1.927870744373918,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00046369413524908473,
      "loss": 5.7248,
      "mean_token_accuracy": 0.20252504646778108,
      "num_tokens": 42346993.0,
      "step": 16705
    },
    {
      "entropy": 6.248994064331055,
      "epoch": 1.9284477784189267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046367165599510207,
      "loss": 5.7783,
      "mean_token_accuracy": 0.1989079162478447,
      "num_tokens": 42358701.0,
      "step": 16710
    },
    {
      "entropy": 6.251170110702515,
      "epoch": 1.9290248124639353,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004636491703952419,
      "loss": 5.8343,
      "mean_token_accuracy": 0.19415884763002395,
      "num_tokens": 42371429.0,
      "step": 16715
    },
    {
      "entropy": 6.263113260269165,
      "epoch": 1.9296018465089442,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046362667845026057,
      "loss": 5.7094,
      "mean_token_accuracy": 0.2015704020857811,
      "num_tokens": 42383346.0,
      "step": 16720
    },
    {
      "entropy": 6.255625009536743,
      "epoch": 1.9301788805539526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046360418016091467,
      "loss": 5.7554,
      "mean_token_accuracy": 0.20293100625276567,
      "num_tokens": 42395247.0,
      "step": 16725
    },
    {
      "entropy": 6.2504359722137455,
      "epoch": 1.9307559145989615,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046358167552796085,
      "loss": 5.7412,
      "mean_token_accuracy": 0.1916235715150833,
      "num_tokens": 42408009.0,
      "step": 16730
    },
    {
      "entropy": 6.213959741592407,
      "epoch": 1.9313329486439699,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046355916455215597,
      "loss": 5.8018,
      "mean_token_accuracy": 0.18905497938394547,
      "num_tokens": 42420367.0,
      "step": 16735
    },
    {
      "entropy": 6.269826936721802,
      "epoch": 1.9319099826889787,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004635366472342573,
      "loss": 5.8266,
      "mean_token_accuracy": 0.19244474619627,
      "num_tokens": 42433298.0,
      "step": 16740
    },
    {
      "entropy": 6.284765005111694,
      "epoch": 1.9324870167339872,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004635141235750222,
      "loss": 5.7734,
      "mean_token_accuracy": 0.1958116129040718,
      "num_tokens": 42446422.0,
      "step": 16745
    },
    {
      "entropy": 6.2704758644104,
      "epoch": 1.933064050778996,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046349159357520815,
      "loss": 5.7711,
      "mean_token_accuracy": 0.18882783353328705,
      "num_tokens": 42458874.0,
      "step": 16750
    },
    {
      "entropy": 6.226549053192139,
      "epoch": 1.9336410848240047,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000463469057235573,
      "loss": 5.7042,
      "mean_token_accuracy": 0.19797363579273225,
      "num_tokens": 42472093.0,
      "step": 16755
    },
    {
      "entropy": 6.102176332473755,
      "epoch": 1.9342181188690133,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046344651455687476,
      "loss": 5.6497,
      "mean_token_accuracy": 0.20643826425075532,
      "num_tokens": 42485396.0,
      "step": 16760
    },
    {
      "entropy": 6.255658769607544,
      "epoch": 1.934795152914022,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046342396553987155,
      "loss": 5.7352,
      "mean_token_accuracy": 0.19764666706323625,
      "num_tokens": 42498379.0,
      "step": 16765
    },
    {
      "entropy": 6.255459642410278,
      "epoch": 1.9353721869590306,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046340141018532186,
      "loss": 5.7513,
      "mean_token_accuracy": 0.1926908478140831,
      "num_tokens": 42512125.0,
      "step": 16770
    },
    {
      "entropy": 6.211010360717774,
      "epoch": 1.9359492210040392,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004633788484939843,
      "loss": 5.6874,
      "mean_token_accuracy": 0.20337988585233688,
      "num_tokens": 42525539.0,
      "step": 16775
    },
    {
      "entropy": 6.207788276672363,
      "epoch": 1.9365262550490479,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046335628046661776,
      "loss": 5.691,
      "mean_token_accuracy": 0.19870318174362184,
      "num_tokens": 42538399.0,
      "step": 16780
    },
    {
      "entropy": 6.290454626083374,
      "epoch": 1.9371032890940567,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046333370610398135,
      "loss": 5.8214,
      "mean_token_accuracy": 0.18948260843753814,
      "num_tokens": 42550607.0,
      "step": 16785
    },
    {
      "entropy": 6.369224977493286,
      "epoch": 1.9376803231390651,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004633111254068342,
      "loss": 5.8571,
      "mean_token_accuracy": 0.1860637202858925,
      "num_tokens": 42563113.0,
      "step": 16790
    },
    {
      "entropy": 6.19832558631897,
      "epoch": 1.938257357184074,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004632885383759359,
      "loss": 5.7358,
      "mean_token_accuracy": 0.19859542846679687,
      "num_tokens": 42574792.0,
      "step": 16795
    },
    {
      "entropy": 6.157328271865845,
      "epoch": 1.9388343912290824,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046326594501204624,
      "loss": 5.6906,
      "mean_token_accuracy": 0.19677329808473587,
      "num_tokens": 42587360.0,
      "step": 16800
    },
    {
      "entropy": 6.239551544189453,
      "epoch": 1.9394114252740913,
      "grad_norm": 0.875,
      "learning_rate": 0.000463243345315925,
      "loss": 5.724,
      "mean_token_accuracy": 0.19548908919095992,
      "num_tokens": 42599901.0,
      "step": 16805
    },
    {
      "entropy": 6.292298984527588,
      "epoch": 1.9399884593190997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046322073928833224,
      "loss": 5.7536,
      "mean_token_accuracy": 0.19571465700864793,
      "num_tokens": 42612127.0,
      "step": 16810
    },
    {
      "entropy": 6.294507312774658,
      "epoch": 1.9405654933641086,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004631981269300285,
      "loss": 5.8631,
      "mean_token_accuracy": 0.19157288670539857,
      "num_tokens": 42624086.0,
      "step": 16815
    },
    {
      "entropy": 6.2822307586669925,
      "epoch": 1.9411425274091172,
      "grad_norm": 0.875,
      "learning_rate": 0.0004631755082417742,
      "loss": 5.8011,
      "mean_token_accuracy": 0.19531398713588716,
      "num_tokens": 42637402.0,
      "step": 16820
    },
    {
      "entropy": 6.295012187957764,
      "epoch": 1.9417195614541258,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004631528832243302,
      "loss": 5.8116,
      "mean_token_accuracy": 0.18646181523799896,
      "num_tokens": 42648162.0,
      "step": 16825
    },
    {
      "entropy": 6.233980464935303,
      "epoch": 1.9422965954991345,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046313025187845735,
      "loss": 5.7494,
      "mean_token_accuracy": 0.19646145403385162,
      "num_tokens": 42658787.0,
      "step": 16830
    },
    {
      "entropy": 6.245473337173462,
      "epoch": 1.942873629544143,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046310761420491705,
      "loss": 5.7408,
      "mean_token_accuracy": 0.18984330743551253,
      "num_tokens": 42672036.0,
      "step": 16835
    },
    {
      "entropy": 6.2862225532531735,
      "epoch": 1.9434506635891518,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004630849702044705,
      "loss": 5.7662,
      "mean_token_accuracy": 0.18849985003471376,
      "num_tokens": 42683691.0,
      "step": 16840
    },
    {
      "entropy": 6.206479740142822,
      "epoch": 1.9440276976341604,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046306231987787937,
      "loss": 5.7195,
      "mean_token_accuracy": 0.19914836585521697,
      "num_tokens": 42697469.0,
      "step": 16845
    },
    {
      "entropy": 6.2218372344970705,
      "epoch": 1.944604731679169,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046303966322590556,
      "loss": 5.728,
      "mean_token_accuracy": 0.19930823594331742,
      "num_tokens": 42710792.0,
      "step": 16850
    },
    {
      "entropy": 6.248984479904175,
      "epoch": 1.9451817657241777,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004630170002493111,
      "loss": 5.7006,
      "mean_token_accuracy": 0.20190690606832504,
      "num_tokens": 42722549.0,
      "step": 16855
    },
    {
      "entropy": 6.196306800842285,
      "epoch": 1.9457587997691865,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046299433094885826,
      "loss": 5.7181,
      "mean_token_accuracy": 0.1972872480750084,
      "num_tokens": 42736113.0,
      "step": 16860
    },
    {
      "entropy": 6.256334400177002,
      "epoch": 1.946335833814195,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046297165532530944,
      "loss": 5.7197,
      "mean_token_accuracy": 0.2009844198822975,
      "num_tokens": 42748801.0,
      "step": 16865
    },
    {
      "entropy": 6.264508056640625,
      "epoch": 1.9469128678592038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004629489733794274,
      "loss": 5.7817,
      "mean_token_accuracy": 0.1920044854283333,
      "num_tokens": 42760738.0,
      "step": 16870
    },
    {
      "entropy": 6.235569095611572,
      "epoch": 1.9474899019042122,
      "grad_norm": 0.984375,
      "learning_rate": 0.000462926285111975,
      "loss": 5.7835,
      "mean_token_accuracy": 0.1918180614709854,
      "num_tokens": 42772470.0,
      "step": 16875
    },
    {
      "entropy": 6.242725419998169,
      "epoch": 1.948066935949221,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046290359052371535,
      "loss": 5.8034,
      "mean_token_accuracy": 0.19629918336868285,
      "num_tokens": 42784919.0,
      "step": 16880
    },
    {
      "entropy": 6.203857469558716,
      "epoch": 1.9486439699942295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004628808896154117,
      "loss": 5.6848,
      "mean_token_accuracy": 0.2041950523853302,
      "num_tokens": 42797722.0,
      "step": 16885
    },
    {
      "entropy": 6.312695503234863,
      "epoch": 1.9492210040392384,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004628581823878277,
      "loss": 5.8037,
      "mean_token_accuracy": 0.19107561111450194,
      "num_tokens": 42809539.0,
      "step": 16890
    },
    {
      "entropy": 6.147143220901489,
      "epoch": 1.949798038084247,
      "grad_norm": 0.953125,
      "learning_rate": 0.000462835468841727,
      "loss": 5.5566,
      "mean_token_accuracy": 0.20410583764314652,
      "num_tokens": 42821519.0,
      "step": 16895
    },
    {
      "entropy": 6.206338357925415,
      "epoch": 1.9503750721292556,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004628127489778737,
      "loss": 5.6709,
      "mean_token_accuracy": 0.20496901422739028,
      "num_tokens": 42833966.0,
      "step": 16900
    },
    {
      "entropy": 6.213610124588013,
      "epoch": 1.9509521061742643,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004627900227970318,
      "loss": 5.7318,
      "mean_token_accuracy": 0.20311963707208633,
      "num_tokens": 42846162.0,
      "step": 16905
    },
    {
      "entropy": 6.236937189102173,
      "epoch": 1.951529140219273,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046276729029996576,
      "loss": 5.7485,
      "mean_token_accuracy": 0.20653784424066543,
      "num_tokens": 42860019.0,
      "step": 16910
    },
    {
      "entropy": 6.261923885345459,
      "epoch": 1.9521061742642816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046274455148744025,
      "loss": 5.8222,
      "mean_token_accuracy": 0.18946326076984404,
      "num_tokens": 42872699.0,
      "step": 16915
    },
    {
      "entropy": 6.229638671875,
      "epoch": 1.9526832083092902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046272180636022,
      "loss": 5.7538,
      "mean_token_accuracy": 0.1984541416168213,
      "num_tokens": 42884907.0,
      "step": 16920
    },
    {
      "entropy": 6.278835010528565,
      "epoch": 1.953260242354299,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046269905491907,
      "loss": 5.7663,
      "mean_token_accuracy": 0.20172708183526994,
      "num_tokens": 42898387.0,
      "step": 16925
    },
    {
      "entropy": 6.284050846099854,
      "epoch": 1.9538372763993075,
      "grad_norm": 0.875,
      "learning_rate": 0.0004626762971647555,
      "loss": 5.7977,
      "mean_token_accuracy": 0.19302857369184495,
      "num_tokens": 42910845.0,
      "step": 16930
    },
    {
      "entropy": 6.245274400711059,
      "epoch": 1.9544143104443163,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046265353309804205,
      "loss": 5.7745,
      "mean_token_accuracy": 0.1925951436161995,
      "num_tokens": 42922494.0,
      "step": 16935
    },
    {
      "entropy": 6.271182346343994,
      "epoch": 1.9549913444893248,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004626307627196952,
      "loss": 5.7322,
      "mean_token_accuracy": 0.19045246690511702,
      "num_tokens": 42934618.0,
      "step": 16940
    },
    {
      "entropy": 6.26016149520874,
      "epoch": 1.9555683785343336,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004626079860304808,
      "loss": 5.6846,
      "mean_token_accuracy": 0.19953396171331406,
      "num_tokens": 42947518.0,
      "step": 16945
    },
    {
      "entropy": 6.165673542022705,
      "epoch": 1.956145412579342,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046258520303116496,
      "loss": 5.7448,
      "mean_token_accuracy": 0.19415563195943833,
      "num_tokens": 42960068.0,
      "step": 16950
    },
    {
      "entropy": 6.247699499130249,
      "epoch": 1.956722446624351,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004625624137225141,
      "loss": 5.7166,
      "mean_token_accuracy": 0.19184452295303345,
      "num_tokens": 42972982.0,
      "step": 16955
    },
    {
      "entropy": 6.170627450942993,
      "epoch": 1.9572994806693593,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004625396181052945,
      "loss": 5.6708,
      "mean_token_accuracy": 0.20777123123407365,
      "num_tokens": 42985040.0,
      "step": 16960
    },
    {
      "entropy": 6.3223930358886715,
      "epoch": 1.9578765147143682,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046251681618027316,
      "loss": 5.7884,
      "mean_token_accuracy": 0.18683255165815355,
      "num_tokens": 42998503.0,
      "step": 16965
    },
    {
      "entropy": 6.185489225387573,
      "epoch": 1.9584535487593768,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004624940079482167,
      "loss": 5.6531,
      "mean_token_accuracy": 0.20618997514247894,
      "num_tokens": 43011334.0,
      "step": 16970
    },
    {
      "entropy": 6.117512512207031,
      "epoch": 1.9590305828043855,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046247119340989254,
      "loss": 5.6636,
      "mean_token_accuracy": 0.21184749007225037,
      "num_tokens": 43024772.0,
      "step": 16975
    },
    {
      "entropy": 6.285856246948242,
      "epoch": 1.959607616849394,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004624483725660678,
      "loss": 5.7181,
      "mean_token_accuracy": 0.20330964475870134,
      "num_tokens": 43036733.0,
      "step": 16980
    },
    {
      "entropy": 6.237366342544556,
      "epoch": 1.9601846508944027,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004624255454175102,
      "loss": 5.7401,
      "mean_token_accuracy": 0.2003675729036331,
      "num_tokens": 43050482.0,
      "step": 16985
    },
    {
      "entropy": 6.334873294830322,
      "epoch": 1.9607616849394114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004624027119649875,
      "loss": 5.8292,
      "mean_token_accuracy": 0.1908613547682762,
      "num_tokens": 43062826.0,
      "step": 16990
    },
    {
      "entropy": 6.218156003952027,
      "epoch": 1.96133871898442,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046237987220926766,
      "loss": 5.6774,
      "mean_token_accuracy": 0.20799431800842286,
      "num_tokens": 43077134.0,
      "step": 16995
    },
    {
      "entropy": 6.153063106536865,
      "epoch": 1.9619157530294289,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046235702615111886,
      "loss": 5.6717,
      "mean_token_accuracy": 0.20705497413873672,
      "num_tokens": 43089871.0,
      "step": 17000
    },
    {
      "entropy": 6.244037055969239,
      "epoch": 1.9624927870744373,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004623341737913096,
      "loss": 5.7721,
      "mean_token_accuracy": 0.19624295234680175,
      "num_tokens": 43102977.0,
      "step": 17005
    },
    {
      "entropy": 6.229075574874878,
      "epoch": 1.9630698211194462,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004623113151306085,
      "loss": 5.7324,
      "mean_token_accuracy": 0.19281073212623595,
      "num_tokens": 43114685.0,
      "step": 17010
    },
    {
      "entropy": 6.201316547393799,
      "epoch": 1.9636468551644546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046228845016978425,
      "loss": 5.6768,
      "mean_token_accuracy": 0.1995202511548996,
      "num_tokens": 43127208.0,
      "step": 17015
    },
    {
      "entropy": 6.219625997543335,
      "epoch": 1.9642238892094634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004622655789096061,
      "loss": 5.652,
      "mean_token_accuracy": 0.2087152749300003,
      "num_tokens": 43140732.0,
      "step": 17020
    },
    {
      "entropy": 6.2197590351104735,
      "epoch": 1.9648009232544719,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046224270135084315,
      "loss": 5.751,
      "mean_token_accuracy": 0.19461191445589066,
      "num_tokens": 43152746.0,
      "step": 17025
    },
    {
      "entropy": 6.198227262496948,
      "epoch": 1.9653779572994807,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046221981749426503,
      "loss": 5.6882,
      "mean_token_accuracy": 0.19643354415893555,
      "num_tokens": 43165608.0,
      "step": 17030
    },
    {
      "entropy": 6.2346631526947025,
      "epoch": 1.9659549913444894,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046219692734064124,
      "loss": 5.7325,
      "mean_token_accuracy": 0.20064710080623627,
      "num_tokens": 43178805.0,
      "step": 17035
    },
    {
      "entropy": 6.28594331741333,
      "epoch": 1.966532025389498,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004621740308907419,
      "loss": 5.8102,
      "mean_token_accuracy": 0.18701407164335251,
      "num_tokens": 43191366.0,
      "step": 17040
    },
    {
      "entropy": 6.269253873825074,
      "epoch": 1.9671090594345066,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004621511281453369,
      "loss": 5.7882,
      "mean_token_accuracy": 0.19155199378728865,
      "num_tokens": 43203748.0,
      "step": 17045
    },
    {
      "entropy": 6.2473005771636965,
      "epoch": 1.9676860934795153,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004621282191051967,
      "loss": 5.7118,
      "mean_token_accuracy": 0.2017536774277687,
      "num_tokens": 43215661.0,
      "step": 17050
    },
    {
      "entropy": 6.263839673995972,
      "epoch": 1.968263127524524,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004621053037710918,
      "loss": 5.7718,
      "mean_token_accuracy": 0.1978829026222229,
      "num_tokens": 43227869.0,
      "step": 17055
    },
    {
      "entropy": 6.212932157516479,
      "epoch": 1.9688401615695326,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000462082382143793,
      "loss": 5.7522,
      "mean_token_accuracy": 0.19169737547636032,
      "num_tokens": 43240620.0,
      "step": 17060
    },
    {
      "entropy": 6.267064094543457,
      "epoch": 1.9694171956145412,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046205945422407113,
      "loss": 5.7077,
      "mean_token_accuracy": 0.20007965266704558,
      "num_tokens": 43252523.0,
      "step": 17065
    },
    {
      "entropy": 6.239709568023682,
      "epoch": 1.9699942296595498,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046203652001269754,
      "loss": 5.7608,
      "mean_token_accuracy": 0.19888664931058883,
      "num_tokens": 43265731.0,
      "step": 17070
    },
    {
      "entropy": 6.0771897315979,
      "epoch": 1.9705712637045587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046201357951044337,
      "loss": 5.6286,
      "mean_token_accuracy": 0.20163885056972503,
      "num_tokens": 43277929.0,
      "step": 17075
    },
    {
      "entropy": 6.259761714935303,
      "epoch": 1.9711482977495671,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046199063271808047,
      "loss": 5.7795,
      "mean_token_accuracy": 0.19385852813720703,
      "num_tokens": 43290011.0,
      "step": 17080
    },
    {
      "entropy": 6.267812061309814,
      "epoch": 1.971725331794576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004619676796363804,
      "loss": 5.7805,
      "mean_token_accuracy": 0.18994874656200408,
      "num_tokens": 43303205.0,
      "step": 17085
    },
    {
      "entropy": 6.118640041351318,
      "epoch": 1.9723023658395844,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046194472026611546,
      "loss": 5.595,
      "mean_token_accuracy": 0.21487024575471877,
      "num_tokens": 43316214.0,
      "step": 17090
    },
    {
      "entropy": 6.183839654922485,
      "epoch": 1.9728793998845933,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004619217546080576,
      "loss": 5.705,
      "mean_token_accuracy": 0.20014588087797164,
      "num_tokens": 43328497.0,
      "step": 17095
    },
    {
      "entropy": 6.28698616027832,
      "epoch": 1.9734564339296017,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004618987826629794,
      "loss": 5.7787,
      "mean_token_accuracy": 0.19396940469741822,
      "num_tokens": 43341081.0,
      "step": 17100
    },
    {
      "entropy": 6.30215311050415,
      "epoch": 1.9740334679746105,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046187580443165344,
      "loss": 5.8302,
      "mean_token_accuracy": 0.18925997167825698,
      "num_tokens": 43354001.0,
      "step": 17105
    },
    {
      "entropy": 6.111357879638672,
      "epoch": 1.9746105020196192,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004618528199148527,
      "loss": 5.5762,
      "mean_token_accuracy": 0.2125083476305008,
      "num_tokens": 43368925.0,
      "step": 17110
    },
    {
      "entropy": 6.164032506942749,
      "epoch": 1.9751875360646278,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046182982911335016,
      "loss": 5.6539,
      "mean_token_accuracy": 0.20521567463874818,
      "num_tokens": 43382063.0,
      "step": 17115
    },
    {
      "entropy": 6.279041814804077,
      "epoch": 1.9757645701096365,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046180683202791905,
      "loss": 5.744,
      "mean_token_accuracy": 0.195831798017025,
      "num_tokens": 43393810.0,
      "step": 17120
    },
    {
      "entropy": 6.1894340991973875,
      "epoch": 1.976341604154645,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000461783828659333,
      "loss": 5.7489,
      "mean_token_accuracy": 0.1993277132511139,
      "num_tokens": 43406500.0,
      "step": 17125
    },
    {
      "entropy": 6.2373778343200685,
      "epoch": 1.9769186381996537,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046176081900836565,
      "loss": 5.7589,
      "mean_token_accuracy": 0.19228785634040832,
      "num_tokens": 43418284.0,
      "step": 17130
    },
    {
      "entropy": 6.219095516204834,
      "epoch": 1.9774956722446624,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046173780307579086,
      "loss": 5.7332,
      "mean_token_accuracy": 0.20107158571481704,
      "num_tokens": 43430913.0,
      "step": 17135
    },
    {
      "entropy": 6.198169422149658,
      "epoch": 1.9780727062896712,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004617147808623829,
      "loss": 5.6841,
      "mean_token_accuracy": 0.2023420214653015,
      "num_tokens": 43444637.0,
      "step": 17140
    },
    {
      "entropy": 6.225779438018799,
      "epoch": 1.9786497403346797,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046169175236891605,
      "loss": 5.7327,
      "mean_token_accuracy": 0.20258403420448304,
      "num_tokens": 43457725.0,
      "step": 17145
    },
    {
      "entropy": 6.300029516220093,
      "epoch": 1.9792267743796885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004616687175961648,
      "loss": 5.7845,
      "mean_token_accuracy": 0.19073337614536284,
      "num_tokens": 43469902.0,
      "step": 17150
    },
    {
      "entropy": 6.264739608764648,
      "epoch": 1.979803808424697,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004616456765449039,
      "loss": 5.7537,
      "mean_token_accuracy": 0.18771969825029372,
      "num_tokens": 43482959.0,
      "step": 17155
    },
    {
      "entropy": 6.206252193450927,
      "epoch": 1.9803808424697058,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046162262921590845,
      "loss": 5.7811,
      "mean_token_accuracy": 0.19611617028713227,
      "num_tokens": 43496051.0,
      "step": 17160
    },
    {
      "entropy": 6.289826488494873,
      "epoch": 1.9809578765147142,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004615995756099535,
      "loss": 5.7525,
      "mean_token_accuracy": 0.20111227929592132,
      "num_tokens": 43509031.0,
      "step": 17165
    },
    {
      "entropy": 6.230772876739502,
      "epoch": 1.981534910559723,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004615765157278146,
      "loss": 5.65,
      "mean_token_accuracy": 0.20207782983779907,
      "num_tokens": 43521577.0,
      "step": 17170
    },
    {
      "entropy": 6.286557340621949,
      "epoch": 1.9821119446047317,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046155344957026726,
      "loss": 5.7764,
      "mean_token_accuracy": 0.19273026585578917,
      "num_tokens": 43536029.0,
      "step": 17175
    },
    {
      "entropy": 6.176950407028198,
      "epoch": 1.9826889786497404,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004615303771380873,
      "loss": 5.6941,
      "mean_token_accuracy": 0.19330438673496247,
      "num_tokens": 43548663.0,
      "step": 17180
    },
    {
      "entropy": 6.2743995666503904,
      "epoch": 1.983266012694749,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046150729843205077,
      "loss": 5.7812,
      "mean_token_accuracy": 0.19549938142299653,
      "num_tokens": 43560302.0,
      "step": 17185
    },
    {
      "entropy": 6.241931533813476,
      "epoch": 1.9838430467397576,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046148421345293384,
      "loss": 5.7318,
      "mean_token_accuracy": 0.19888069480657578,
      "num_tokens": 43572657.0,
      "step": 17190
    },
    {
      "entropy": 6.236357593536377,
      "epoch": 1.9844200807847663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004614611222015131,
      "loss": 5.8108,
      "mean_token_accuracy": 0.1903862848877907,
      "num_tokens": 43584973.0,
      "step": 17195
    },
    {
      "entropy": 6.31862244606018,
      "epoch": 1.984997114829775,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046143802467856507,
      "loss": 5.8048,
      "mean_token_accuracy": 0.1940244749188423,
      "num_tokens": 43597457.0,
      "step": 17200
    },
    {
      "entropy": 6.268091821670533,
      "epoch": 1.9855741488747836,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046141492088486673,
      "loss": 5.7435,
      "mean_token_accuracy": 0.19540723264217377,
      "num_tokens": 43610256.0,
      "step": 17205
    },
    {
      "entropy": 6.309213304519654,
      "epoch": 1.9861511829197922,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004613918108211951,
      "loss": 5.7369,
      "mean_token_accuracy": 0.19972355365753175,
      "num_tokens": 43622765.0,
      "step": 17210
    },
    {
      "entropy": 6.218349552154541,
      "epoch": 1.986728216964801,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004613686944883275,
      "loss": 5.7438,
      "mean_token_accuracy": 0.19382604509592055,
      "num_tokens": 43634353.0,
      "step": 17215
    },
    {
      "entropy": 6.2674988269805905,
      "epoch": 1.9873052510098095,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046134557188704146,
      "loss": 5.7579,
      "mean_token_accuracy": 0.196334633231163,
      "num_tokens": 43647778.0,
      "step": 17220
    },
    {
      "entropy": 6.241354846954346,
      "epoch": 1.9878822850548183,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046132244301811466,
      "loss": 5.7469,
      "mean_token_accuracy": 0.19919382482767106,
      "num_tokens": 43659759.0,
      "step": 17225
    },
    {
      "entropy": 6.2110288619995115,
      "epoch": 1.9884593190998268,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046129930788232497,
      "loss": 5.6806,
      "mean_token_accuracy": 0.19585457742214202,
      "num_tokens": 43673223.0,
      "step": 17230
    },
    {
      "entropy": 6.2823878765106205,
      "epoch": 1.9890363531448356,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046127616648045073,
      "loss": 5.7574,
      "mean_token_accuracy": 0.19816111922264099,
      "num_tokens": 43684657.0,
      "step": 17235
    },
    {
      "entropy": 6.271449613571167,
      "epoch": 1.989613387189844,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046125301881327007,
      "loss": 5.828,
      "mean_token_accuracy": 0.1909281387925148,
      "num_tokens": 43697583.0,
      "step": 17240
    },
    {
      "entropy": 6.196461534500122,
      "epoch": 1.990190421234853,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046122986488156167,
      "loss": 5.6856,
      "mean_token_accuracy": 0.2068835034966469,
      "num_tokens": 43711175.0,
      "step": 17245
    },
    {
      "entropy": 6.169440698623657,
      "epoch": 1.9907674552798615,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046120670468610426,
      "loss": 5.6936,
      "mean_token_accuracy": 0.20736287236213685,
      "num_tokens": 43724595.0,
      "step": 17250
    },
    {
      "entropy": 6.2320794582366945,
      "epoch": 1.9913444893248702,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046118353822767683,
      "loss": 5.6756,
      "mean_token_accuracy": 0.21006713062524796,
      "num_tokens": 43737192.0,
      "step": 17255
    },
    {
      "entropy": 6.214487457275391,
      "epoch": 1.9919215233698788,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004611603655070586,
      "loss": 5.7221,
      "mean_token_accuracy": 0.20066475570201875,
      "num_tokens": 43749543.0,
      "step": 17260
    },
    {
      "entropy": 6.204158592224121,
      "epoch": 1.9924985574148875,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046113718652502896,
      "loss": 5.7577,
      "mean_token_accuracy": 0.1970704585313797,
      "num_tokens": 43762754.0,
      "step": 17265
    },
    {
      "entropy": 6.291552257537842,
      "epoch": 1.993075591459896,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004611140012823675,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1978034183382988,
      "num_tokens": 43774489.0,
      "step": 17270
    },
    {
      "entropy": 6.240046882629395,
      "epoch": 1.9936526255049047,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046109080977985395,
      "loss": 5.7203,
      "mean_token_accuracy": 0.19953835755586624,
      "num_tokens": 43786519.0,
      "step": 17275
    },
    {
      "entropy": 6.093880462646484,
      "epoch": 1.9942296595499136,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046106761201826854,
      "loss": 5.6953,
      "mean_token_accuracy": 0.20718604624271392,
      "num_tokens": 43800359.0,
      "step": 17280
    },
    {
      "entropy": 6.153894138336182,
      "epoch": 1.994806693594922,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046104440799839145,
      "loss": 5.6467,
      "mean_token_accuracy": 0.20157096534967422,
      "num_tokens": 43812130.0,
      "step": 17285
    },
    {
      "entropy": 6.202395725250244,
      "epoch": 1.9953837276399309,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000461021197721003,
      "loss": 5.6824,
      "mean_token_accuracy": 0.20786392986774443,
      "num_tokens": 43823544.0,
      "step": 17290
    },
    {
      "entropy": 6.32968578338623,
      "epoch": 1.9959607616849393,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046099798118688407,
      "loss": 5.7894,
      "mean_token_accuracy": 0.19760541766881942,
      "num_tokens": 43836283.0,
      "step": 17295
    },
    {
      "entropy": 6.229362726211548,
      "epoch": 1.9965377957299482,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004609747583968154,
      "loss": 5.7852,
      "mean_token_accuracy": 0.193734011054039,
      "num_tokens": 43849077.0,
      "step": 17300
    },
    {
      "entropy": 6.271792459487915,
      "epoch": 1.9971148297749566,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004609515293515781,
      "loss": 5.7774,
      "mean_token_accuracy": 0.19543831646442414,
      "num_tokens": 43862805.0,
      "step": 17305
    },
    {
      "entropy": 6.303545761108398,
      "epoch": 1.9976918638199654,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004609282940519535,
      "loss": 5.7654,
      "mean_token_accuracy": 0.19415387511253357,
      "num_tokens": 43875420.0,
      "step": 17310
    },
    {
      "entropy": 6.254008197784424,
      "epoch": 1.998268897864974,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004609050524987231,
      "loss": 5.7875,
      "mean_token_accuracy": 0.189268496632576,
      "num_tokens": 43888445.0,
      "step": 17315
    },
    {
      "entropy": 6.233069515228271,
      "epoch": 1.9988459319099827,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004608818046926686,
      "loss": 5.8142,
      "mean_token_accuracy": 0.1950289487838745,
      "num_tokens": 43900943.0,
      "step": 17320
    },
    {
      "entropy": 6.210719203948974,
      "epoch": 1.9994229659549914,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004608585506345719,
      "loss": 5.7597,
      "mean_token_accuracy": 0.20078433454036712,
      "num_tokens": 43913951.0,
      "step": 17325
    },
    {
      "entropy": 6.241820812225342,
      "epoch": 2.0,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004608352903252152,
      "loss": 5.7103,
      "mean_token_accuracy": 0.20115090012550355,
      "num_tokens": 43926234.0,
      "step": 17330
    },
    {
      "entropy": 6.155938911437988,
      "epoch": 2.000577034045009,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046081202376538084,
      "loss": 5.6062,
      "mean_token_accuracy": 0.2129983901977539,
      "num_tokens": 43939519.0,
      "step": 17335
    },
    {
      "entropy": 6.186594152450562,
      "epoch": 2.0011540680900173,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004607887509558513,
      "loss": 5.6362,
      "mean_token_accuracy": 0.2054543301463127,
      "num_tokens": 43953209.0,
      "step": 17340
    },
    {
      "entropy": 6.23337082862854,
      "epoch": 2.001731102135026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004607654718974094,
      "loss": 5.6116,
      "mean_token_accuracy": 0.20926240235567092,
      "num_tokens": 43965578.0,
      "step": 17345
    },
    {
      "entropy": 6.2014381885528564,
      "epoch": 2.0023081361800346,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004607421865908382,
      "loss": 5.6198,
      "mean_token_accuracy": 0.20223724395036696,
      "num_tokens": 43978508.0,
      "step": 17350
    },
    {
      "entropy": 6.253764247894287,
      "epoch": 2.0028851702250434,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004607188950369207,
      "loss": 5.7173,
      "mean_token_accuracy": 0.19965245872735976,
      "num_tokens": 43992266.0,
      "step": 17355
    },
    {
      "entropy": 6.262997150421143,
      "epoch": 2.003462204270052,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004606955972364405,
      "loss": 5.676,
      "mean_token_accuracy": 0.19863019436597823,
      "num_tokens": 44004603.0,
      "step": 17360
    },
    {
      "entropy": 6.254293298721313,
      "epoch": 2.0040392383150607,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004606722931901812,
      "loss": 5.6507,
      "mean_token_accuracy": 0.1994758814573288,
      "num_tokens": 44017578.0,
      "step": 17365
    },
    {
      "entropy": 6.210956048965454,
      "epoch": 2.004616272360069,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004606489828989264,
      "loss": 5.6666,
      "mean_token_accuracy": 0.2062393993139267,
      "num_tokens": 44031745.0,
      "step": 17370
    },
    {
      "entropy": 6.23971152305603,
      "epoch": 2.005193306405078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004606256663634604,
      "loss": 5.7204,
      "mean_token_accuracy": 0.19800833612680435,
      "num_tokens": 44043027.0,
      "step": 17375
    },
    {
      "entropy": 6.2065952777862545,
      "epoch": 2.0057703404500864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004606023435845672,
      "loss": 5.6308,
      "mean_token_accuracy": 0.2112123966217041,
      "num_tokens": 44056211.0,
      "step": 17380
    },
    {
      "entropy": 6.128069686889648,
      "epoch": 2.0063473744950953,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004605790145630314,
      "loss": 5.5406,
      "mean_token_accuracy": 0.21449308842420578,
      "num_tokens": 44069861.0,
      "step": 17385
    },
    {
      "entropy": 6.292887544631958,
      "epoch": 2.0069244085401037,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004605556792996377,
      "loss": 5.703,
      "mean_token_accuracy": 0.1974819242954254,
      "num_tokens": 44083373.0,
      "step": 17390
    },
    {
      "entropy": 6.155897426605224,
      "epoch": 2.0075014425851125,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004605323377951709,
      "loss": 5.5914,
      "mean_token_accuracy": 0.20667856186628342,
      "num_tokens": 44096138.0,
      "step": 17395
    },
    {
      "entropy": 6.197297620773315,
      "epoch": 2.008078476630121,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000460508990050416,
      "loss": 5.6447,
      "mean_token_accuracy": 0.19865068793296814,
      "num_tokens": 44107591.0,
      "step": 17400
    },
    {
      "entropy": 6.245768070220947,
      "epoch": 2.00865551067513,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004604856360661584,
      "loss": 5.645,
      "mean_token_accuracy": 0.20221610218286515,
      "num_tokens": 44119733.0,
      "step": 17405
    },
    {
      "entropy": 6.312962770462036,
      "epoch": 2.0092325447201387,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004604622758431835,
      "loss": 5.7533,
      "mean_token_accuracy": 0.19784328639507293,
      "num_tokens": 44131967.0,
      "step": 17410
    },
    {
      "entropy": 6.226692390441895,
      "epoch": 2.009809578765147,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046043890938227724,
      "loss": 5.6886,
      "mean_token_accuracy": 0.202842678129673,
      "num_tokens": 44144426.0,
      "step": 17415
    },
    {
      "entropy": 6.222259902954102,
      "epoch": 2.010386612810156,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046041553668422526,
      "loss": 5.7247,
      "mean_token_accuracy": 0.19307588040828705,
      "num_tokens": 44156298.0,
      "step": 17420
    },
    {
      "entropy": 6.202986001968384,
      "epoch": 2.0109636468551644,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046039215774981376,
      "loss": 5.6133,
      "mean_token_accuracy": 0.20076507031917573,
      "num_tokens": 44167955.0,
      "step": 17425
    },
    {
      "entropy": 6.236855745315552,
      "epoch": 2.0115406809001732,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046036877257982917,
      "loss": 5.6622,
      "mean_token_accuracy": 0.20241550952196122,
      "num_tokens": 44180555.0,
      "step": 17430
    },
    {
      "entropy": 6.184055376052856,
      "epoch": 2.0121177149451817,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000460345381175058,
      "loss": 5.6619,
      "mean_token_accuracy": 0.2000580132007599,
      "num_tokens": 44194335.0,
      "step": 17435
    },
    {
      "entropy": 6.253341913223267,
      "epoch": 2.0126947489901905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000460321983536287,
      "loss": 5.6521,
      "mean_token_accuracy": 0.20659084767103195,
      "num_tokens": 44207258.0,
      "step": 17440
    },
    {
      "entropy": 6.285744667053223,
      "epoch": 2.013271783035199,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046029857966430315,
      "loss": 5.6693,
      "mean_token_accuracy": 0.20483236461877824,
      "num_tokens": 44220424.0,
      "step": 17445
    },
    {
      "entropy": 6.145317506790161,
      "epoch": 2.013848817080208,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046027516955989364,
      "loss": 5.6097,
      "mean_token_accuracy": 0.2023573562502861,
      "num_tokens": 44233394.0,
      "step": 17450
    },
    {
      "entropy": 6.281408739089966,
      "epoch": 2.014425851125216,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046025175322384577,
      "loss": 5.6546,
      "mean_token_accuracy": 0.19403218626976013,
      "num_tokens": 44246255.0,
      "step": 17455
    },
    {
      "entropy": 6.259836101531983,
      "epoch": 2.015002885170225,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046022833065694727,
      "loss": 5.6518,
      "mean_token_accuracy": 0.19955547899007797,
      "num_tokens": 44258315.0,
      "step": 17460
    },
    {
      "entropy": 6.235680866241455,
      "epoch": 2.0155799192152335,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046020490185998575,
      "loss": 5.7546,
      "mean_token_accuracy": 0.19254150241613388,
      "num_tokens": 44271667.0,
      "step": 17465
    },
    {
      "entropy": 6.3066315174102785,
      "epoch": 2.0161569532602424,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004601814668337495,
      "loss": 5.6808,
      "mean_token_accuracy": 0.20038435161113738,
      "num_tokens": 44284831.0,
      "step": 17470
    },
    {
      "entropy": 6.21392035484314,
      "epoch": 2.016733987305251,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046015802557902654,
      "loss": 5.6753,
      "mean_token_accuracy": 0.20335264056921004,
      "num_tokens": 44297161.0,
      "step": 17475
    },
    {
      "entropy": 6.265213680267334,
      "epoch": 2.0173110213502596,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046013457809660537,
      "loss": 5.6421,
      "mean_token_accuracy": 0.20377608984708787,
      "num_tokens": 44309537.0,
      "step": 17480
    },
    {
      "entropy": 6.177024602890015,
      "epoch": 2.0178880553952685,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046011112438727454,
      "loss": 5.6294,
      "mean_token_accuracy": 0.21052780151367187,
      "num_tokens": 44320761.0,
      "step": 17485
    },
    {
      "entropy": 6.187991380691528,
      "epoch": 2.018465089440277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004600876644518231,
      "loss": 5.5466,
      "mean_token_accuracy": 0.2109512582421303,
      "num_tokens": 44333323.0,
      "step": 17490
    },
    {
      "entropy": 6.099303674697876,
      "epoch": 2.0190421234852858,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046006419829104,
      "loss": 5.4833,
      "mean_token_accuracy": 0.21557884514331818,
      "num_tokens": 44345919.0,
      "step": 17495
    },
    {
      "entropy": 6.23404335975647,
      "epoch": 2.019619157530294,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004600407259057145,
      "loss": 5.699,
      "mean_token_accuracy": 0.20145083218812943,
      "num_tokens": 44359702.0,
      "step": 17500
    },
    {
      "entropy": 6.164773797988891,
      "epoch": 2.020196191575303,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004600172472966361,
      "loss": 5.5994,
      "mean_token_accuracy": 0.20531328916549682,
      "num_tokens": 44374186.0,
      "step": 17505
    },
    {
      "entropy": 6.272966957092285,
      "epoch": 2.0207732256203115,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00045999376246459446,
      "loss": 5.7512,
      "mean_token_accuracy": 0.19523853659629822,
      "num_tokens": 44387870.0,
      "step": 17510
    },
    {
      "entropy": 6.251287126541138,
      "epoch": 2.0213502596653203,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004599702714103795,
      "loss": 5.6835,
      "mean_token_accuracy": 0.19467540830373764,
      "num_tokens": 44400706.0,
      "step": 17515
    },
    {
      "entropy": 6.274653720855713,
      "epoch": 2.0219272937103288,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004599467741347814,
      "loss": 5.6843,
      "mean_token_accuracy": 0.19732244163751603,
      "num_tokens": 44413166.0,
      "step": 17520
    },
    {
      "entropy": 6.259235906600952,
      "epoch": 2.0225043277553376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004599232706385904,
      "loss": 5.6911,
      "mean_token_accuracy": 0.20414672046899796,
      "num_tokens": 44425265.0,
      "step": 17525
    },
    {
      "entropy": 6.2070684909820555,
      "epoch": 2.023081361800346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000459899760922597,
      "loss": 5.6112,
      "mean_token_accuracy": 0.20335202366113664,
      "num_tokens": 44436990.0,
      "step": 17530
    },
    {
      "entropy": 6.254583263397217,
      "epoch": 2.023658395845355,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004598762449875921,
      "loss": 5.6502,
      "mean_token_accuracy": 0.20472092777490616,
      "num_tokens": 44450503.0,
      "step": 17535
    },
    {
      "entropy": 6.274939250946045,
      "epoch": 2.0242354298903633,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004598527228343665,
      "loss": 5.697,
      "mean_token_accuracy": 0.19625728875398635,
      "num_tokens": 44463559.0,
      "step": 17540
    },
    {
      "entropy": 6.273628997802734,
      "epoch": 2.024812463935372,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004598291944637112,
      "loss": 5.7402,
      "mean_token_accuracy": 0.19684628993272782,
      "num_tokens": 44475064.0,
      "step": 17545
    },
    {
      "entropy": 6.2887876510620115,
      "epoch": 2.025389497980381,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045980565987641797,
      "loss": 5.6639,
      "mean_token_accuracy": 0.2009371191263199,
      "num_tokens": 44487094.0,
      "step": 17550
    },
    {
      "entropy": 6.262852191925049,
      "epoch": 2.0259665320253895,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045978211907327804,
      "loss": 5.6959,
      "mean_token_accuracy": 0.19581854492425918,
      "num_tokens": 44498950.0,
      "step": 17555
    },
    {
      "entropy": 6.22926607131958,
      "epoch": 2.0265435660703983,
      "grad_norm": 1.0,
      "learning_rate": 0.0004597585720550834,
      "loss": 5.6676,
      "mean_token_accuracy": 0.20075897723436356,
      "num_tokens": 44511277.0,
      "step": 17560
    },
    {
      "entropy": 6.173655700683594,
      "epoch": 2.0271206001154067,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000459735018822626,
      "loss": 5.6426,
      "mean_token_accuracy": 0.20742131769657135,
      "num_tokens": 44525595.0,
      "step": 17565
    },
    {
      "entropy": 6.272472858428955,
      "epoch": 2.0276976341604156,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045971145937669794,
      "loss": 5.6906,
      "mean_token_accuracy": 0.2004339799284935,
      "num_tokens": 44537470.0,
      "step": 17570
    },
    {
      "entropy": 6.307222652435303,
      "epoch": 2.028274668205424,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004596878937180917,
      "loss": 5.6756,
      "mean_token_accuracy": 0.19540244936943055,
      "num_tokens": 44550594.0,
      "step": 17575
    },
    {
      "entropy": 6.265674877166748,
      "epoch": 2.028851702250433,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004596643218475999,
      "loss": 5.6786,
      "mean_token_accuracy": 0.19477606266736985,
      "num_tokens": 44563204.0,
      "step": 17580
    },
    {
      "entropy": 6.221870183944702,
      "epoch": 2.0294287362954413,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045964074376601534,
      "loss": 5.6943,
      "mean_token_accuracy": 0.19705665111541748,
      "num_tokens": 44577377.0,
      "step": 17585
    },
    {
      "entropy": 6.2684094429016115,
      "epoch": 2.03000577034045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045961715947413106,
      "loss": 5.6935,
      "mean_token_accuracy": 0.20129784494638442,
      "num_tokens": 44589658.0,
      "step": 17590
    },
    {
      "entropy": 6.163085222244263,
      "epoch": 2.0305828043854586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004595935689727404,
      "loss": 5.6868,
      "mean_token_accuracy": 0.2033739924430847,
      "num_tokens": 44602477.0,
      "step": 17595
    },
    {
      "entropy": 6.281161975860596,
      "epoch": 2.0311598384304674,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004595699722626367,
      "loss": 5.6751,
      "mean_token_accuracy": 0.1965423047542572,
      "num_tokens": 44613734.0,
      "step": 17600
    },
    {
      "entropy": 6.339623880386353,
      "epoch": 2.031736872475476,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045954636934461367,
      "loss": 5.7766,
      "mean_token_accuracy": 0.19242643266916276,
      "num_tokens": 44626608.0,
      "step": 17605
    },
    {
      "entropy": 6.127445840835572,
      "epoch": 2.0323139065204847,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004595227602194652,
      "loss": 5.547,
      "mean_token_accuracy": 0.21846822649240494,
      "num_tokens": 44640982.0,
      "step": 17610
    },
    {
      "entropy": 6.189198398590088,
      "epoch": 2.0328909405654936,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004594991448879853,
      "loss": 5.6817,
      "mean_token_accuracy": 0.20118973553180694,
      "num_tokens": 44653546.0,
      "step": 17615
    },
    {
      "entropy": 6.206907892227173,
      "epoch": 2.033467974610502,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004594755233509683,
      "loss": 5.6131,
      "mean_token_accuracy": 0.20262075662612916,
      "num_tokens": 44667555.0,
      "step": 17620
    },
    {
      "entropy": 6.15839958190918,
      "epoch": 2.034045008655511,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045945189560920875,
      "loss": 5.5396,
      "mean_token_accuracy": 0.20959313213825226,
      "num_tokens": 44680205.0,
      "step": 17625
    },
    {
      "entropy": 6.273732662200928,
      "epoch": 2.0346220427005193,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004594282616635013,
      "loss": 5.6721,
      "mean_token_accuracy": 0.19900794178247452,
      "num_tokens": 44692838.0,
      "step": 17630
    },
    {
      "entropy": 6.235873889923096,
      "epoch": 2.035199076745528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004594046215146409,
      "loss": 5.6294,
      "mean_token_accuracy": 0.2072541669011116,
      "num_tokens": 44706010.0,
      "step": 17635
    },
    {
      "entropy": 6.180500555038452,
      "epoch": 2.0357761107905366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045938097516342257,
      "loss": 5.6333,
      "mean_token_accuracy": 0.2143148437142372,
      "num_tokens": 44718857.0,
      "step": 17640
    },
    {
      "entropy": 6.253361225128174,
      "epoch": 2.0363531448355454,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045935732261064184,
      "loss": 5.6436,
      "mean_token_accuracy": 0.2046870470046997,
      "num_tokens": 44730876.0,
      "step": 17645
    },
    {
      "entropy": 6.155157232284546,
      "epoch": 2.036930178880554,
      "grad_norm": 1.0,
      "learning_rate": 0.00045933366385709405,
      "loss": 5.6029,
      "mean_token_accuracy": 0.2016918659210205,
      "num_tokens": 44743236.0,
      "step": 17650
    },
    {
      "entropy": 6.130135631561279,
      "epoch": 2.0375072129255627,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004593099989035751,
      "loss": 5.579,
      "mean_token_accuracy": 0.21147882491350173,
      "num_tokens": 44755762.0,
      "step": 17655
    },
    {
      "entropy": 6.316247892379761,
      "epoch": 2.038084246970571,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004592863277508809,
      "loss": 5.7116,
      "mean_token_accuracy": 0.19873490333557128,
      "num_tokens": 44768439.0,
      "step": 17660
    },
    {
      "entropy": 6.2354902744293215,
      "epoch": 2.03866128101558,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004592626503998076,
      "loss": 5.6826,
      "mean_token_accuracy": 0.1987271189689636,
      "num_tokens": 44781751.0,
      "step": 17665
    },
    {
      "entropy": 6.141412353515625,
      "epoch": 2.0392383150605884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004592389668511515,
      "loss": 5.5403,
      "mean_token_accuracy": 0.205668169260025,
      "num_tokens": 44794221.0,
      "step": 17670
    },
    {
      "entropy": 6.281309366226196,
      "epoch": 2.0398153491055973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004592152771057093,
      "loss": 5.6685,
      "mean_token_accuracy": 0.20133744329214096,
      "num_tokens": 44807027.0,
      "step": 17675
    },
    {
      "entropy": 6.167712211608887,
      "epoch": 2.0403923831506057,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045919158116427785,
      "loss": 5.6766,
      "mean_token_accuracy": 0.20474430918693542,
      "num_tokens": 44820399.0,
      "step": 17680
    },
    {
      "entropy": 6.292107629776001,
      "epoch": 2.0409694171956145,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045916787902765396,
      "loss": 5.767,
      "mean_token_accuracy": 0.18854909092187883,
      "num_tokens": 44833179.0,
      "step": 17685
    },
    {
      "entropy": 6.211412382125855,
      "epoch": 2.0415464512406234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004591441706966349,
      "loss": 5.5655,
      "mean_token_accuracy": 0.2210731253027916,
      "num_tokens": 44846169.0,
      "step": 17690
    },
    {
      "entropy": 6.109423685073852,
      "epoch": 2.042123485285632,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004591204561720183,
      "loss": 5.511,
      "mean_token_accuracy": 0.20630994588136672,
      "num_tokens": 44858400.0,
      "step": 17695
    },
    {
      "entropy": 6.221507835388183,
      "epoch": 2.0427005193306407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004590967354546015,
      "loss": 5.7366,
      "mean_token_accuracy": 0.1998888745903969,
      "num_tokens": 44870291.0,
      "step": 17700
    },
    {
      "entropy": 6.32669882774353,
      "epoch": 2.043277553375649,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004590730085451824,
      "loss": 5.7005,
      "mean_token_accuracy": 0.20282694101333618,
      "num_tokens": 44883893.0,
      "step": 17705
    },
    {
      "entropy": 6.22974419593811,
      "epoch": 2.043854587420658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045904927544455914,
      "loss": 5.5991,
      "mean_token_accuracy": 0.20218637734651565,
      "num_tokens": 44896075.0,
      "step": 17710
    },
    {
      "entropy": 6.12322301864624,
      "epoch": 2.0444316214656664,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045902553615353005,
      "loss": 5.584,
      "mean_token_accuracy": 0.20640041679143906,
      "num_tokens": 44908752.0,
      "step": 17715
    },
    {
      "entropy": 6.312625122070313,
      "epoch": 2.0450086555106752,
      "grad_norm": 0.875,
      "learning_rate": 0.0004590017906728933,
      "loss": 5.6668,
      "mean_token_accuracy": 0.2084189996123314,
      "num_tokens": 44920782.0,
      "step": 17720
    },
    {
      "entropy": 6.188219499588013,
      "epoch": 2.0455856895556837,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045897803900344774,
      "loss": 5.5748,
      "mean_token_accuracy": 0.20670071095228196,
      "num_tokens": 44934565.0,
      "step": 17725
    },
    {
      "entropy": 6.154708814620972,
      "epoch": 2.0461627236006925,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004589542811459923,
      "loss": 5.5965,
      "mean_token_accuracy": 0.21104211509227752,
      "num_tokens": 44948232.0,
      "step": 17730
    },
    {
      "entropy": 6.261731672286987,
      "epoch": 2.046739757645701,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004589305171013259,
      "loss": 5.6244,
      "mean_token_accuracy": 0.20190979689359664,
      "num_tokens": 44960092.0,
      "step": 17735
    },
    {
      "entropy": 6.319637632369995,
      "epoch": 2.04731679169071,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000458906746870248,
      "loss": 5.7451,
      "mean_token_accuracy": 0.19333603233098984,
      "num_tokens": 44972065.0,
      "step": 17740
    },
    {
      "entropy": 6.241170454025268,
      "epoch": 2.047893825735718,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000458882970453558,
      "loss": 5.6873,
      "mean_token_accuracy": 0.21004485934972764,
      "num_tokens": 44985142.0,
      "step": 17745
    },
    {
      "entropy": 6.251197671890258,
      "epoch": 2.048470859780727,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004588591878520556,
      "loss": 5.7335,
      "mean_token_accuracy": 0.19356470704078674,
      "num_tokens": 44998131.0,
      "step": 17750
    },
    {
      "entropy": 6.193807935714721,
      "epoch": 2.049047893825736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004588353990665407,
      "loss": 5.5746,
      "mean_token_accuracy": 0.2091870754957199,
      "num_tokens": 45012715.0,
      "step": 17755
    },
    {
      "entropy": 6.257633352279663,
      "epoch": 2.0496249278707444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004588116040978136,
      "loss": 5.7328,
      "mean_token_accuracy": 0.1942736968398094,
      "num_tokens": 45025214.0,
      "step": 17760
    },
    {
      "entropy": 6.1944324493408205,
      "epoch": 2.050201961915753,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045878780294667437,
      "loss": 5.5597,
      "mean_token_accuracy": 0.2213586837053299,
      "num_tokens": 45037956.0,
      "step": 17765
    },
    {
      "entropy": 6.208847618103027,
      "epoch": 2.0507789959607616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004587639956139237,
      "loss": 5.7184,
      "mean_token_accuracy": 0.19553436189889908,
      "num_tokens": 45052074.0,
      "step": 17770
    },
    {
      "entropy": 6.219709634780884,
      "epoch": 2.0513560300057705,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004587401821003624,
      "loss": 5.644,
      "mean_token_accuracy": 0.19805888533592225,
      "num_tokens": 45064408.0,
      "step": 17775
    },
    {
      "entropy": 6.1760063648223875,
      "epoch": 2.051933064050779,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045871636240679133,
      "loss": 5.5629,
      "mean_token_accuracy": 0.2161258950829506,
      "num_tokens": 45077944.0,
      "step": 17780
    },
    {
      "entropy": 6.157691383361817,
      "epoch": 2.0525100980957878,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004586925365340117,
      "loss": 5.6168,
      "mean_token_accuracy": 0.20498253554105758,
      "num_tokens": 45090682.0,
      "step": 17785
    },
    {
      "entropy": 6.198795175552368,
      "epoch": 2.053087132140796,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004586687044828247,
      "loss": 5.6275,
      "mean_token_accuracy": 0.2042268231511116,
      "num_tokens": 45102567.0,
      "step": 17790
    },
    {
      "entropy": 6.158082008361816,
      "epoch": 2.053664166185805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004586448662540322,
      "loss": 5.5962,
      "mean_token_accuracy": 0.20935477763414384,
      "num_tokens": 45114585.0,
      "step": 17795
    },
    {
      "entropy": 6.27517032623291,
      "epoch": 2.0542412002308135,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004586210218484358,
      "loss": 5.7028,
      "mean_token_accuracy": 0.199089315533638,
      "num_tokens": 45128491.0,
      "step": 17800
    },
    {
      "entropy": 6.214361238479614,
      "epoch": 2.0548182342758223,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045859717126683745,
      "loss": 5.7209,
      "mean_token_accuracy": 0.19616287797689438,
      "num_tokens": 45141646.0,
      "step": 17805
    },
    {
      "entropy": 6.1508544921875,
      "epoch": 2.0553952683208307,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045857331451003953,
      "loss": 5.5566,
      "mean_token_accuracy": 0.21111010313034057,
      "num_tokens": 45153797.0,
      "step": 17810
    },
    {
      "entropy": 6.183637237548828,
      "epoch": 2.0559723023658396,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045854945157884435,
      "loss": 5.6696,
      "mean_token_accuracy": 0.20431995689868926,
      "num_tokens": 45166626.0,
      "step": 17815
    },
    {
      "entropy": 6.200638723373413,
      "epoch": 2.056549336410848,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045852558247405455,
      "loss": 5.6339,
      "mean_token_accuracy": 0.20421143025159835,
      "num_tokens": 45181974.0,
      "step": 17820
    },
    {
      "entropy": 6.32881007194519,
      "epoch": 2.057126370455857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045850170719647287,
      "loss": 5.7287,
      "mean_token_accuracy": 0.19897275567054748,
      "num_tokens": 45194406.0,
      "step": 17825
    },
    {
      "entropy": 6.195233964920044,
      "epoch": 2.0577034045008658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045847782574690254,
      "loss": 5.6237,
      "mean_token_accuracy": 0.20583680272102356,
      "num_tokens": 45206777.0,
      "step": 17830
    },
    {
      "entropy": 6.160016250610352,
      "epoch": 2.058280438545874,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045845393812614664,
      "loss": 5.6188,
      "mean_token_accuracy": 0.20355250239372252,
      "num_tokens": 45219718.0,
      "step": 17835
    },
    {
      "entropy": 6.2939612865448,
      "epoch": 2.058857472590883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004584300443350086,
      "loss": 5.679,
      "mean_token_accuracy": 0.20269724428653718,
      "num_tokens": 45231038.0,
      "step": 17840
    },
    {
      "entropy": 6.221446418762207,
      "epoch": 2.0594345066358914,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004584061443742922,
      "loss": 5.7275,
      "mean_token_accuracy": 0.19740950912237168,
      "num_tokens": 45242835.0,
      "step": 17845
    },
    {
      "entropy": 6.253048229217529,
      "epoch": 2.0600115406809003,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045838223824480124,
      "loss": 5.7082,
      "mean_token_accuracy": 0.1985023394227028,
      "num_tokens": 45254857.0,
      "step": 17850
    },
    {
      "entropy": 6.1964263916015625,
      "epoch": 2.0605885747259087,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004583583259473397,
      "loss": 5.6521,
      "mean_token_accuracy": 0.20834221690893173,
      "num_tokens": 45267897.0,
      "step": 17855
    },
    {
      "entropy": 6.214391469955444,
      "epoch": 2.0611656087709176,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045833440748271203,
      "loss": 5.6625,
      "mean_token_accuracy": 0.21355814188718797,
      "num_tokens": 45280209.0,
      "step": 17860
    },
    {
      "entropy": 6.21958966255188,
      "epoch": 2.061742642815926,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045831048285172266,
      "loss": 5.6569,
      "mean_token_accuracy": 0.20677920132875444,
      "num_tokens": 45291834.0,
      "step": 17865
    },
    {
      "entropy": 6.278505563735962,
      "epoch": 2.062319676860935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004582865520551762,
      "loss": 5.7545,
      "mean_token_accuracy": 0.19581874012947081,
      "num_tokens": 45305057.0,
      "step": 17870
    },
    {
      "entropy": 6.1761250495910645,
      "epoch": 2.0628967109059433,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045826261509387755,
      "loss": 5.6204,
      "mean_token_accuracy": 0.20212469696998597,
      "num_tokens": 45318212.0,
      "step": 17875
    },
    {
      "entropy": 6.3260908126831055,
      "epoch": 2.063473744950952,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045823867196863197,
      "loss": 5.7534,
      "mean_token_accuracy": 0.19136424660682677,
      "num_tokens": 45330981.0,
      "step": 17880
    },
    {
      "entropy": 6.161239862442017,
      "epoch": 2.0640507789959606,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004582147226802446,
      "loss": 5.59,
      "mean_token_accuracy": 0.20502999126911164,
      "num_tokens": 45344404.0,
      "step": 17885
    },
    {
      "entropy": 6.1044426441192625,
      "epoch": 2.0646278130409694,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004581907672295211,
      "loss": 5.6203,
      "mean_token_accuracy": 0.21126181334257127,
      "num_tokens": 45357708.0,
      "step": 17890
    },
    {
      "entropy": 6.083464002609253,
      "epoch": 2.065204847085978,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045816680561726716,
      "loss": 5.5307,
      "mean_token_accuracy": 0.21751691550016403,
      "num_tokens": 45370625.0,
      "step": 17895
    },
    {
      "entropy": 6.267791080474853,
      "epoch": 2.0657818811309867,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004581428378442886,
      "loss": 5.6142,
      "mean_token_accuracy": 0.2051943302154541,
      "num_tokens": 45381864.0,
      "step": 17900
    },
    {
      "entropy": 6.221700620651245,
      "epoch": 2.0663589151759956,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00045811886391139173,
      "loss": 5.6732,
      "mean_token_accuracy": 0.19928453415632247,
      "num_tokens": 45394624.0,
      "step": 17905
    },
    {
      "entropy": 6.1344873905181885,
      "epoch": 2.066935949221004,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045809488381938284,
      "loss": 5.5327,
      "mean_token_accuracy": 0.21862836182117462,
      "num_tokens": 45408551.0,
      "step": 17910
    },
    {
      "entropy": 6.297921419143677,
      "epoch": 2.067512983266013,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004580708975690684,
      "loss": 5.7122,
      "mean_token_accuracy": 0.19909276366233825,
      "num_tokens": 45420432.0,
      "step": 17915
    },
    {
      "entropy": 6.321915578842163,
      "epoch": 2.0680900173110213,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004580469051612554,
      "loss": 5.8181,
      "mean_token_accuracy": 0.18725140541791915,
      "num_tokens": 45432914.0,
      "step": 17920
    },
    {
      "entropy": 6.224918508529663,
      "epoch": 2.06866705135603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045802290659675057,
      "loss": 5.6254,
      "mean_token_accuracy": 0.2068016141653061,
      "num_tokens": 45445501.0,
      "step": 17925
    },
    {
      "entropy": 6.128627586364746,
      "epoch": 2.0692440854010385,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045799890187636123,
      "loss": 5.5594,
      "mean_token_accuracy": 0.2027588188648224,
      "num_tokens": 45458920.0,
      "step": 17930
    },
    {
      "entropy": 6.315451335906983,
      "epoch": 2.0698211194460474,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045797489100089464,
      "loss": 5.7671,
      "mean_token_accuracy": 0.19484665244817734,
      "num_tokens": 45471438.0,
      "step": 17935
    },
    {
      "entropy": 6.245334959030151,
      "epoch": 2.070398153491056,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004579508739711585,
      "loss": 5.6362,
      "mean_token_accuracy": 0.20459053069353103,
      "num_tokens": 45483956.0,
      "step": 17940
    },
    {
      "entropy": 6.2623810291290285,
      "epoch": 2.0709751875360647,
      "grad_norm": 1.0,
      "learning_rate": 0.00045792685078796075,
      "loss": 5.7391,
      "mean_token_accuracy": 0.1916007474064827,
      "num_tokens": 45496499.0,
      "step": 17945
    },
    {
      "entropy": 6.2876989364624025,
      "epoch": 2.071552221581073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000457902821452109,
      "loss": 5.6473,
      "mean_token_accuracy": 0.2057687446475029,
      "num_tokens": 45509069.0,
      "step": 17950
    },
    {
      "entropy": 6.318817281723023,
      "epoch": 2.072129255626082,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045787878596441193,
      "loss": 5.781,
      "mean_token_accuracy": 0.19031604528427123,
      "num_tokens": 45521707.0,
      "step": 17955
    },
    {
      "entropy": 6.165262126922608,
      "epoch": 2.0727062896710904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004578547443256776,
      "loss": 5.5912,
      "mean_token_accuracy": 0.21902539432048798,
      "num_tokens": 45533618.0,
      "step": 17960
    },
    {
      "entropy": 6.189864587783814,
      "epoch": 2.0732833237160992,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004578306965367148,
      "loss": 5.6456,
      "mean_token_accuracy": 0.210878786444664,
      "num_tokens": 45546487.0,
      "step": 17965
    },
    {
      "entropy": 6.218348979949951,
      "epoch": 2.073860357761108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045780664259833235,
      "loss": 5.6663,
      "mean_token_accuracy": 0.2008408173918724,
      "num_tokens": 45560258.0,
      "step": 17970
    },
    {
      "entropy": 6.3292594909667965,
      "epoch": 2.0744373918061165,
      "grad_norm": 1.0,
      "learning_rate": 0.00045778258251133924,
      "loss": 5.7244,
      "mean_token_accuracy": 0.19564768075942993,
      "num_tokens": 45573152.0,
      "step": 17975
    },
    {
      "entropy": 6.337434244155884,
      "epoch": 2.0750144258511254,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045775851627654474,
      "loss": 5.7305,
      "mean_token_accuracy": 0.19415022879838945,
      "num_tokens": 45586042.0,
      "step": 17980
    },
    {
      "entropy": 6.327477693557739,
      "epoch": 2.075591459896134,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004577344438947584,
      "loss": 5.7815,
      "mean_token_accuracy": 0.19349619150161743,
      "num_tokens": 45598761.0,
      "step": 17985
    },
    {
      "entropy": 6.215632438659668,
      "epoch": 2.0761684939411427,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045771036536678984,
      "loss": 5.6165,
      "mean_token_accuracy": 0.205550654232502,
      "num_tokens": 45610988.0,
      "step": 17990
    },
    {
      "entropy": 6.202718114852905,
      "epoch": 2.076745527986151,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00045768628069344885,
      "loss": 5.6077,
      "mean_token_accuracy": 0.2161363333463669,
      "num_tokens": 45623689.0,
      "step": 17995
    },
    {
      "entropy": 6.255098104476929,
      "epoch": 2.07732256203116,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004576621898755455,
      "loss": 5.6891,
      "mean_token_accuracy": 0.2092033326625824,
      "num_tokens": 45637283.0,
      "step": 18000
    },
    {
      "epoch": 2.07732256203116,
      "eval_entropy": 6.061120188368427,
      "eval_loss": 5.789821147918701,
      "eval_mean_token_accuracy": 0.2038638797594737,
      "eval_num_tokens": 45637283.0,
      "eval_runtime": 17.5495,
      "eval_samples_per_second": 1603.233,
      "eval_steps_per_second": 200.404,
      "step": 18000
    },
    {
      "entropy": 6.271256685256958,
      "epoch": 2.0778995960761684,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045763809291389024,
      "loss": 5.7231,
      "mean_token_accuracy": 0.19220100492238998,
      "num_tokens": 45651139.0,
      "step": 18005
    },
    {
      "entropy": 6.281512689590454,
      "epoch": 2.0784766301211772,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004576139898092933,
      "loss": 5.6888,
      "mean_token_accuracy": 0.20105655640363693,
      "num_tokens": 45663696.0,
      "step": 18010
    },
    {
      "entropy": 6.263245868682861,
      "epoch": 2.0790536641661856,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004575898805625657,
      "loss": 5.6764,
      "mean_token_accuracy": 0.20521681159734725,
      "num_tokens": 45676272.0,
      "step": 18015
    },
    {
      "entropy": 6.234762716293335,
      "epoch": 2.0796306982111945,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045756576517451804,
      "loss": 5.6834,
      "mean_token_accuracy": 0.20475297272205353,
      "num_tokens": 45688031.0,
      "step": 18020
    },
    {
      "entropy": 6.275010824203491,
      "epoch": 2.080207732256203,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045754164364596156,
      "loss": 5.7356,
      "mean_token_accuracy": 0.19875845760107042,
      "num_tokens": 45700286.0,
      "step": 18025
    },
    {
      "entropy": 6.231034851074218,
      "epoch": 2.080784766301212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004575175159777076,
      "loss": 5.6489,
      "mean_token_accuracy": 0.20674524307250977,
      "num_tokens": 45713935.0,
      "step": 18030
    },
    {
      "entropy": 6.215792417526245,
      "epoch": 2.0813618003462206,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004574933821705676,
      "loss": 5.6528,
      "mean_token_accuracy": 0.1962132826447487,
      "num_tokens": 45726815.0,
      "step": 18035
    },
    {
      "entropy": 6.230903005599975,
      "epoch": 2.081938834391229,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004574692422253534,
      "loss": 5.6225,
      "mean_token_accuracy": 0.2043220043182373,
      "num_tokens": 45738727.0,
      "step": 18040
    },
    {
      "entropy": 6.221548175811767,
      "epoch": 2.082515868436238,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045744509614287687,
      "loss": 5.6577,
      "mean_token_accuracy": 0.20010620653629302,
      "num_tokens": 45751221.0,
      "step": 18045
    },
    {
      "entropy": 6.275927448272705,
      "epoch": 2.0830929024812463,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004574209439239501,
      "loss": 5.6424,
      "mean_token_accuracy": 0.20071506649255752,
      "num_tokens": 45764680.0,
      "step": 18050
    },
    {
      "entropy": 6.251017332077026,
      "epoch": 2.083669936526255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045739678556938563,
      "loss": 5.6505,
      "mean_token_accuracy": 0.19949812889099122,
      "num_tokens": 45778324.0,
      "step": 18055
    },
    {
      "entropy": 6.177297067642212,
      "epoch": 2.0842469705712636,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045737262107999575,
      "loss": 5.5744,
      "mean_token_accuracy": 0.20188492238521577,
      "num_tokens": 45789412.0,
      "step": 18060
    },
    {
      "entropy": 6.270459794998169,
      "epoch": 2.0848240046162725,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004573484504565934,
      "loss": 5.7941,
      "mean_token_accuracy": 0.19221320003271103,
      "num_tokens": 45802264.0,
      "step": 18065
    },
    {
      "entropy": 6.272039175033569,
      "epoch": 2.085401038661281,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004573242736999915,
      "loss": 5.6458,
      "mean_token_accuracy": 0.2073688417673111,
      "num_tokens": 45815411.0,
      "step": 18070
    },
    {
      "entropy": 6.309269189834595,
      "epoch": 2.0859780727062898,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045730009081100314,
      "loss": 5.8004,
      "mean_token_accuracy": 0.19237782508134843,
      "num_tokens": 45828001.0,
      "step": 18075
    },
    {
      "entropy": 6.25848274230957,
      "epoch": 2.086555106751298,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045727590179044195,
      "loss": 5.7069,
      "mean_token_accuracy": 0.19830467849969863,
      "num_tokens": 45841001.0,
      "step": 18080
    },
    {
      "entropy": 6.272468519210816,
      "epoch": 2.087132140796307,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004572517066391211,
      "loss": 5.726,
      "mean_token_accuracy": 0.1950818806886673,
      "num_tokens": 45855160.0,
      "step": 18085
    },
    {
      "entropy": 6.1138811111450195,
      "epoch": 2.0877091748413155,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045722750535785484,
      "loss": 5.5177,
      "mean_token_accuracy": 0.2084098443388939,
      "num_tokens": 45867822.0,
      "step": 18090
    },
    {
      "entropy": 6.305677652359009,
      "epoch": 2.0882862088863243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045720329794745685,
      "loss": 5.734,
      "mean_token_accuracy": 0.1883520469069481,
      "num_tokens": 45878990.0,
      "step": 18095
    },
    {
      "entropy": 6.269919300079346,
      "epoch": 2.0888632429313327,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004571790844087415,
      "loss": 5.6603,
      "mean_token_accuracy": 0.20223036855459214,
      "num_tokens": 45891411.0,
      "step": 18100
    },
    {
      "entropy": 6.21648211479187,
      "epoch": 2.0894402769763416,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004571548647425231,
      "loss": 5.6139,
      "mean_token_accuracy": 0.20800613164901732,
      "num_tokens": 45903589.0,
      "step": 18105
    },
    {
      "entropy": 6.215238428115844,
      "epoch": 2.0900173110213505,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004571306389496164,
      "loss": 5.6214,
      "mean_token_accuracy": 0.20546383112668992,
      "num_tokens": 45915454.0,
      "step": 18110
    },
    {
      "entropy": 6.24914813041687,
      "epoch": 2.090594345066359,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045710640703083594,
      "loss": 5.628,
      "mean_token_accuracy": 0.20552445948123932,
      "num_tokens": 45928974.0,
      "step": 18115
    },
    {
      "entropy": 6.292775392532349,
      "epoch": 2.0911713791113677,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045708216898699707,
      "loss": 5.6497,
      "mean_token_accuracy": 0.20035794526338577,
      "num_tokens": 45941040.0,
      "step": 18120
    },
    {
      "entropy": 6.258070278167724,
      "epoch": 2.091748413156376,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045705792481891483,
      "loss": 5.6825,
      "mean_token_accuracy": 0.2039056807756424,
      "num_tokens": 45954760.0,
      "step": 18125
    },
    {
      "entropy": 6.240582704544067,
      "epoch": 2.092325447201385,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045703367452740477,
      "loss": 5.7393,
      "mean_token_accuracy": 0.2046695962548256,
      "num_tokens": 45967735.0,
      "step": 18130
    },
    {
      "entropy": 6.318571758270264,
      "epoch": 2.0929024812463934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045700941811328247,
      "loss": 5.7815,
      "mean_token_accuracy": 0.1897900477051735,
      "num_tokens": 45979318.0,
      "step": 18135
    },
    {
      "entropy": 6.168387460708618,
      "epoch": 2.0934795152914023,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045698515557736374,
      "loss": 5.6664,
      "mean_token_accuracy": 0.2009122595191002,
      "num_tokens": 45992487.0,
      "step": 18140
    },
    {
      "entropy": 6.3405050277709964,
      "epoch": 2.0940565493364107,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045696088692046477,
      "loss": 5.7137,
      "mean_token_accuracy": 0.19912787675857543,
      "num_tokens": 46004103.0,
      "step": 18145
    },
    {
      "entropy": 6.327506828308105,
      "epoch": 2.0946335833814196,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045693661214340174,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1949118822813034,
      "num_tokens": 46016498.0,
      "step": 18150
    },
    {
      "entropy": 6.216175842285156,
      "epoch": 2.095210617426428,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045691233124699117,
      "loss": 5.6833,
      "mean_token_accuracy": 0.2000325307250023,
      "num_tokens": 46028179.0,
      "step": 18155
    },
    {
      "entropy": 6.232275342941284,
      "epoch": 2.095787651471437,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004568880442320497,
      "loss": 5.6576,
      "mean_token_accuracy": 0.19335602968931198,
      "num_tokens": 46040152.0,
      "step": 18160
    },
    {
      "entropy": 6.243351840972901,
      "epoch": 2.0963646855164453,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045686375109939417,
      "loss": 5.6916,
      "mean_token_accuracy": 0.2010435089468956,
      "num_tokens": 46052663.0,
      "step": 18165
    },
    {
      "entropy": 6.147784948348999,
      "epoch": 2.096941719561454,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004568394518498417,
      "loss": 5.6448,
      "mean_token_accuracy": 0.20705157816410064,
      "num_tokens": 46066793.0,
      "step": 18170
    },
    {
      "entropy": 6.240902853012085,
      "epoch": 2.0975187536064626,
      "grad_norm": 0.859375,
      "learning_rate": 0.00045681514648420965,
      "loss": 5.693,
      "mean_token_accuracy": 0.2024614155292511,
      "num_tokens": 46080913.0,
      "step": 18175
    },
    {
      "entropy": 6.275468397140503,
      "epoch": 2.0980957876514714,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004567908350033154,
      "loss": 5.705,
      "mean_token_accuracy": 0.19357870817184447,
      "num_tokens": 46093344.0,
      "step": 18180
    },
    {
      "entropy": 6.243587875366211,
      "epoch": 2.0986728216964803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004567665174079767,
      "loss": 5.6188,
      "mean_token_accuracy": 0.2064504310488701,
      "num_tokens": 46106967.0,
      "step": 18185
    },
    {
      "entropy": 6.2748072147369385,
      "epoch": 2.0992498557414887,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045674219369901153,
      "loss": 5.6654,
      "mean_token_accuracy": 0.20344507098197936,
      "num_tokens": 46118564.0,
      "step": 18190
    },
    {
      "entropy": 6.313588333129883,
      "epoch": 2.0998268897864976,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004567178638772379,
      "loss": 5.7706,
      "mean_token_accuracy": 0.18828553706407547,
      "num_tokens": 46130057.0,
      "step": 18195
    },
    {
      "entropy": 6.184890365600586,
      "epoch": 2.100403923831506,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004566935279434742,
      "loss": 5.6006,
      "mean_token_accuracy": 0.21139217913150787,
      "num_tokens": 46142730.0,
      "step": 18200
    },
    {
      "entropy": 6.265673589706421,
      "epoch": 2.100980957876515,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045666918589853895,
      "loss": 5.6718,
      "mean_token_accuracy": 0.20348992496728896,
      "num_tokens": 46153669.0,
      "step": 18205
    },
    {
      "entropy": 6.212975645065308,
      "epoch": 2.1015579919215233,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045664483774325094,
      "loss": 5.677,
      "mean_token_accuracy": 0.2022017002105713,
      "num_tokens": 46166472.0,
      "step": 18210
    },
    {
      "entropy": 6.25945553779602,
      "epoch": 2.102135025966532,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004566204834784289,
      "loss": 5.7078,
      "mean_token_accuracy": 0.19803351461887359,
      "num_tokens": 46178417.0,
      "step": 18215
    },
    {
      "entropy": 6.21232533454895,
      "epoch": 2.1027120600115405,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00045659612310489225,
      "loss": 5.6247,
      "mean_token_accuracy": 0.20136779248714448,
      "num_tokens": 46191537.0,
      "step": 18220
    },
    {
      "entropy": 6.1611487865448,
      "epoch": 2.1032890940565494,
      "grad_norm": 4.90625,
      "learning_rate": 0.00045657175662346014,
      "loss": 5.5483,
      "mean_token_accuracy": 0.20845748037099837,
      "num_tokens": 46205716.0,
      "step": 18225
    },
    {
      "entropy": 6.300152015686035,
      "epoch": 2.103866128101558,
      "grad_norm": 0.875,
      "learning_rate": 0.0004565473840349522,
      "loss": 5.7055,
      "mean_token_accuracy": 0.1970550909638405,
      "num_tokens": 46219050.0,
      "step": 18230
    },
    {
      "entropy": 6.290921354293824,
      "epoch": 2.1044431621465667,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045652300534018816,
      "loss": 5.6968,
      "mean_token_accuracy": 0.20341092795133592,
      "num_tokens": 46231045.0,
      "step": 18235
    },
    {
      "entropy": 6.33854489326477,
      "epoch": 2.105020196191575,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000456498620539988,
      "loss": 5.7969,
      "mean_token_accuracy": 0.19549209624528885,
      "num_tokens": 46244629.0,
      "step": 18240
    },
    {
      "entropy": 6.290593385696411,
      "epoch": 2.105597230236584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004564742296351719,
      "loss": 5.6947,
      "mean_token_accuracy": 0.2022603139281273,
      "num_tokens": 46256960.0,
      "step": 18245
    },
    {
      "entropy": 6.182830715179444,
      "epoch": 2.106174264281593,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045644983262656014,
      "loss": 5.6175,
      "mean_token_accuracy": 0.2079408958554268,
      "num_tokens": 46268558.0,
      "step": 18250
    },
    {
      "entropy": 6.250482368469238,
      "epoch": 2.1067512983266012,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004564254295149735,
      "loss": 5.6682,
      "mean_token_accuracy": 0.20333527773618698,
      "num_tokens": 46281529.0,
      "step": 18255
    },
    {
      "entropy": 6.220935535430908,
      "epoch": 2.10732833237161,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045640102030123264,
      "loss": 5.6632,
      "mean_token_accuracy": 0.20304493010044097,
      "num_tokens": 46293598.0,
      "step": 18260
    },
    {
      "entropy": 6.2396129131317135,
      "epoch": 2.1079053664166185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045637660498615865,
      "loss": 5.6284,
      "mean_token_accuracy": 0.2031083658337593,
      "num_tokens": 46304252.0,
      "step": 18265
    },
    {
      "entropy": 6.2506492137908936,
      "epoch": 2.1084824004616274,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004563521835705725,
      "loss": 5.7081,
      "mean_token_accuracy": 0.20356660783290864,
      "num_tokens": 46316295.0,
      "step": 18270
    },
    {
      "entropy": 6.204360580444336,
      "epoch": 2.109059434506636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045632775605529587,
      "loss": 5.6344,
      "mean_token_accuracy": 0.19993555545806885,
      "num_tokens": 46328929.0,
      "step": 18275
    },
    {
      "entropy": 6.294415426254273,
      "epoch": 2.1096364685516447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004563033224411501,
      "loss": 5.7137,
      "mean_token_accuracy": 0.19647678434848787,
      "num_tokens": 46340784.0,
      "step": 18280
    },
    {
      "entropy": 6.261148262023926,
      "epoch": 2.110213502596653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004562788827289572,
      "loss": 5.6481,
      "mean_token_accuracy": 0.20099495947360993,
      "num_tokens": 46354188.0,
      "step": 18285
    },
    {
      "entropy": 6.1608076095581055,
      "epoch": 2.110790536641662,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045625443691953914,
      "loss": 5.5593,
      "mean_token_accuracy": 0.21348249912261963,
      "num_tokens": 46366076.0,
      "step": 18290
    },
    {
      "entropy": 6.219788026809693,
      "epoch": 2.1113675706866704,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004562299850137181,
      "loss": 5.6035,
      "mean_token_accuracy": 0.20397830456495286,
      "num_tokens": 46376954.0,
      "step": 18295
    },
    {
      "entropy": 6.236584854125977,
      "epoch": 2.111944604731679,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004562055270123166,
      "loss": 5.6365,
      "mean_token_accuracy": 0.19959541410207748,
      "num_tokens": 46387951.0,
      "step": 18300
    },
    {
      "entropy": 6.222868156433106,
      "epoch": 2.1125216387766876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045618106291615715,
      "loss": 5.7218,
      "mean_token_accuracy": 0.20118409246206284,
      "num_tokens": 46399889.0,
      "step": 18305
    },
    {
      "entropy": 6.194531965255737,
      "epoch": 2.1130986728216965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004561565927260627,
      "loss": 5.7268,
      "mean_token_accuracy": 0.19808163344860077,
      "num_tokens": 46412420.0,
      "step": 18310
    },
    {
      "entropy": 6.247423267364502,
      "epoch": 2.1136757068667054,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004561321164428561,
      "loss": 5.7022,
      "mean_token_accuracy": 0.19544601291418076,
      "num_tokens": 46424758.0,
      "step": 18315
    },
    {
      "entropy": 6.21748685836792,
      "epoch": 2.114252740911714,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004561076340673609,
      "loss": 5.7238,
      "mean_token_accuracy": 0.19610183238983153,
      "num_tokens": 46437163.0,
      "step": 18320
    },
    {
      "entropy": 6.239074563980102,
      "epoch": 2.1148297749567226,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004560831456004003,
      "loss": 5.6065,
      "mean_token_accuracy": 0.20686981379985808,
      "num_tokens": 46450780.0,
      "step": 18325
    },
    {
      "entropy": 6.250164604187011,
      "epoch": 2.115406809001731,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004560586510427981,
      "loss": 5.6773,
      "mean_token_accuracy": 0.20282908231019975,
      "num_tokens": 46463879.0,
      "step": 18330
    },
    {
      "entropy": 6.224255990982056,
      "epoch": 2.11598384304674,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004560341503953781,
      "loss": 5.6954,
      "mean_token_accuracy": 0.1984913483262062,
      "num_tokens": 46477624.0,
      "step": 18335
    },
    {
      "entropy": 6.2712701797485355,
      "epoch": 2.1165608770917483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004560096436589643,
      "loss": 5.6821,
      "mean_token_accuracy": 0.1961820214986801,
      "num_tokens": 46489745.0,
      "step": 18340
    },
    {
      "entropy": 6.257243824005127,
      "epoch": 2.117137911136757,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045598513083438115,
      "loss": 5.6336,
      "mean_token_accuracy": 0.20528088212013246,
      "num_tokens": 46501429.0,
      "step": 18345
    },
    {
      "entropy": 6.2360915184021,
      "epoch": 2.1177149451817656,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045596061192245297,
      "loss": 5.75,
      "mean_token_accuracy": 0.19197311997413635,
      "num_tokens": 46515094.0,
      "step": 18350
    },
    {
      "entropy": 6.226460266113281,
      "epoch": 2.1182919792267745,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004559360869240045,
      "loss": 5.6397,
      "mean_token_accuracy": 0.20181107372045518,
      "num_tokens": 46527659.0,
      "step": 18355
    },
    {
      "entropy": 6.202995872497558,
      "epoch": 2.118869013271783,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004559115558398606,
      "loss": 5.6128,
      "mean_token_accuracy": 0.201771542429924,
      "num_tokens": 46540190.0,
      "step": 18360
    },
    {
      "entropy": 6.337617874145508,
      "epoch": 2.1194460473167918,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004558870186708465,
      "loss": 5.7314,
      "mean_token_accuracy": 0.19478038102388381,
      "num_tokens": 46552349.0,
      "step": 18365
    },
    {
      "entropy": 6.285080814361573,
      "epoch": 2.1200230813618,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045586247541778724,
      "loss": 5.7142,
      "mean_token_accuracy": 0.1966187462210655,
      "num_tokens": 46564693.0,
      "step": 18370
    },
    {
      "entropy": 6.13067364692688,
      "epoch": 2.120600115406809,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004558379260815085,
      "loss": 5.601,
      "mean_token_accuracy": 0.2058517187833786,
      "num_tokens": 46577035.0,
      "step": 18375
    },
    {
      "entropy": 6.207986783981323,
      "epoch": 2.1211771494518175,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004558133706628359,
      "loss": 5.6556,
      "mean_token_accuracy": 0.19956648498773574,
      "num_tokens": 46589670.0,
      "step": 18380
    },
    {
      "entropy": 6.339266967773438,
      "epoch": 2.1217541834968263,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045578880916259554,
      "loss": 5.7049,
      "mean_token_accuracy": 0.19442971497774125,
      "num_tokens": 46601320.0,
      "step": 18385
    },
    {
      "entropy": 6.247182321548462,
      "epoch": 2.122331217541835,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004557642415816132,
      "loss": 5.6679,
      "mean_token_accuracy": 0.2010486125946045,
      "num_tokens": 46615065.0,
      "step": 18390
    },
    {
      "entropy": 6.160168170928955,
      "epoch": 2.1229082515868436,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004557396679207155,
      "loss": 5.6144,
      "mean_token_accuracy": 0.2045021042227745,
      "num_tokens": 46628847.0,
      "step": 18395
    },
    {
      "entropy": 6.237725067138672,
      "epoch": 2.1234852856318525,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045571508818072887,
      "loss": 5.6154,
      "mean_token_accuracy": 0.20226229578256608,
      "num_tokens": 46642214.0,
      "step": 18400
    },
    {
      "entropy": 6.289147472381591,
      "epoch": 2.124062319676861,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004556905023624799,
      "loss": 5.664,
      "mean_token_accuracy": 0.20270753502845765,
      "num_tokens": 46655920.0,
      "step": 18405
    },
    {
      "entropy": 6.26250958442688,
      "epoch": 2.1246393537218697,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045566591046679577,
      "loss": 5.7294,
      "mean_token_accuracy": 0.19266002327203752,
      "num_tokens": 46668006.0,
      "step": 18410
    },
    {
      "entropy": 6.149018621444702,
      "epoch": 2.125216387766878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045564131249450343,
      "loss": 5.5514,
      "mean_token_accuracy": 0.21318518817424775,
      "num_tokens": 46681523.0,
      "step": 18415
    },
    {
      "entropy": 6.265885972976685,
      "epoch": 2.125793421811887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004556167084464302,
      "loss": 5.6601,
      "mean_token_accuracy": 0.20284263044595718,
      "num_tokens": 46693904.0,
      "step": 18420
    },
    {
      "entropy": 6.207275247573852,
      "epoch": 2.1263704558568954,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004555920983234038,
      "loss": 5.6725,
      "mean_token_accuracy": 0.19513762444257737,
      "num_tokens": 46705210.0,
      "step": 18425
    },
    {
      "entropy": 6.205043745040894,
      "epoch": 2.1269474899019043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045556748212625183,
      "loss": 5.6281,
      "mean_token_accuracy": 0.2072788342833519,
      "num_tokens": 46717594.0,
      "step": 18430
    },
    {
      "entropy": 6.313850831985474,
      "epoch": 2.1275245239469127,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004555428598558023,
      "loss": 5.7962,
      "mean_token_accuracy": 0.1880889430642128,
      "num_tokens": 46730363.0,
      "step": 18435
    },
    {
      "entropy": 6.167889070510864,
      "epoch": 2.1281015579919216,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004555182315128833,
      "loss": 5.6695,
      "mean_token_accuracy": 0.2026648074388504,
      "num_tokens": 46742554.0,
      "step": 18440
    },
    {
      "entropy": 6.264118432998657,
      "epoch": 2.12867859203693,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004554935970983234,
      "loss": 5.7368,
      "mean_token_accuracy": 0.20292575657367706,
      "num_tokens": 46754907.0,
      "step": 18445
    },
    {
      "entropy": 6.273162460327148,
      "epoch": 2.129255626081939,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004554689566129509,
      "loss": 5.709,
      "mean_token_accuracy": 0.19691937267780305,
      "num_tokens": 46767463.0,
      "step": 18450
    },
    {
      "entropy": 6.293329524993896,
      "epoch": 2.1298326601269473,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045544431005759467,
      "loss": 5.7368,
      "mean_token_accuracy": 0.19693097323179246,
      "num_tokens": 46781523.0,
      "step": 18455
    },
    {
      "entropy": 6.260634040832519,
      "epoch": 2.130409694171956,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00045541965743308376,
      "loss": 5.7082,
      "mean_token_accuracy": 0.20587607026100158,
      "num_tokens": 46796557.0,
      "step": 18460
    },
    {
      "entropy": 6.272769212722778,
      "epoch": 2.130986728216965,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004553949987402473,
      "loss": 5.6985,
      "mean_token_accuracy": 0.19824606776237488,
      "num_tokens": 46809142.0,
      "step": 18465
    },
    {
      "entropy": 6.278567361831665,
      "epoch": 2.1315637622619734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004553703339799146,
      "loss": 5.7389,
      "mean_token_accuracy": 0.19780928790569305,
      "num_tokens": 46821654.0,
      "step": 18470
    },
    {
      "entropy": 6.185671472549439,
      "epoch": 2.1321407963069823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004553456631529154,
      "loss": 5.6149,
      "mean_token_accuracy": 0.2034536361694336,
      "num_tokens": 46833262.0,
      "step": 18475
    },
    {
      "entropy": 6.299640560150147,
      "epoch": 2.1327178303519907,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004553209862600794,
      "loss": 5.7283,
      "mean_token_accuracy": 0.19782658964395522,
      "num_tokens": 46846194.0,
      "step": 18480
    },
    {
      "entropy": 6.172132205963135,
      "epoch": 2.1332948643969996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004552963033022365,
      "loss": 5.6077,
      "mean_token_accuracy": 0.20716050565242766,
      "num_tokens": 46858395.0,
      "step": 18485
    },
    {
      "entropy": 6.272927618026733,
      "epoch": 2.133871898442008,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045527161428021706,
      "loss": 5.6997,
      "mean_token_accuracy": 0.2083545967936516,
      "num_tokens": 46871609.0,
      "step": 18490
    },
    {
      "entropy": 6.293195915222168,
      "epoch": 2.134448932487017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004552469191948514,
      "loss": 5.6987,
      "mean_token_accuracy": 0.19334774166345597,
      "num_tokens": 46883177.0,
      "step": 18495
    },
    {
      "entropy": 6.272311735153198,
      "epoch": 2.1350259665320253,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004552222180469702,
      "loss": 5.584,
      "mean_token_accuracy": 0.20692434757947922,
      "num_tokens": 46895678.0,
      "step": 18500
    },
    {
      "entropy": 6.191584157943725,
      "epoch": 2.135603000577034,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00045519751083740413,
      "loss": 5.6407,
      "mean_token_accuracy": 0.19992784708738326,
      "num_tokens": 46909041.0,
      "step": 18505
    },
    {
      "entropy": 6.228591775894165,
      "epoch": 2.1361800346220425,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045517279756698435,
      "loss": 5.6018,
      "mean_token_accuracy": 0.2105460360646248,
      "num_tokens": 46921386.0,
      "step": 18510
    },
    {
      "entropy": 6.236222553253174,
      "epoch": 2.1367570686670514,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045514807823654203,
      "loss": 5.6401,
      "mean_token_accuracy": 0.20905629098415374,
      "num_tokens": 46933627.0,
      "step": 18515
    },
    {
      "entropy": 6.281707429885865,
      "epoch": 2.13733410271206,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004551233528469086,
      "loss": 5.7237,
      "mean_token_accuracy": 0.19393283128738403,
      "num_tokens": 46945108.0,
      "step": 18520
    },
    {
      "entropy": 6.234498453140259,
      "epoch": 2.1379111367570687,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045509862139891553,
      "loss": 5.6165,
      "mean_token_accuracy": 0.20893581062555314,
      "num_tokens": 46958366.0,
      "step": 18525
    },
    {
      "entropy": 6.242971467971802,
      "epoch": 2.138488170802077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045507388389339496,
      "loss": 5.626,
      "mean_token_accuracy": 0.20295826643705367,
      "num_tokens": 46972001.0,
      "step": 18530
    },
    {
      "entropy": 6.227196788787841,
      "epoch": 2.139065204847086,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045504914033117866,
      "loss": 5.6415,
      "mean_token_accuracy": 0.20377167612314223,
      "num_tokens": 46984365.0,
      "step": 18535
    },
    {
      "entropy": 6.212011671066284,
      "epoch": 2.139642238892095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045502439071309897,
      "loss": 5.5737,
      "mean_token_accuracy": 0.21961427181959153,
      "num_tokens": 46996838.0,
      "step": 18540
    },
    {
      "entropy": 6.25760440826416,
      "epoch": 2.1402192729371032,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045499963503998835,
      "loss": 5.7338,
      "mean_token_accuracy": 0.1998489499092102,
      "num_tokens": 47008979.0,
      "step": 18545
    },
    {
      "entropy": 6.2248228073120115,
      "epoch": 2.140796306982112,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004549748733126794,
      "loss": 5.6787,
      "mean_token_accuracy": 0.20151789635419845,
      "num_tokens": 47021092.0,
      "step": 18550
    },
    {
      "entropy": 6.147092342376709,
      "epoch": 2.1413733410271205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000454950105532005,
      "loss": 5.6052,
      "mean_token_accuracy": 0.20899777561426164,
      "num_tokens": 47034277.0,
      "step": 18555
    },
    {
      "entropy": 6.20290994644165,
      "epoch": 2.1419503750721294,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045492533169879816,
      "loss": 5.651,
      "mean_token_accuracy": 0.19967204332351685,
      "num_tokens": 47046783.0,
      "step": 18560
    },
    {
      "entropy": 6.211876773834229,
      "epoch": 2.142527409117138,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045490055181389216,
      "loss": 5.6726,
      "mean_token_accuracy": 0.20224044770002364,
      "num_tokens": 47059358.0,
      "step": 18565
    },
    {
      "entropy": 6.290266132354736,
      "epoch": 2.1431044431621467,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045487576587812046,
      "loss": 5.6694,
      "mean_token_accuracy": 0.19567059874534606,
      "num_tokens": 47071492.0,
      "step": 18570
    },
    {
      "entropy": 6.23031439781189,
      "epoch": 2.143681477207155,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045485097389231675,
      "loss": 5.6986,
      "mean_token_accuracy": 0.19965904504060744,
      "num_tokens": 47083196.0,
      "step": 18575
    },
    {
      "entropy": 6.277791690826416,
      "epoch": 2.144258511252164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004548261758573149,
      "loss": 5.7075,
      "mean_token_accuracy": 0.2062839761376381,
      "num_tokens": 47095193.0,
      "step": 18580
    },
    {
      "entropy": 6.2877357482910154,
      "epoch": 2.1448355452971724,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004548013717739489,
      "loss": 5.6913,
      "mean_token_accuracy": 0.2042230486869812,
      "num_tokens": 47107751.0,
      "step": 18585
    },
    {
      "entropy": 6.206104326248169,
      "epoch": 2.145412579342181,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004547765616430531,
      "loss": 5.6768,
      "mean_token_accuracy": 0.20299719423055648,
      "num_tokens": 47121336.0,
      "step": 18590
    },
    {
      "entropy": 6.20489730834961,
      "epoch": 2.14598961338719,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004547517454654619,
      "loss": 5.6586,
      "mean_token_accuracy": 0.19978562891483306,
      "num_tokens": 47135565.0,
      "step": 18595
    },
    {
      "entropy": 6.275424671173096,
      "epoch": 2.1465666474321985,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045472692324201005,
      "loss": 5.6812,
      "mean_token_accuracy": 0.20549759566783904,
      "num_tokens": 47147703.0,
      "step": 18600
    },
    {
      "entropy": 6.284392070770264,
      "epoch": 2.1471436814772074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045470209497353243,
      "loss": 5.6388,
      "mean_token_accuracy": 0.20480807572603227,
      "num_tokens": 47159309.0,
      "step": 18605
    },
    {
      "entropy": 6.25921664237976,
      "epoch": 2.1477207155222158,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004546772606608641,
      "loss": 5.7099,
      "mean_token_accuracy": 0.20396180599927902,
      "num_tokens": 47171275.0,
      "step": 18610
    },
    {
      "entropy": 6.19076099395752,
      "epoch": 2.1482977495672246,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004546524203048404,
      "loss": 5.6396,
      "mean_token_accuracy": 0.20729872435331345,
      "num_tokens": 47183556.0,
      "step": 18615
    },
    {
      "entropy": 6.201742553710938,
      "epoch": 2.148874783612233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004546275739062967,
      "loss": 5.6569,
      "mean_token_accuracy": 0.19873298853635787,
      "num_tokens": 47196863.0,
      "step": 18620
    },
    {
      "entropy": 6.286324119567871,
      "epoch": 2.149451817657242,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004546027214660688,
      "loss": 5.7267,
      "mean_token_accuracy": 0.1972798749804497,
      "num_tokens": 47208717.0,
      "step": 18625
    },
    {
      "entropy": 6.236415529251099,
      "epoch": 2.1500288517022503,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004545778629849926,
      "loss": 5.6972,
      "mean_token_accuracy": 0.20324379950761795,
      "num_tokens": 47222404.0,
      "step": 18630
    },
    {
      "entropy": 6.2396653175354,
      "epoch": 2.150605885747259,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004545529984639042,
      "loss": 5.6628,
      "mean_token_accuracy": 0.20132592767477037,
      "num_tokens": 47235005.0,
      "step": 18635
    },
    {
      "entropy": 6.252484321594238,
      "epoch": 2.1511829197922676,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004545281279036398,
      "loss": 5.7348,
      "mean_token_accuracy": 0.20025120228528975,
      "num_tokens": 47248591.0,
      "step": 18640
    },
    {
      "entropy": 6.22770848274231,
      "epoch": 2.1517599538372765,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004545032513050361,
      "loss": 5.6313,
      "mean_token_accuracy": 0.19892991930246354,
      "num_tokens": 47261231.0,
      "step": 18645
    },
    {
      "entropy": 6.261441040039062,
      "epoch": 2.152336987882285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004544783686689296,
      "loss": 5.6625,
      "mean_token_accuracy": 0.1971624195575714,
      "num_tokens": 47272891.0,
      "step": 18650
    },
    {
      "entropy": 6.271063661575317,
      "epoch": 2.1529140219272938,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045445347999615736,
      "loss": 5.707,
      "mean_token_accuracy": 0.19718798696994783,
      "num_tokens": 47284730.0,
      "step": 18655
    },
    {
      "entropy": 6.306783390045166,
      "epoch": 2.153491055972302,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045442858528755653,
      "loss": 5.661,
      "mean_token_accuracy": 0.19824289083480834,
      "num_tokens": 47297485.0,
      "step": 18660
    },
    {
      "entropy": 6.117976570129395,
      "epoch": 2.154068090017311,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045440368454396435,
      "loss": 5.5836,
      "mean_token_accuracy": 0.20566565692424774,
      "num_tokens": 47310925.0,
      "step": 18665
    },
    {
      "entropy": 6.215226793289185,
      "epoch": 2.15464512406232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004543787777662183,
      "loss": 5.66,
      "mean_token_accuracy": 0.19447725713253022,
      "num_tokens": 47322645.0,
      "step": 18670
    },
    {
      "entropy": 6.253636455535888,
      "epoch": 2.1552221581073283,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045435386495515617,
      "loss": 5.7144,
      "mean_token_accuracy": 0.20357380211353301,
      "num_tokens": 47334582.0,
      "step": 18675
    },
    {
      "entropy": 6.1429496765136715,
      "epoch": 2.155799192152337,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004543289461116159,
      "loss": 5.5992,
      "mean_token_accuracy": 0.20072960257530212,
      "num_tokens": 47347792.0,
      "step": 18680
    },
    {
      "entropy": 6.203429460525513,
      "epoch": 2.1563762261973456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004543040212364356,
      "loss": 5.6829,
      "mean_token_accuracy": 0.19770924001932144,
      "num_tokens": 47361064.0,
      "step": 18685
    },
    {
      "entropy": 6.29300742149353,
      "epoch": 2.1569532602423545,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004542790903304536,
      "loss": 5.6932,
      "mean_token_accuracy": 0.19712911546230316,
      "num_tokens": 47374329.0,
      "step": 18690
    },
    {
      "entropy": 6.226617622375488,
      "epoch": 2.157530294287363,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004542541533945085,
      "loss": 5.6857,
      "mean_token_accuracy": 0.20336962640285491,
      "num_tokens": 47387755.0,
      "step": 18695
    },
    {
      "entropy": 6.2593629360198975,
      "epoch": 2.1581073283323717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045422921042943885,
      "loss": 5.7034,
      "mean_token_accuracy": 0.1937658190727234,
      "num_tokens": 47400970.0,
      "step": 18700
    },
    {
      "entropy": 6.119820308685303,
      "epoch": 2.15868436237738,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000454204261436084,
      "loss": 5.5642,
      "mean_token_accuracy": 0.21209942549467087,
      "num_tokens": 47414091.0,
      "step": 18705
    },
    {
      "entropy": 6.145029067993164,
      "epoch": 2.159261396422389,
      "grad_norm": 1.0,
      "learning_rate": 0.00045417930641528255,
      "loss": 5.4413,
      "mean_token_accuracy": 0.2311820924282074,
      "num_tokens": 47428485.0,
      "step": 18710
    },
    {
      "entropy": 6.215566825866699,
      "epoch": 2.1598384304673974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045415434536787424,
      "loss": 5.7094,
      "mean_token_accuracy": 0.19985218942165375,
      "num_tokens": 47442474.0,
      "step": 18715
    },
    {
      "entropy": 6.189802265167236,
      "epoch": 2.1604154645124063,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004541293782946985,
      "loss": 5.6265,
      "mean_token_accuracy": 0.20974376201629638,
      "num_tokens": 47456286.0,
      "step": 18720
    },
    {
      "entropy": 6.194776391983032,
      "epoch": 2.1609924985574147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004541044051965952,
      "loss": 5.6079,
      "mean_token_accuracy": 0.2135043904185295,
      "num_tokens": 47468946.0,
      "step": 18725
    },
    {
      "entropy": 6.262034320831299,
      "epoch": 2.1615695326024236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004540794260744041,
      "loss": 5.5847,
      "mean_token_accuracy": 0.207892245054245,
      "num_tokens": 47481877.0,
      "step": 18730
    },
    {
      "entropy": 6.198784399032593,
      "epoch": 2.162146566647432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004540544409289655,
      "loss": 5.6426,
      "mean_token_accuracy": 0.21184716522693633,
      "num_tokens": 47493564.0,
      "step": 18735
    },
    {
      "entropy": 6.211164140701294,
      "epoch": 2.162723600692441,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004540294497611197,
      "loss": 5.7208,
      "mean_token_accuracy": 0.1968499466776848,
      "num_tokens": 47506565.0,
      "step": 18740
    },
    {
      "entropy": 6.2459207534790036,
      "epoch": 2.1633006347374497,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045400445257170725,
      "loss": 5.6451,
      "mean_token_accuracy": 0.20258357524871826,
      "num_tokens": 47518384.0,
      "step": 18745
    },
    {
      "entropy": 6.164866399765015,
      "epoch": 2.163877668782458,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004539794493615689,
      "loss": 5.6199,
      "mean_token_accuracy": 0.2032683551311493,
      "num_tokens": 47530260.0,
      "step": 18750
    },
    {
      "entropy": 6.2766376495361325,
      "epoch": 2.164454702827467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004539544401315458,
      "loss": 5.7144,
      "mean_token_accuracy": 0.19290636777877807,
      "num_tokens": 47544651.0,
      "step": 18755
    },
    {
      "entropy": 6.28360424041748,
      "epoch": 2.1650317368724754,
      "grad_norm": 0.890625,
      "learning_rate": 0.000453929424882479,
      "loss": 5.7018,
      "mean_token_accuracy": 0.19478310495615006,
      "num_tokens": 47556854.0,
      "step": 18760
    },
    {
      "entropy": 6.248087406158447,
      "epoch": 2.1656087709174843,
      "grad_norm": 1.0,
      "learning_rate": 0.00045390440361520987,
      "loss": 5.7009,
      "mean_token_accuracy": 0.19815675020217896,
      "num_tokens": 47569372.0,
      "step": 18765
    },
    {
      "entropy": 6.23117733001709,
      "epoch": 2.1661858049624927,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004538793763305799,
      "loss": 5.6511,
      "mean_token_accuracy": 0.19637321680784225,
      "num_tokens": 47581111.0,
      "step": 18770
    },
    {
      "entropy": 6.2339768409729,
      "epoch": 2.1667628390075016,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045385434302943104,
      "loss": 5.6709,
      "mean_token_accuracy": 0.20395976155996323,
      "num_tokens": 47593224.0,
      "step": 18775
    },
    {
      "entropy": 6.217057895660401,
      "epoch": 2.16733987305251,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004538293037126052,
      "loss": 5.6057,
      "mean_token_accuracy": 0.21004017293453217,
      "num_tokens": 47606959.0,
      "step": 18780
    },
    {
      "entropy": 6.187237787246704,
      "epoch": 2.167916907097519,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045380425838094444,
      "loss": 5.653,
      "mean_token_accuracy": 0.20198656916618346,
      "num_tokens": 47620159.0,
      "step": 18785
    },
    {
      "entropy": 6.316193914413452,
      "epoch": 2.1684939411425272,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045377920703529133,
      "loss": 5.7508,
      "mean_token_accuracy": 0.19582061767578124,
      "num_tokens": 47632325.0,
      "step": 18790
    },
    {
      "entropy": 6.251670169830322,
      "epoch": 2.169070975187536,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004537541496764885,
      "loss": 5.6154,
      "mean_token_accuracy": 0.2114711284637451,
      "num_tokens": 47644247.0,
      "step": 18795
    },
    {
      "entropy": 6.227293634414673,
      "epoch": 2.1696480092325445,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004537290863053786,
      "loss": 5.7016,
      "mean_token_accuracy": 0.20073844343423844,
      "num_tokens": 47655724.0,
      "step": 18800
    },
    {
      "entropy": 6.138774108886719,
      "epoch": 2.1702250432775534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045370401692280455,
      "loss": 5.6195,
      "mean_token_accuracy": 0.19920498728752137,
      "num_tokens": 47667298.0,
      "step": 18805
    },
    {
      "entropy": 6.221665906906128,
      "epoch": 2.170802077322562,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045367894152960976,
      "loss": 5.67,
      "mean_token_accuracy": 0.2062709778547287,
      "num_tokens": 47680393.0,
      "step": 18810
    },
    {
      "entropy": 6.125773811340332,
      "epoch": 2.1713791113675707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045365386012663744,
      "loss": 5.6087,
      "mean_token_accuracy": 0.20824531465768814,
      "num_tokens": 47691766.0,
      "step": 18815
    },
    {
      "entropy": 6.283021116256714,
      "epoch": 2.1719561454125795,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045362877271473137,
      "loss": 5.715,
      "mean_token_accuracy": 0.19979382753372193,
      "num_tokens": 47704710.0,
      "step": 18820
    },
    {
      "entropy": 6.200976848602295,
      "epoch": 2.172533179457588,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045360367929473517,
      "loss": 5.6108,
      "mean_token_accuracy": 0.2037011981010437,
      "num_tokens": 47716869.0,
      "step": 18825
    },
    {
      "entropy": 6.197491264343261,
      "epoch": 2.173110213502597,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000453578579867493,
      "loss": 5.6381,
      "mean_token_accuracy": 0.1991969898343086,
      "num_tokens": 47729294.0,
      "step": 18830
    },
    {
      "entropy": 6.275630617141724,
      "epoch": 2.1736872475476052,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045355347443384896,
      "loss": 5.7129,
      "mean_token_accuracy": 0.1988064631819725,
      "num_tokens": 47741130.0,
      "step": 18835
    },
    {
      "entropy": 6.221657752990723,
      "epoch": 2.174264281592614,
      "grad_norm": 0.875,
      "learning_rate": 0.00045352836299464755,
      "loss": 5.7063,
      "mean_token_accuracy": 0.19925190806388854,
      "num_tokens": 47753337.0,
      "step": 18840
    },
    {
      "entropy": 6.258538913726807,
      "epoch": 2.1748413156376225,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004535032455507333,
      "loss": 5.658,
      "mean_token_accuracy": 0.19879949539899827,
      "num_tokens": 47766105.0,
      "step": 18845
    },
    {
      "entropy": 6.239501428604126,
      "epoch": 2.1754183496826314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045347812210295107,
      "loss": 5.6134,
      "mean_token_accuracy": 0.21042157858610153,
      "num_tokens": 47777887.0,
      "step": 18850
    },
    {
      "entropy": 6.269635200500488,
      "epoch": 2.17599538372764,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045345299265214583,
      "loss": 5.7022,
      "mean_token_accuracy": 0.20443666279315947,
      "num_tokens": 47789860.0,
      "step": 18855
    },
    {
      "entropy": 6.249324131011963,
      "epoch": 2.1765724177726486,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045342785719916284,
      "loss": 5.643,
      "mean_token_accuracy": 0.19899480491876603,
      "num_tokens": 47801987.0,
      "step": 18860
    },
    {
      "entropy": 6.27932596206665,
      "epoch": 2.177149451817657,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045340271574484755,
      "loss": 5.7377,
      "mean_token_accuracy": 0.19762785732746124,
      "num_tokens": 47814586.0,
      "step": 18865
    },
    {
      "entropy": 6.190877485275268,
      "epoch": 2.177726485862666,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004533775682900454,
      "loss": 5.5725,
      "mean_token_accuracy": 0.20968065708875655,
      "num_tokens": 47827138.0,
      "step": 18870
    },
    {
      "entropy": 6.2957573413848875,
      "epoch": 2.1783035199076743,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004533524148356025,
      "loss": 5.7367,
      "mean_token_accuracy": 0.19704822450876236,
      "num_tokens": 47839748.0,
      "step": 18875
    },
    {
      "entropy": 6.078841209411621,
      "epoch": 2.178880553952683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004533272553823646,
      "loss": 5.4906,
      "mean_token_accuracy": 0.21344491839408875,
      "num_tokens": 47852542.0,
      "step": 18880
    },
    {
      "entropy": 6.231060552597046,
      "epoch": 2.179457587997692,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045330208993117816,
      "loss": 5.6586,
      "mean_token_accuracy": 0.20557589530944825,
      "num_tokens": 47866192.0,
      "step": 18885
    },
    {
      "entropy": 6.155361318588257,
      "epoch": 2.1800346220427005,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004532769184828894,
      "loss": 5.5934,
      "mean_token_accuracy": 0.20994996428489685,
      "num_tokens": 47878164.0,
      "step": 18890
    },
    {
      "entropy": 6.233050155639648,
      "epoch": 2.1806116560877093,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004532517410383451,
      "loss": 5.6408,
      "mean_token_accuracy": 0.2007750764489174,
      "num_tokens": 47890040.0,
      "step": 18895
    },
    {
      "entropy": 6.25622148513794,
      "epoch": 2.1811886901327178,
      "grad_norm": 0.953125,
      "learning_rate": 0.000453226557598392,
      "loss": 5.6789,
      "mean_token_accuracy": 0.19771016836166383,
      "num_tokens": 47902949.0,
      "step": 18900
    },
    {
      "entropy": 6.2645423889160154,
      "epoch": 2.1817657241777266,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004532013681638771,
      "loss": 5.6704,
      "mean_token_accuracy": 0.20155248641967774,
      "num_tokens": 47915674.0,
      "step": 18905
    },
    {
      "entropy": 6.191791534423828,
      "epoch": 2.182342758222735,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004531761727356478,
      "loss": 5.5637,
      "mean_token_accuracy": 0.21473615914583205,
      "num_tokens": 47929358.0,
      "step": 18910
    },
    {
      "entropy": 6.309220218658448,
      "epoch": 2.182919792267744,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004531509713145514,
      "loss": 5.7168,
      "mean_token_accuracy": 0.20053535103797912,
      "num_tokens": 47940790.0,
      "step": 18915
    },
    {
      "entropy": 6.255091667175293,
      "epoch": 2.1834968263127523,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045312576390143557,
      "loss": 5.6692,
      "mean_token_accuracy": 0.2040617659687996,
      "num_tokens": 47952651.0,
      "step": 18920
    },
    {
      "entropy": 6.2789088726043705,
      "epoch": 2.184073860357761,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045310055049714815,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1954498276114464,
      "num_tokens": 47964378.0,
      "step": 18925
    },
    {
      "entropy": 6.319014072418213,
      "epoch": 2.1846508944027696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045307533110253726,
      "loss": 5.7863,
      "mean_token_accuracy": 0.1934896543622017,
      "num_tokens": 47976724.0,
      "step": 18930
    },
    {
      "entropy": 6.2694634914398195,
      "epoch": 2.1852279284477785,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00045305010571845096,
      "loss": 5.7069,
      "mean_token_accuracy": 0.19618794471025466,
      "num_tokens": 47989364.0,
      "step": 18935
    },
    {
      "entropy": 6.203828763961792,
      "epoch": 2.185804962492787,
      "grad_norm": 1.0,
      "learning_rate": 0.0004530248743457378,
      "loss": 5.6853,
      "mean_token_accuracy": 0.20574698746204376,
      "num_tokens": 48001845.0,
      "step": 18940
    },
    {
      "entropy": 6.228608512878418,
      "epoch": 2.1863819965377957,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004529996369852465,
      "loss": 5.7367,
      "mean_token_accuracy": 0.19824532717466353,
      "num_tokens": 48016789.0,
      "step": 18945
    },
    {
      "entropy": 6.330625247955322,
      "epoch": 2.1869590305828046,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045297439363782576,
      "loss": 5.7447,
      "mean_token_accuracy": 0.18936055451631545,
      "num_tokens": 48029016.0,
      "step": 18950
    },
    {
      "entropy": 6.146222877502441,
      "epoch": 2.187536064627813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004529491443043247,
      "loss": 5.6022,
      "mean_token_accuracy": 0.2103421062231064,
      "num_tokens": 48041998.0,
      "step": 18955
    },
    {
      "entropy": 6.2504833221435545,
      "epoch": 2.188113098672822,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004529238889855926,
      "loss": 5.6749,
      "mean_token_accuracy": 0.2022068127989769,
      "num_tokens": 48054716.0,
      "step": 18960
    },
    {
      "entropy": 6.224620723724366,
      "epoch": 2.1886901327178303,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004528986276824789,
      "loss": 5.6152,
      "mean_token_accuracy": 0.20955174416303635,
      "num_tokens": 48068444.0,
      "step": 18965
    },
    {
      "entropy": 6.225123596191406,
      "epoch": 2.189267166762839,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004528733603958331,
      "loss": 5.6319,
      "mean_token_accuracy": 0.2119954138994217,
      "num_tokens": 48081494.0,
      "step": 18970
    },
    {
      "entropy": 6.205834436416626,
      "epoch": 2.1898442008078476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004528480871265053,
      "loss": 5.6534,
      "mean_token_accuracy": 0.1976575180888176,
      "num_tokens": 48094092.0,
      "step": 18975
    },
    {
      "entropy": 6.199561691284179,
      "epoch": 2.1904212348528564,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045282280787534537,
      "loss": 5.5952,
      "mean_token_accuracy": 0.20846834033727646,
      "num_tokens": 48106061.0,
      "step": 18980
    },
    {
      "entropy": 6.213763570785522,
      "epoch": 2.190998268897865,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004527975226432036,
      "loss": 5.6025,
      "mean_token_accuracy": 0.20764440596103667,
      "num_tokens": 48118686.0,
      "step": 18985
    },
    {
      "entropy": 6.269301557540894,
      "epoch": 2.1915753029428737,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045277223143093057,
      "loss": 5.6579,
      "mean_token_accuracy": 0.20267911553382872,
      "num_tokens": 48130946.0,
      "step": 18990
    },
    {
      "entropy": 6.15597243309021,
      "epoch": 2.192152336987882,
      "grad_norm": 1.0,
      "learning_rate": 0.00045274693423937674,
      "loss": 5.5367,
      "mean_token_accuracy": 0.21818872839212416,
      "num_tokens": 48143923.0,
      "step": 18995
    },
    {
      "entropy": 6.2669604301452635,
      "epoch": 2.192729371032891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045272163106939314,
      "loss": 5.7056,
      "mean_token_accuracy": 0.19601131826639176,
      "num_tokens": 48156089.0,
      "step": 19000
    },
    {
      "entropy": 6.21266508102417,
      "epoch": 2.1933064050778994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045269632192183076,
      "loss": 5.5948,
      "mean_token_accuracy": 0.20693731158971787,
      "num_tokens": 48169930.0,
      "step": 19005
    },
    {
      "entropy": 6.282389211654663,
      "epoch": 2.1938834391229083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045267100679754075,
      "loss": 5.6868,
      "mean_token_accuracy": 0.19795607775449753,
      "num_tokens": 48181378.0,
      "step": 19010
    },
    {
      "entropy": 6.189380741119384,
      "epoch": 2.1944604731679167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004526456856973748,
      "loss": 5.7024,
      "mean_token_accuracy": 0.20305758863687515,
      "num_tokens": 48193141.0,
      "step": 19015
    },
    {
      "entropy": 6.155608510971069,
      "epoch": 2.1950375072129256,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004526203586221844,
      "loss": 5.5725,
      "mean_token_accuracy": 0.21505974531173705,
      "num_tokens": 48206817.0,
      "step": 19020
    },
    {
      "entropy": 6.30601315498352,
      "epoch": 2.1956145412579344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045259502557282145,
      "loss": 5.6739,
      "mean_token_accuracy": 0.20411995947360992,
      "num_tokens": 48220104.0,
      "step": 19025
    },
    {
      "entropy": 6.247973203659058,
      "epoch": 2.196191575302943,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004525696865501381,
      "loss": 5.6927,
      "mean_token_accuracy": 0.20016210675239562,
      "num_tokens": 48232446.0,
      "step": 19030
    },
    {
      "entropy": 6.205334806442261,
      "epoch": 2.1967686093479517,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004525443415549865,
      "loss": 5.7301,
      "mean_token_accuracy": 0.19787056893110275,
      "num_tokens": 48244071.0,
      "step": 19035
    },
    {
      "entropy": 6.3160703659057615,
      "epoch": 2.19734564339296,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045251899058821915,
      "loss": 5.6888,
      "mean_token_accuracy": 0.19176996648311614,
      "num_tokens": 48256808.0,
      "step": 19040
    },
    {
      "entropy": 6.2396503448486325,
      "epoch": 2.197922677437969,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004524936336506887,
      "loss": 5.5606,
      "mean_token_accuracy": 0.2135269299149513,
      "num_tokens": 48269635.0,
      "step": 19045
    },
    {
      "entropy": 6.217072677612305,
      "epoch": 2.1984997114829774,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004524682707432482,
      "loss": 5.676,
      "mean_token_accuracy": 0.19969352632761,
      "num_tokens": 48282211.0,
      "step": 19050
    },
    {
      "entropy": 6.260828447341919,
      "epoch": 2.1990767455279863,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045244290186675034,
      "loss": 5.7264,
      "mean_token_accuracy": 0.1954602465033531,
      "num_tokens": 48295617.0,
      "step": 19055
    },
    {
      "entropy": 6.2889610767364506,
      "epoch": 2.1996537795729947,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004524175270220489,
      "loss": 5.7131,
      "mean_token_accuracy": 0.19686751961708068,
      "num_tokens": 48307885.0,
      "step": 19060
    },
    {
      "entropy": 6.274149560928345,
      "epoch": 2.2002308136180035,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045239214620999683,
      "loss": 5.6317,
      "mean_token_accuracy": 0.21213814467191697,
      "num_tokens": 48320451.0,
      "step": 19065
    },
    {
      "entropy": 6.240833950042725,
      "epoch": 2.200807847663012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004523667594314481,
      "loss": 5.611,
      "mean_token_accuracy": 0.20494293570518493,
      "num_tokens": 48333283.0,
      "step": 19070
    },
    {
      "entropy": 6.30566143989563,
      "epoch": 2.201384881708021,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00045234136668725655,
      "loss": 5.7509,
      "mean_token_accuracy": 0.19305981695652008,
      "num_tokens": 48347197.0,
      "step": 19075
    },
    {
      "entropy": 6.344265031814575,
      "epoch": 2.2019619157530292,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045231596797827616,
      "loss": 5.7597,
      "mean_token_accuracy": 0.19303266555070878,
      "num_tokens": 48359356.0,
      "step": 19080
    },
    {
      "entropy": 6.2445234775543215,
      "epoch": 2.202538949798038,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045229056330536134,
      "loss": 5.6688,
      "mean_token_accuracy": 0.2031550630927086,
      "num_tokens": 48373343.0,
      "step": 19085
    },
    {
      "entropy": 6.170627450942993,
      "epoch": 2.2031159838430465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045226515266936644,
      "loss": 5.6049,
      "mean_token_accuracy": 0.20834530144929886,
      "num_tokens": 48385614.0,
      "step": 19090
    },
    {
      "entropy": 6.231019735336304,
      "epoch": 2.2036930178880554,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004522397360711462,
      "loss": 5.658,
      "mean_token_accuracy": 0.20207223445177078,
      "num_tokens": 48397814.0,
      "step": 19095
    },
    {
      "entropy": 6.206200885772705,
      "epoch": 2.2042700519330642,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004522143135115555,
      "loss": 5.6498,
      "mean_token_accuracy": 0.21277464628219606,
      "num_tokens": 48411092.0,
      "step": 19100
    },
    {
      "entropy": 6.230615329742432,
      "epoch": 2.2048470859780727,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045218888499144933,
      "loss": 5.6634,
      "mean_token_accuracy": 0.19754871875047683,
      "num_tokens": 48423203.0,
      "step": 19105
    },
    {
      "entropy": 6.259657001495361,
      "epoch": 2.2054241200230815,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045216345051168314,
      "loss": 5.6656,
      "mean_token_accuracy": 0.20753274410963057,
      "num_tokens": 48435562.0,
      "step": 19110
    },
    {
      "entropy": 6.235284042358399,
      "epoch": 2.20600115406809,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004521380100731122,
      "loss": 5.6546,
      "mean_token_accuracy": 0.20100843608379365,
      "num_tokens": 48449021.0,
      "step": 19115
    },
    {
      "entropy": 6.203287935256958,
      "epoch": 2.206578188113099,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045211256367659234,
      "loss": 5.6131,
      "mean_token_accuracy": 0.2070888951420784,
      "num_tokens": 48462411.0,
      "step": 19120
    },
    {
      "entropy": 6.263839530944824,
      "epoch": 2.207155222158107,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004520871113229793,
      "loss": 5.6813,
      "mean_token_accuracy": 0.20064661800861358,
      "num_tokens": 48474397.0,
      "step": 19125
    },
    {
      "entropy": 6.310215663909912,
      "epoch": 2.207732256203116,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00045206165301312927,
      "loss": 5.7176,
      "mean_token_accuracy": 0.2060359612107277,
      "num_tokens": 48490549.0,
      "step": 19130
    },
    {
      "entropy": 6.320865488052368,
      "epoch": 2.2083092902481245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004520361887478985,
      "loss": 5.6977,
      "mean_token_accuracy": 0.20557867735624313,
      "num_tokens": 48503692.0,
      "step": 19135
    },
    {
      "entropy": 6.1898720264434814,
      "epoch": 2.2088863242931334,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004520107185281435,
      "loss": 5.6339,
      "mean_token_accuracy": 0.20308961123228073,
      "num_tokens": 48515574.0,
      "step": 19140
    },
    {
      "entropy": 6.25698070526123,
      "epoch": 2.209463358338142,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004519852423547208,
      "loss": 5.7365,
      "mean_token_accuracy": 0.19310335516929628,
      "num_tokens": 48527832.0,
      "step": 19145
    },
    {
      "entropy": 6.262687158584595,
      "epoch": 2.2100403923831506,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004519597602284875,
      "loss": 5.6503,
      "mean_token_accuracy": 0.20530754178762436,
      "num_tokens": 48539558.0,
      "step": 19150
    },
    {
      "entropy": 6.194172620773315,
      "epoch": 2.210617426428159,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004519342721503005,
      "loss": 5.6382,
      "mean_token_accuracy": 0.20811931490898133,
      "num_tokens": 48553360.0,
      "step": 19155
    },
    {
      "entropy": 6.273493719100952,
      "epoch": 2.211194460473168,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004519087781210171,
      "loss": 5.6838,
      "mean_token_accuracy": 0.1971590891480446,
      "num_tokens": 48567416.0,
      "step": 19160
    },
    {
      "entropy": 6.285200786590576,
      "epoch": 2.2117714945181763,
      "grad_norm": 0.921875,
      "learning_rate": 0.000451883278141495,
      "loss": 5.692,
      "mean_token_accuracy": 0.20410846322774887,
      "num_tokens": 48580086.0,
      "step": 19165
    },
    {
      "entropy": 6.224889087677002,
      "epoch": 2.212348528563185,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045185777221259157,
      "loss": 5.5675,
      "mean_token_accuracy": 0.2167016312479973,
      "num_tokens": 48591758.0,
      "step": 19170
    },
    {
      "entropy": 6.0322558879852295,
      "epoch": 2.212925562608194,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004518322603351648,
      "loss": 5.4721,
      "mean_token_accuracy": 0.22335913628339768,
      "num_tokens": 48604994.0,
      "step": 19175
    },
    {
      "entropy": 6.253607511520386,
      "epoch": 2.2135025966532025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045180674251007287,
      "loss": 5.7186,
      "mean_token_accuracy": 0.19760302156209947,
      "num_tokens": 48617652.0,
      "step": 19180
    },
    {
      "entropy": 6.238452911376953,
      "epoch": 2.2140796306982113,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045178121873817395,
      "loss": 5.5807,
      "mean_token_accuracy": 0.21068335473537445,
      "num_tokens": 48629696.0,
      "step": 19185
    },
    {
      "entropy": 6.228818607330322,
      "epoch": 2.2146566647432198,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004517556890203265,
      "loss": 5.6589,
      "mean_token_accuracy": 0.20333070009946824,
      "num_tokens": 48641297.0,
      "step": 19190
    },
    {
      "entropy": 6.250348091125488,
      "epoch": 2.2152336987882286,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045173015335738925,
      "loss": 5.6696,
      "mean_token_accuracy": 0.20270389169454575,
      "num_tokens": 48654384.0,
      "step": 19195
    },
    {
      "entropy": 6.183960008621216,
      "epoch": 2.215810732833237,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045170461175022116,
      "loss": 5.5799,
      "mean_token_accuracy": 0.20867810249328614,
      "num_tokens": 48666868.0,
      "step": 19200
    },
    {
      "entropy": 6.277534627914429,
      "epoch": 2.216387766878246,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004516790641996812,
      "loss": 5.7178,
      "mean_token_accuracy": 0.19755952358245848,
      "num_tokens": 48679780.0,
      "step": 19205
    },
    {
      "entropy": 6.256493186950683,
      "epoch": 2.2169648009232543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004516535107066286,
      "loss": 5.6662,
      "mean_token_accuracy": 0.19854632019996643,
      "num_tokens": 48692438.0,
      "step": 19210
    },
    {
      "entropy": 6.228190994262695,
      "epoch": 2.217541834968263,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045162795127192296,
      "loss": 5.6702,
      "mean_token_accuracy": 0.20286177843809128,
      "num_tokens": 48705627.0,
      "step": 19215
    },
    {
      "entropy": 6.2406003952026365,
      "epoch": 2.2181188690132716,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000451602385896424,
      "loss": 5.6372,
      "mean_token_accuracy": 0.19649529904127122,
      "num_tokens": 48717481.0,
      "step": 19220
    },
    {
      "entropy": 6.220306348800659,
      "epoch": 2.2186959030582805,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045157681458099145,
      "loss": 5.6905,
      "mean_token_accuracy": 0.19923168271780015,
      "num_tokens": 48729821.0,
      "step": 19225
    },
    {
      "entropy": 6.2750246047973635,
      "epoch": 2.2192729371032893,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004515512373264854,
      "loss": 5.6498,
      "mean_token_accuracy": 0.2031153231859207,
      "num_tokens": 48742632.0,
      "step": 19230
    },
    {
      "entropy": 6.224585437774659,
      "epoch": 2.2198499711482977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045152565413376623,
      "loss": 5.6312,
      "mean_token_accuracy": 0.20760363042354585,
      "num_tokens": 48754712.0,
      "step": 19235
    },
    {
      "entropy": 6.19266939163208,
      "epoch": 2.2204270051933066,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004515000650036944,
      "loss": 5.6569,
      "mean_token_accuracy": 0.20226311236619948,
      "num_tokens": 48767170.0,
      "step": 19240
    },
    {
      "entropy": 6.316464185714722,
      "epoch": 2.221004039238315,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004514744699371305,
      "loss": 5.7268,
      "mean_token_accuracy": 0.19159384518861772,
      "num_tokens": 48780392.0,
      "step": 19245
    },
    {
      "entropy": 6.318859100341797,
      "epoch": 2.221581073283324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045144886893493547,
      "loss": 5.6815,
      "mean_token_accuracy": 0.20438247323036193,
      "num_tokens": 48792100.0,
      "step": 19250
    },
    {
      "entropy": 6.20794095993042,
      "epoch": 2.2221581073283323,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00045142326199797047,
      "loss": 5.7112,
      "mean_token_accuracy": 0.20482142567634581,
      "num_tokens": 48805192.0,
      "step": 19255
    },
    {
      "entropy": 6.216684246063233,
      "epoch": 2.222735141373341,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045139764912709656,
      "loss": 5.5727,
      "mean_token_accuracy": 0.20802184641361238,
      "num_tokens": 48819735.0,
      "step": 19260
    },
    {
      "entropy": 6.219668865203857,
      "epoch": 2.2233121754183496,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004513720303231754,
      "loss": 5.6415,
      "mean_token_accuracy": 0.20419298112392426,
      "num_tokens": 48833785.0,
      "step": 19265
    },
    {
      "entropy": 6.256777238845825,
      "epoch": 2.2238892094633584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045134640558706864,
      "loss": 5.6478,
      "mean_token_accuracy": 0.20405424535274505,
      "num_tokens": 48846491.0,
      "step": 19270
    },
    {
      "entropy": 6.247850704193115,
      "epoch": 2.224466243508367,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000451320774919638,
      "loss": 5.5917,
      "mean_token_accuracy": 0.20437240600585938,
      "num_tokens": 48858269.0,
      "step": 19275
    },
    {
      "entropy": 6.26996693611145,
      "epoch": 2.2250432775533757,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045129513832174587,
      "loss": 5.7521,
      "mean_token_accuracy": 0.19192806482315064,
      "num_tokens": 48870051.0,
      "step": 19280
    },
    {
      "entropy": 6.239027261734009,
      "epoch": 2.225620311598384,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045126949579425414,
      "loss": 5.6675,
      "mean_token_accuracy": 0.2041851282119751,
      "num_tokens": 48882933.0,
      "step": 19285
    },
    {
      "entropy": 6.260211515426636,
      "epoch": 2.226197345643393,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045124384733802563,
      "loss": 5.663,
      "mean_token_accuracy": 0.19803516715765,
      "num_tokens": 48895641.0,
      "step": 19290
    },
    {
      "entropy": 6.3270751953125,
      "epoch": 2.2267743796884014,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004512181929539228,
      "loss": 5.7586,
      "mean_token_accuracy": 0.1929812714457512,
      "num_tokens": 48907509.0,
      "step": 19295
    },
    {
      "entropy": 6.267719268798828,
      "epoch": 2.2273514137334103,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045119253264280855,
      "loss": 5.6195,
      "mean_token_accuracy": 0.20867607295513152,
      "num_tokens": 48920270.0,
      "step": 19300
    },
    {
      "entropy": 6.201306915283203,
      "epoch": 2.227928447778419,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045116686640554607,
      "loss": 5.6118,
      "mean_token_accuracy": 0.2091464564204216,
      "num_tokens": 48932630.0,
      "step": 19305
    },
    {
      "entropy": 6.253885221481323,
      "epoch": 2.2285054818234276,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045114119424299843,
      "loss": 5.654,
      "mean_token_accuracy": 0.20716083645820618,
      "num_tokens": 48944759.0,
      "step": 19310
    },
    {
      "entropy": 6.130707025527954,
      "epoch": 2.2290825158684364,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004511155161560293,
      "loss": 5.5887,
      "mean_token_accuracy": 0.21271215230226517,
      "num_tokens": 48956835.0,
      "step": 19315
    },
    {
      "entropy": 6.159862995147705,
      "epoch": 2.229659549913445,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045108983214550225,
      "loss": 5.5569,
      "mean_token_accuracy": 0.21337656378746034,
      "num_tokens": 48970552.0,
      "step": 19320
    },
    {
      "entropy": 6.2575013637542725,
      "epoch": 2.2302365839584537,
      "grad_norm": 1.0,
      "learning_rate": 0.0004510641422122811,
      "loss": 5.7425,
      "mean_token_accuracy": 0.1989905297756195,
      "num_tokens": 48981206.0,
      "step": 19325
    },
    {
      "entropy": 6.25875039100647,
      "epoch": 2.230813618003462,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045103844635723005,
      "loss": 5.6434,
      "mean_token_accuracy": 0.2020503655076027,
      "num_tokens": 48992849.0,
      "step": 19330
    },
    {
      "entropy": 6.298848867416382,
      "epoch": 2.231390652048471,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045101274458121335,
      "loss": 5.812,
      "mean_token_accuracy": 0.19482233226299286,
      "num_tokens": 49006094.0,
      "step": 19335
    },
    {
      "entropy": 6.1478112697601315,
      "epoch": 2.2319676860934794,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045098703688509537,
      "loss": 5.6553,
      "mean_token_accuracy": 0.1996007651090622,
      "num_tokens": 49019118.0,
      "step": 19340
    },
    {
      "entropy": 6.202544641494751,
      "epoch": 2.2325447201384883,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004509613232697408,
      "loss": 5.5626,
      "mean_token_accuracy": 0.2080274999141693,
      "num_tokens": 49031217.0,
      "step": 19345
    },
    {
      "entropy": 6.285206413269043,
      "epoch": 2.2331217541834967,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004509356037360145,
      "loss": 5.6815,
      "mean_token_accuracy": 0.2019059970974922,
      "num_tokens": 49043251.0,
      "step": 19350
    },
    {
      "entropy": 6.277829027175903,
      "epoch": 2.2336987882285055,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004509098782847816,
      "loss": 5.6894,
      "mean_token_accuracy": 0.19635870605707167,
      "num_tokens": 49055227.0,
      "step": 19355
    },
    {
      "entropy": 6.284742736816407,
      "epoch": 2.234275822273514,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004508841469169073,
      "loss": 5.7336,
      "mean_token_accuracy": 0.19619537144899368,
      "num_tokens": 49067810.0,
      "step": 19360
    },
    {
      "entropy": 6.213280439376831,
      "epoch": 2.234852856318523,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045085840963325716,
      "loss": 5.603,
      "mean_token_accuracy": 0.215190526843071,
      "num_tokens": 49079898.0,
      "step": 19365
    },
    {
      "entropy": 6.276435327529907,
      "epoch": 2.2354298903635312,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004508326664346967,
      "loss": 5.6916,
      "mean_token_accuracy": 0.20098729729652404,
      "num_tokens": 49092961.0,
      "step": 19370
    },
    {
      "entropy": 6.212544679641724,
      "epoch": 2.23600692440854,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045080691732209176,
      "loss": 5.6461,
      "mean_token_accuracy": 0.20732251107692717,
      "num_tokens": 49106670.0,
      "step": 19375
    },
    {
      "entropy": 6.0918388843536375,
      "epoch": 2.236583958453549,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045078116229630864,
      "loss": 5.5094,
      "mean_token_accuracy": 0.21907488703727723,
      "num_tokens": 49119947.0,
      "step": 19380
    },
    {
      "entropy": 6.2327533721923825,
      "epoch": 2.2371609924985574,
      "grad_norm": 1.0,
      "learning_rate": 0.00045075540135821343,
      "loss": 5.6948,
      "mean_token_accuracy": 0.19821955412626266,
      "num_tokens": 49132486.0,
      "step": 19385
    },
    {
      "entropy": 6.275893497467041,
      "epoch": 2.2377380265435662,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004507296345086725,
      "loss": 5.6732,
      "mean_token_accuracy": 0.19729887545108796,
      "num_tokens": 49145121.0,
      "step": 19390
    },
    {
      "entropy": 6.2653038024902346,
      "epoch": 2.2383150605885747,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004507038617485526,
      "loss": 5.7055,
      "mean_token_accuracy": 0.1971891388297081,
      "num_tokens": 49157247.0,
      "step": 19395
    },
    {
      "entropy": 6.2951830387115475,
      "epoch": 2.2388920946335835,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045067808307872064,
      "loss": 5.7452,
      "mean_token_accuracy": 0.1944376692175865,
      "num_tokens": 49170886.0,
      "step": 19400
    },
    {
      "entropy": 6.258055877685547,
      "epoch": 2.239469128678592,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045065229850004365,
      "loss": 5.6957,
      "mean_token_accuracy": 0.19348233193159103,
      "num_tokens": 49183087.0,
      "step": 19405
    },
    {
      "entropy": 6.274082326889038,
      "epoch": 2.240046162723601,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004506265080133888,
      "loss": 5.6745,
      "mean_token_accuracy": 0.19739380329847336,
      "num_tokens": 49194703.0,
      "step": 19410
    },
    {
      "entropy": 6.251322984695435,
      "epoch": 2.240623196768609,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045060071161962364,
      "loss": 5.7208,
      "mean_token_accuracy": 0.19921400845050813,
      "num_tokens": 49208459.0,
      "step": 19415
    },
    {
      "entropy": 6.198825693130493,
      "epoch": 2.241200230813618,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004505749093196158,
      "loss": 5.5735,
      "mean_token_accuracy": 0.20759802162647248,
      "num_tokens": 49221043.0,
      "step": 19420
    },
    {
      "entropy": 6.31164140701294,
      "epoch": 2.2417772648586265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000450549101114233,
      "loss": 5.6975,
      "mean_token_accuracy": 0.19763408452272416,
      "num_tokens": 49233738.0,
      "step": 19425
    },
    {
      "entropy": 6.265511894226075,
      "epoch": 2.2423542989036354,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004505232870043434,
      "loss": 5.6994,
      "mean_token_accuracy": 0.199758380651474,
      "num_tokens": 49246244.0,
      "step": 19430
    },
    {
      "entropy": 6.272655296325683,
      "epoch": 2.2429313329486438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004504974669908152,
      "loss": 5.7314,
      "mean_token_accuracy": 0.19492171257734298,
      "num_tokens": 49258126.0,
      "step": 19435
    },
    {
      "entropy": 6.264100408554077,
      "epoch": 2.2435083669936526,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045047164107451696,
      "loss": 5.7166,
      "mean_token_accuracy": 0.20415278077125548,
      "num_tokens": 49270303.0,
      "step": 19440
    },
    {
      "entropy": 6.126714706420898,
      "epoch": 2.244085401038661,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004504458092563172,
      "loss": 5.4872,
      "mean_token_accuracy": 0.21517646163702012,
      "num_tokens": 49283295.0,
      "step": 19445
    },
    {
      "entropy": 6.278569841384888,
      "epoch": 2.24466243508367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045041997153708475,
      "loss": 5.6806,
      "mean_token_accuracy": 0.20142019242048265,
      "num_tokens": 49296003.0,
      "step": 19450
    },
    {
      "entropy": 6.294100952148438,
      "epoch": 2.2452394691286788,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045039412791768877,
      "loss": 5.7371,
      "mean_token_accuracy": 0.20109089761972426,
      "num_tokens": 49307982.0,
      "step": 19455
    },
    {
      "entropy": 6.258247518539429,
      "epoch": 2.245816503173687,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045036827839899833,
      "loss": 5.6922,
      "mean_token_accuracy": 0.19863855093717575,
      "num_tokens": 49319220.0,
      "step": 19460
    },
    {
      "entropy": 6.169884586334229,
      "epoch": 2.246393537218696,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045034242298188303,
      "loss": 5.6321,
      "mean_token_accuracy": 0.20230182856321335,
      "num_tokens": 49331683.0,
      "step": 19465
    },
    {
      "entropy": 6.175655651092529,
      "epoch": 2.2469705712637045,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004503165616672124,
      "loss": 5.5481,
      "mean_token_accuracy": 0.21665727198123932,
      "num_tokens": 49344815.0,
      "step": 19470
    },
    {
      "entropy": 6.292679357528686,
      "epoch": 2.2475476053087133,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004502906944558563,
      "loss": 5.738,
      "mean_token_accuracy": 0.19713768512010574,
      "num_tokens": 49357399.0,
      "step": 19475
    },
    {
      "entropy": 6.288967847824097,
      "epoch": 2.2481246393537218,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004502648213486848,
      "loss": 5.7242,
      "mean_token_accuracy": 0.19424415081739427,
      "num_tokens": 49369641.0,
      "step": 19480
    },
    {
      "entropy": 6.231660032272339,
      "epoch": 2.2487016733987306,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045023894234656807,
      "loss": 5.6679,
      "mean_token_accuracy": 0.2042401909828186,
      "num_tokens": 49381043.0,
      "step": 19485
    },
    {
      "entropy": 6.290990161895752,
      "epoch": 2.249278707443739,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004502130574503766,
      "loss": 5.7031,
      "mean_token_accuracy": 0.2015619918704033,
      "num_tokens": 49392862.0,
      "step": 19490
    },
    {
      "entropy": 6.3085236072540285,
      "epoch": 2.249855741488748,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004501871666609809,
      "loss": 5.7413,
      "mean_token_accuracy": 0.20127873867750168,
      "num_tokens": 49405566.0,
      "step": 19495
    },
    {
      "entropy": 6.268811511993408,
      "epoch": 2.2504327755337563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004501612699792519,
      "loss": 5.6711,
      "mean_token_accuracy": 0.20113914906978608,
      "num_tokens": 49418048.0,
      "step": 19500
    },
    {
      "entropy": 6.209052467346192,
      "epoch": 2.251009809578765,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004501353674060606,
      "loss": 5.668,
      "mean_token_accuracy": 0.20543440133333207,
      "num_tokens": 49430129.0,
      "step": 19505
    },
    {
      "entropy": 6.268482971191406,
      "epoch": 2.251586843623774,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004501094589422782,
      "loss": 5.7131,
      "mean_token_accuracy": 0.1948510617017746,
      "num_tokens": 49443047.0,
      "step": 19510
    },
    {
      "entropy": 6.2901740074157715,
      "epoch": 2.2521638776687825,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045008354458877614,
      "loss": 5.7149,
      "mean_token_accuracy": 0.2022533968091011,
      "num_tokens": 49456061.0,
      "step": 19515
    },
    {
      "entropy": 6.109235620498657,
      "epoch": 2.252740911713791,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000450057624346426,
      "loss": 5.4558,
      "mean_token_accuracy": 0.21882125586271287,
      "num_tokens": 49470044.0,
      "step": 19520
    },
    {
      "entropy": 6.237044191360473,
      "epoch": 2.2533179457587997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045003169821609967,
      "loss": 5.6801,
      "mean_token_accuracy": 0.19884446263313293,
      "num_tokens": 49482232.0,
      "step": 19525
    },
    {
      "entropy": 6.233609342575074,
      "epoch": 2.2538949798038086,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045000576619866914,
      "loss": 5.6115,
      "mean_token_accuracy": 0.20279134958982467,
      "num_tokens": 49493639.0,
      "step": 19530
    },
    {
      "entropy": 6.317775630950928,
      "epoch": 2.254472013848817,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044997982829500657,
      "loss": 5.7753,
      "mean_token_accuracy": 0.19574931710958482,
      "num_tokens": 49506233.0,
      "step": 19535
    },
    {
      "entropy": 6.212774848937988,
      "epoch": 2.255049047893826,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044995388450598436,
      "loss": 5.688,
      "mean_token_accuracy": 0.2011435702443123,
      "num_tokens": 49518479.0,
      "step": 19540
    },
    {
      "entropy": 6.255437326431275,
      "epoch": 2.2556260819388343,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004499279348324751,
      "loss": 5.6915,
      "mean_token_accuracy": 0.19872308373451233,
      "num_tokens": 49531293.0,
      "step": 19545
    },
    {
      "entropy": 6.173596620559692,
      "epoch": 2.256203115983843,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044990197927535173,
      "loss": 5.5739,
      "mean_token_accuracy": 0.20866711735725402,
      "num_tokens": 49543492.0,
      "step": 19550
    },
    {
      "entropy": 6.260227680206299,
      "epoch": 2.2567801500288516,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044987601783548703,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1971447914838791,
      "num_tokens": 49555460.0,
      "step": 19555
    },
    {
      "entropy": 6.1519458293914795,
      "epoch": 2.2573571840738604,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004498500505137545,
      "loss": 5.5325,
      "mean_token_accuracy": 0.20627312064170839,
      "num_tokens": 49568077.0,
      "step": 19560
    },
    {
      "entropy": 6.223918151855469,
      "epoch": 2.257934218118869,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004498240773110273,
      "loss": 5.6707,
      "mean_token_accuracy": 0.20031350702047349,
      "num_tokens": 49580201.0,
      "step": 19565
    },
    {
      "entropy": 6.243503475189209,
      "epoch": 2.2585112521638777,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004497980982281791,
      "loss": 5.6325,
      "mean_token_accuracy": 0.20472485721111297,
      "num_tokens": 49592469.0,
      "step": 19570
    },
    {
      "entropy": 6.2865403175354,
      "epoch": 2.259088286208886,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004497721132660837,
      "loss": 5.7287,
      "mean_token_accuracy": 0.1972167506814003,
      "num_tokens": 49604961.0,
      "step": 19575
    },
    {
      "entropy": 6.24722809791565,
      "epoch": 2.259665320253895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044974612242561516,
      "loss": 5.694,
      "mean_token_accuracy": 0.19423324465751649,
      "num_tokens": 49617435.0,
      "step": 19580
    },
    {
      "entropy": 6.268092012405395,
      "epoch": 2.260242354298904,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004497201257076475,
      "loss": 5.6829,
      "mean_token_accuracy": 0.20406584739685057,
      "num_tokens": 49629799.0,
      "step": 19585
    },
    {
      "entropy": 6.135132360458374,
      "epoch": 2.2608193883439123,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004496941231130552,
      "loss": 5.5275,
      "mean_token_accuracy": 0.20646347403526305,
      "num_tokens": 49642937.0,
      "step": 19590
    },
    {
      "entropy": 6.213338136672974,
      "epoch": 2.261396422388921,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004496681146427129,
      "loss": 5.6333,
      "mean_token_accuracy": 0.20206339359283448,
      "num_tokens": 49655731.0,
      "step": 19595
    },
    {
      "entropy": 6.283998298645019,
      "epoch": 2.2619734564339296,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044964210029749523,
      "loss": 5.6733,
      "mean_token_accuracy": 0.19964399486780166,
      "num_tokens": 49668515.0,
      "step": 19600
    },
    {
      "entropy": 6.273973274230957,
      "epoch": 2.2625504904789384,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044961608007827736,
      "loss": 5.7051,
      "mean_token_accuracy": 0.19750163853168487,
      "num_tokens": 49681246.0,
      "step": 19605
    },
    {
      "entropy": 6.219721126556396,
      "epoch": 2.263127524523947,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004495900539859344,
      "loss": 5.6893,
      "mean_token_accuracy": 0.19896354228258134,
      "num_tokens": 49694234.0,
      "step": 19610
    },
    {
      "entropy": 6.174768018722534,
      "epoch": 2.2637045585689557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044956402202134157,
      "loss": 5.5626,
      "mean_token_accuracy": 0.21196575164794923,
      "num_tokens": 49707511.0,
      "step": 19615
    },
    {
      "entropy": 6.232049798965454,
      "epoch": 2.264281592613964,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044953798418537465,
      "loss": 5.6734,
      "mean_token_accuracy": 0.21190411895513533,
      "num_tokens": 49720348.0,
      "step": 19620
    },
    {
      "entropy": 6.137292528152466,
      "epoch": 2.264858626658973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004495119404789093,
      "loss": 5.5218,
      "mean_token_accuracy": 0.20552247017621994,
      "num_tokens": 49733777.0,
      "step": 19625
    },
    {
      "entropy": 6.233484411239624,
      "epoch": 2.2654356607039814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004494858909028216,
      "loss": 5.6017,
      "mean_token_accuracy": 0.20797090530395507,
      "num_tokens": 49746962.0,
      "step": 19630
    },
    {
      "entropy": 6.319221830368042,
      "epoch": 2.2660126947489903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004494598354579875,
      "loss": 5.735,
      "mean_token_accuracy": 0.19510992020368575,
      "num_tokens": 49760211.0,
      "step": 19635
    },
    {
      "entropy": 6.194464254379272,
      "epoch": 2.2665897287939987,
      "grad_norm": 0.875,
      "learning_rate": 0.0004494337741452836,
      "loss": 5.5803,
      "mean_token_accuracy": 0.21507738530635834,
      "num_tokens": 49773215.0,
      "step": 19640
    },
    {
      "entropy": 6.22137222290039,
      "epoch": 2.2671667628390075,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004494077069655863,
      "loss": 5.6395,
      "mean_token_accuracy": 0.2070373088121414,
      "num_tokens": 49785521.0,
      "step": 19645
    },
    {
      "entropy": 6.209274816513061,
      "epoch": 2.267743796884016,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004493816339197724,
      "loss": 5.6304,
      "mean_token_accuracy": 0.20317748337984085,
      "num_tokens": 49796449.0,
      "step": 19650
    },
    {
      "entropy": 6.225429582595825,
      "epoch": 2.268320830929025,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044935555500871897,
      "loss": 5.6795,
      "mean_token_accuracy": 0.20555828213691713,
      "num_tokens": 49810222.0,
      "step": 19655
    },
    {
      "entropy": 6.261484289169312,
      "epoch": 2.2688978649740337,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000449329470233303,
      "loss": 5.6491,
      "mean_token_accuracy": 0.20452033132314681,
      "num_tokens": 49822799.0,
      "step": 19660
    },
    {
      "entropy": 6.242724132537842,
      "epoch": 2.269474899019042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044930337959440183,
      "loss": 5.6407,
      "mean_token_accuracy": 0.20070089995861054,
      "num_tokens": 49835782.0,
      "step": 19665
    },
    {
      "entropy": 6.258236217498779,
      "epoch": 2.270051933064051,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004492772830928931,
      "loss": 5.6624,
      "mean_token_accuracy": 0.20089285224676132,
      "num_tokens": 49849204.0,
      "step": 19670
    },
    {
      "entropy": 6.282140111923217,
      "epoch": 2.2706289671090594,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044925118072965456,
      "loss": 5.7453,
      "mean_token_accuracy": 0.1938568726181984,
      "num_tokens": 49860769.0,
      "step": 19675
    },
    {
      "entropy": 6.245398902893067,
      "epoch": 2.2712060011540682,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000449225072505564,
      "loss": 5.6386,
      "mean_token_accuracy": 0.205486498773098,
      "num_tokens": 49873923.0,
      "step": 19680
    },
    {
      "entropy": 6.236215114593506,
      "epoch": 2.2717830351990767,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044919895842149976,
      "loss": 5.6438,
      "mean_token_accuracy": 0.20596978217363357,
      "num_tokens": 49886968.0,
      "step": 19685
    },
    {
      "entropy": 6.140633821487427,
      "epoch": 2.2723600692440855,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044917283847834005,
      "loss": 5.6018,
      "mean_token_accuracy": 0.208343805372715,
      "num_tokens": 49899096.0,
      "step": 19690
    },
    {
      "entropy": 6.207383632659912,
      "epoch": 2.272937103289094,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004491467126769635,
      "loss": 5.6052,
      "mean_token_accuracy": 0.20171039700508117,
      "num_tokens": 49911887.0,
      "step": 19695
    },
    {
      "entropy": 6.253566074371338,
      "epoch": 2.273514137334103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044912058101824866,
      "loss": 5.6058,
      "mean_token_accuracy": 0.21103622615337372,
      "num_tokens": 49925188.0,
      "step": 19700
    },
    {
      "entropy": 6.178601169586182,
      "epoch": 2.274091171379111,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00044909444350307463,
      "loss": 5.6028,
      "mean_token_accuracy": 0.20649515241384506,
      "num_tokens": 49938568.0,
      "step": 19705
    },
    {
      "entropy": 6.173976993560791,
      "epoch": 2.27466820542412,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004490683001323205,
      "loss": 5.5498,
      "mean_token_accuracy": 0.21556459218263627,
      "num_tokens": 49950630.0,
      "step": 19710
    },
    {
      "entropy": 6.2054931163787845,
      "epoch": 2.2752452394691285,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044904215090686554,
      "loss": 5.6026,
      "mean_token_accuracy": 0.21213731914758682,
      "num_tokens": 49962242.0,
      "step": 19715
    },
    {
      "entropy": 6.264598274230957,
      "epoch": 2.2758222735141374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044901599582758926,
      "loss": 5.7216,
      "mean_token_accuracy": 0.1935809999704361,
      "num_tokens": 49974604.0,
      "step": 19720
    },
    {
      "entropy": 6.23988299369812,
      "epoch": 2.2763993075591458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044898983489537143,
      "loss": 5.6376,
      "mean_token_accuracy": 0.20834981948137282,
      "num_tokens": 49986326.0,
      "step": 19725
    },
    {
      "entropy": 6.248817586898804,
      "epoch": 2.2769763416041546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000448963668111092,
      "loss": 5.5986,
      "mean_token_accuracy": 0.21556760966777802,
      "num_tokens": 49998752.0,
      "step": 19730
    },
    {
      "entropy": 6.265404748916626,
      "epoch": 2.2775533756491635,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044893749547563085,
      "loss": 5.702,
      "mean_token_accuracy": 0.204070346057415,
      "num_tokens": 50011839.0,
      "step": 19735
    },
    {
      "entropy": 6.199413204193116,
      "epoch": 2.278130409694172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044891131698986846,
      "loss": 5.5867,
      "mean_token_accuracy": 0.2110982283949852,
      "num_tokens": 50024554.0,
      "step": 19740
    },
    {
      "entropy": 6.2423789501190186,
      "epoch": 2.2787074437391808,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004488851326546854,
      "loss": 5.7213,
      "mean_token_accuracy": 0.20015700459480285,
      "num_tokens": 50038868.0,
      "step": 19745
    },
    {
      "entropy": 6.29875431060791,
      "epoch": 2.279284477784189,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004488589424709622,
      "loss": 5.6735,
      "mean_token_accuracy": 0.20198557227849961,
      "num_tokens": 50050556.0,
      "step": 19750
    },
    {
      "entropy": 6.283203840255737,
      "epoch": 2.279861511829198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004488327464395799,
      "loss": 5.7024,
      "mean_token_accuracy": 0.20245268791913987,
      "num_tokens": 50063169.0,
      "step": 19755
    },
    {
      "entropy": 6.279094171524048,
      "epoch": 2.2804385458742065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004488065445614195,
      "loss": 5.7086,
      "mean_token_accuracy": 0.20200192034244538,
      "num_tokens": 50076563.0,
      "step": 19760
    },
    {
      "entropy": 6.210533189773559,
      "epoch": 2.2810155799192153,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004487803368373623,
      "loss": 5.6668,
      "mean_token_accuracy": 0.20192975401878357,
      "num_tokens": 50089584.0,
      "step": 19765
    },
    {
      "entropy": 6.214915132522583,
      "epoch": 2.2815926139642237,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004487541232682898,
      "loss": 5.6054,
      "mean_token_accuracy": 0.20104413330554963,
      "num_tokens": 50101566.0,
      "step": 19770
    },
    {
      "entropy": 6.273401737213135,
      "epoch": 2.2821696480092326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004487279038550836,
      "loss": 5.6783,
      "mean_token_accuracy": 0.20586044043302537,
      "num_tokens": 50114378.0,
      "step": 19775
    },
    {
      "entropy": 6.209450101852417,
      "epoch": 2.282746682054241,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004487016785986258,
      "loss": 5.6368,
      "mean_token_accuracy": 0.20735076069831848,
      "num_tokens": 50126658.0,
      "step": 19780
    },
    {
      "entropy": 6.242487382888794,
      "epoch": 2.28332371609925,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044867544749979817,
      "loss": 5.6664,
      "mean_token_accuracy": 0.19945186823606492,
      "num_tokens": 50139802.0,
      "step": 19785
    },
    {
      "entropy": 6.336682415008545,
      "epoch": 2.2839007501442588,
      "grad_norm": 1.0,
      "learning_rate": 0.00044864921055948317,
      "loss": 5.7687,
      "mean_token_accuracy": 0.19500694572925567,
      "num_tokens": 50152158.0,
      "step": 19790
    },
    {
      "entropy": 6.30319299697876,
      "epoch": 2.284477784189267,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044862296777856326,
      "loss": 5.655,
      "mean_token_accuracy": 0.20215387195348739,
      "num_tokens": 50165099.0,
      "step": 19795
    },
    {
      "entropy": 6.177653408050537,
      "epoch": 2.2850548182342756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004485967191579211,
      "loss": 5.6256,
      "mean_token_accuracy": 0.20661345720291138,
      "num_tokens": 50177073.0,
      "step": 19800
    },
    {
      "entropy": 6.260256958007813,
      "epoch": 2.2856318522792844,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004485704646984394,
      "loss": 5.7014,
      "mean_token_accuracy": 0.200007364153862,
      "num_tokens": 50190458.0,
      "step": 19805
    },
    {
      "entropy": 6.212049961090088,
      "epoch": 2.2862088863242933,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004485442044010015,
      "loss": 5.6207,
      "mean_token_accuracy": 0.2056537538766861,
      "num_tokens": 50201726.0,
      "step": 19810
    },
    {
      "entropy": 6.189518547058105,
      "epoch": 2.2867859203693017,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004485179382664904,
      "loss": 5.596,
      "mean_token_accuracy": 0.21296337842941285,
      "num_tokens": 50215550.0,
      "step": 19815
    },
    {
      "entropy": 6.21746597290039,
      "epoch": 2.2873629544143106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004484916662957896,
      "loss": 5.6267,
      "mean_token_accuracy": 0.20205569863319398,
      "num_tokens": 50227982.0,
      "step": 19820
    },
    {
      "entropy": 6.273596143722534,
      "epoch": 2.287939988459319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004484653884897829,
      "loss": 5.6713,
      "mean_token_accuracy": 0.19844297468662261,
      "num_tokens": 50239804.0,
      "step": 19825
    },
    {
      "entropy": 6.293114852905274,
      "epoch": 2.288517022504328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044843910484935394,
      "loss": 5.6805,
      "mean_token_accuracy": 0.20734443366527558,
      "num_tokens": 50252192.0,
      "step": 19830
    },
    {
      "entropy": 6.308523654937744,
      "epoch": 2.2890940565493363,
      "grad_norm": 1.0,
      "learning_rate": 0.0004484128153753868,
      "loss": 5.7675,
      "mean_token_accuracy": 0.19301671385765076,
      "num_tokens": 50265194.0,
      "step": 19835
    },
    {
      "entropy": 6.21258020401001,
      "epoch": 2.289671090594345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044838652006876583,
      "loss": 5.6059,
      "mean_token_accuracy": 0.21008452028036118,
      "num_tokens": 50277704.0,
      "step": 19840
    },
    {
      "entropy": 6.180286884307861,
      "epoch": 2.2902481246393536,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044836021893037537,
      "loss": 5.5926,
      "mean_token_accuracy": 0.20745245963335038,
      "num_tokens": 50290670.0,
      "step": 19845
    },
    {
      "entropy": 6.18676266670227,
      "epoch": 2.2908251586843624,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004483339119611001,
      "loss": 5.6055,
      "mean_token_accuracy": 0.206353996694088,
      "num_tokens": 50302764.0,
      "step": 19850
    },
    {
      "entropy": 6.224746513366699,
      "epoch": 2.291402192729371,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044830759916182476,
      "loss": 5.6555,
      "mean_token_accuracy": 0.20963870286941527,
      "num_tokens": 50316824.0,
      "step": 19855
    },
    {
      "entropy": 6.386384534835815,
      "epoch": 2.2919792267743797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004482812805334344,
      "loss": 5.7115,
      "mean_token_accuracy": 0.19778601378202437,
      "num_tokens": 50329259.0,
      "step": 19860
    },
    {
      "entropy": 6.171825981140136,
      "epoch": 2.2925562608193886,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004482549560768142,
      "loss": 5.6044,
      "mean_token_accuracy": 0.2155290201306343,
      "num_tokens": 50341165.0,
      "step": 19865
    },
    {
      "entropy": 6.097897148132324,
      "epoch": 2.293133294864397,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004482286257928497,
      "loss": 5.609,
      "mean_token_accuracy": 0.20484520643949508,
      "num_tokens": 50354655.0,
      "step": 19870
    },
    {
      "entropy": 6.255611276626587,
      "epoch": 2.293710328909406,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004482022896824263,
      "loss": 5.7013,
      "mean_token_accuracy": 0.20133717656135558,
      "num_tokens": 50366917.0,
      "step": 19875
    },
    {
      "entropy": 6.2481482982635494,
      "epoch": 2.2942873629544143,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044817594774643004,
      "loss": 5.7083,
      "mean_token_accuracy": 0.20200784653425216,
      "num_tokens": 50380415.0,
      "step": 19880
    },
    {
      "entropy": 6.214859533309936,
      "epoch": 2.294864396999423,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044814959998574675,
      "loss": 5.6189,
      "mean_token_accuracy": 0.20611760318279265,
      "num_tokens": 50393051.0,
      "step": 19885
    },
    {
      "entropy": 6.259435749053955,
      "epoch": 2.2954414310444315,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044812324640126265,
      "loss": 5.707,
      "mean_token_accuracy": 0.196713587641716,
      "num_tokens": 50407517.0,
      "step": 19890
    },
    {
      "entropy": 6.303005743026733,
      "epoch": 2.2960184650894404,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004480968869938642,
      "loss": 5.7476,
      "mean_token_accuracy": 0.19690272808074952,
      "num_tokens": 50419656.0,
      "step": 19895
    },
    {
      "entropy": 6.205738210678101,
      "epoch": 2.296595499134449,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00044807052176443793,
      "loss": 5.5631,
      "mean_token_accuracy": 0.20255490392446518,
      "num_tokens": 50433798.0,
      "step": 19900
    },
    {
      "entropy": 6.29276967048645,
      "epoch": 2.2971725331794577,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044804415071387067,
      "loss": 5.7365,
      "mean_token_accuracy": 0.19925104081630707,
      "num_tokens": 50445459.0,
      "step": 19905
    },
    {
      "entropy": 6.234588241577148,
      "epoch": 2.297749567224466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004480177738430492,
      "loss": 5.6154,
      "mean_token_accuracy": 0.21303294450044633,
      "num_tokens": 50458649.0,
      "step": 19910
    },
    {
      "entropy": 6.103732299804688,
      "epoch": 2.298326601269475,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044799139115286104,
      "loss": 5.4755,
      "mean_token_accuracy": 0.2213372528553009,
      "num_tokens": 50471878.0,
      "step": 19915
    },
    {
      "entropy": 6.285918092727661,
      "epoch": 2.2989036353144834,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004479650026441933,
      "loss": 5.726,
      "mean_token_accuracy": 0.19926259964704512,
      "num_tokens": 50484033.0,
      "step": 19920
    },
    {
      "entropy": 6.263997268676758,
      "epoch": 2.2994806693594922,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044793860831793356,
      "loss": 5.663,
      "mean_token_accuracy": 0.20496127605438233,
      "num_tokens": 50496072.0,
      "step": 19925
    },
    {
      "entropy": 6.26498966217041,
      "epoch": 2.3000577034045007,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044791220817496963,
      "loss": 5.6294,
      "mean_token_accuracy": 0.2116732567548752,
      "num_tokens": 50508572.0,
      "step": 19930
    },
    {
      "entropy": 6.232687902450562,
      "epoch": 2.3006347374495095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004478858022161895,
      "loss": 5.7477,
      "mean_token_accuracy": 0.20238634943962097,
      "num_tokens": 50521946.0,
      "step": 19935
    },
    {
      "entropy": 6.272780323028565,
      "epoch": 2.3012117714945184,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004478593904424813,
      "loss": 5.608,
      "mean_token_accuracy": 0.20416843593120576,
      "num_tokens": 50534517.0,
      "step": 19940
    },
    {
      "entropy": 6.23766565322876,
      "epoch": 2.301788805539527,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004478329728547334,
      "loss": 5.6649,
      "mean_token_accuracy": 0.20591583997011184,
      "num_tokens": 50546479.0,
      "step": 19945
    },
    {
      "entropy": 6.286223268508911,
      "epoch": 2.3023658395845357,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044780654945383424,
      "loss": 5.7424,
      "mean_token_accuracy": 0.19686770141124726,
      "num_tokens": 50559579.0,
      "step": 19950
    },
    {
      "entropy": 6.254808044433593,
      "epoch": 2.302942873629544,
      "grad_norm": 1.0,
      "learning_rate": 0.00044778012024067267,
      "loss": 5.5773,
      "mean_token_accuracy": 0.20634620487689972,
      "num_tokens": 50572810.0,
      "step": 19955
    },
    {
      "entropy": 6.263852596282959,
      "epoch": 2.303519907674553,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004477536852161376,
      "loss": 5.6596,
      "mean_token_accuracy": 0.1994602084159851,
      "num_tokens": 50585514.0,
      "step": 19960
    },
    {
      "entropy": 6.223838663101196,
      "epoch": 2.3040969417195614,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004477272443811181,
      "loss": 5.6893,
      "mean_token_accuracy": 0.20446572154760362,
      "num_tokens": 50597897.0,
      "step": 19965
    },
    {
      "entropy": 6.204572677612305,
      "epoch": 2.3046739757645702,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004477007977365035,
      "loss": 5.6756,
      "mean_token_accuracy": 0.20004230439662934,
      "num_tokens": 50610251.0,
      "step": 19970
    },
    {
      "entropy": 6.237534761428833,
      "epoch": 2.3052510098095786,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004476743452831834,
      "loss": 5.6514,
      "mean_token_accuracy": 0.2027403861284256,
      "num_tokens": 50622212.0,
      "step": 19975
    },
    {
      "entropy": 6.207398223876953,
      "epoch": 2.3058280438545875,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044764788702204747,
      "loss": 5.6407,
      "mean_token_accuracy": 0.19937213361263276,
      "num_tokens": 50636648.0,
      "step": 19980
    },
    {
      "entropy": 6.310660123825073,
      "epoch": 2.306405077899596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004476214229539856,
      "loss": 5.7105,
      "mean_token_accuracy": 0.20629100501537323,
      "num_tokens": 50648033.0,
      "step": 19985
    },
    {
      "entropy": 6.279502248764038,
      "epoch": 2.306982111944605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004475949530798879,
      "loss": 5.74,
      "mean_token_accuracy": 0.19943794161081313,
      "num_tokens": 50660547.0,
      "step": 19990
    },
    {
      "entropy": 6.229419755935669,
      "epoch": 2.307559145989613,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044756847740064475,
      "loss": 5.6284,
      "mean_token_accuracy": 0.2128899797797203,
      "num_tokens": 50672194.0,
      "step": 19995
    },
    {
      "entropy": 6.17949800491333,
      "epoch": 2.308136180034622,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004475419959171465,
      "loss": 5.6518,
      "mean_token_accuracy": 0.20565163642168044,
      "num_tokens": 50685896.0,
      "step": 20000
    },
    {
      "entropy": 6.149283123016358,
      "epoch": 2.3087132140796305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000447515508630284,
      "loss": 5.5984,
      "mean_token_accuracy": 0.20428095310926436,
      "num_tokens": 50699255.0,
      "step": 20005
    },
    {
      "entropy": 6.189637041091919,
      "epoch": 2.3092902481246393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044748901554094806,
      "loss": 5.5861,
      "mean_token_accuracy": 0.20944172590970994,
      "num_tokens": 50711098.0,
      "step": 20010
    },
    {
      "entropy": 6.2908350944519045,
      "epoch": 2.309867282169648,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004474625166500297,
      "loss": 5.7942,
      "mean_token_accuracy": 0.19428673386573792,
      "num_tokens": 50724821.0,
      "step": 20015
    },
    {
      "entropy": 6.321254825592041,
      "epoch": 2.3104443162146566,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044743601195842026,
      "loss": 5.6965,
      "mean_token_accuracy": 0.19857280552387238,
      "num_tokens": 50736572.0,
      "step": 20020
    },
    {
      "entropy": 6.030602979660034,
      "epoch": 2.3110213502596655,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00044740950146701127,
      "loss": 5.4352,
      "mean_token_accuracy": 0.21739086359739304,
      "num_tokens": 50749743.0,
      "step": 20025
    },
    {
      "entropy": 6.243865823745727,
      "epoch": 2.311598384304674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004473829851766943,
      "loss": 5.6412,
      "mean_token_accuracy": 0.21195403039455413,
      "num_tokens": 50761558.0,
      "step": 20030
    },
    {
      "entropy": 6.29378981590271,
      "epoch": 2.3121754183496828,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004473564630883612,
      "loss": 5.7132,
      "mean_token_accuracy": 0.19658023416996,
      "num_tokens": 50773963.0,
      "step": 20035
    },
    {
      "entropy": 6.205898332595825,
      "epoch": 2.312752452394691,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004473299352029042,
      "loss": 5.6463,
      "mean_token_accuracy": 0.2047144815325737,
      "num_tokens": 50787160.0,
      "step": 20040
    },
    {
      "entropy": 6.241002798080444,
      "epoch": 2.3133294864397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004473034015212154,
      "loss": 5.5231,
      "mean_token_accuracy": 0.21583717912435532,
      "num_tokens": 50800014.0,
      "step": 20045
    },
    {
      "entropy": 6.192015171051025,
      "epoch": 2.3139065204847085,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004472768620441872,
      "loss": 5.5899,
      "mean_token_accuracy": 0.20334323346614838,
      "num_tokens": 50812149.0,
      "step": 20050
    },
    {
      "entropy": 6.259647798538208,
      "epoch": 2.3144835545297173,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044725031677271234,
      "loss": 5.7097,
      "mean_token_accuracy": 0.2004058539867401,
      "num_tokens": 50825044.0,
      "step": 20055
    },
    {
      "entropy": 6.340881776809693,
      "epoch": 2.3150605885747257,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004472237657076837,
      "loss": 5.7665,
      "mean_token_accuracy": 0.20082692354917525,
      "num_tokens": 50838813.0,
      "step": 20060
    },
    {
      "entropy": 6.233511161804199,
      "epoch": 2.3156376226197346,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004471972088499942,
      "loss": 5.5484,
      "mean_token_accuracy": 0.20299635976552963,
      "num_tokens": 50853029.0,
      "step": 20065
    },
    {
      "entropy": 6.2278190612792965,
      "epoch": 2.3162146566647435,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004471706462005371,
      "loss": 5.6423,
      "mean_token_accuracy": 0.20507512837648392,
      "num_tokens": 50866383.0,
      "step": 20070
    },
    {
      "entropy": 6.2081972599029545,
      "epoch": 2.316791690709752,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004471440777602059,
      "loss": 5.6411,
      "mean_token_accuracy": 0.20972357988357543,
      "num_tokens": 50880537.0,
      "step": 20075
    },
    {
      "entropy": 6.1785074234008786,
      "epoch": 2.3173687247547603,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044711750352989407,
      "loss": 5.622,
      "mean_token_accuracy": 0.2089190348982811,
      "num_tokens": 50892213.0,
      "step": 20080
    },
    {
      "entropy": 6.224743223190307,
      "epoch": 2.317945758799769,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004470909235104956,
      "loss": 5.6722,
      "mean_token_accuracy": 0.20047866106033324,
      "num_tokens": 50904022.0,
      "step": 20085
    },
    {
      "entropy": 6.22236590385437,
      "epoch": 2.318522792844778,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004470643377029043,
      "loss": 5.643,
      "mean_token_accuracy": 0.2028649106621742,
      "num_tokens": 50916766.0,
      "step": 20090
    },
    {
      "entropy": 6.2523521900177,
      "epoch": 2.3190998268897864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004470377461080145,
      "loss": 5.6173,
      "mean_token_accuracy": 0.20249716192483902,
      "num_tokens": 50928842.0,
      "step": 20095
    },
    {
      "entropy": 6.224369239807129,
      "epoch": 2.3196768609347953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004470111487267206,
      "loss": 5.6533,
      "mean_token_accuracy": 0.2025792270898819,
      "num_tokens": 50941079.0,
      "step": 20100
    },
    {
      "entropy": 6.196924352645874,
      "epoch": 2.3202538949798037,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004469845455599171,
      "loss": 5.6179,
      "mean_token_accuracy": 0.20616311132907866,
      "num_tokens": 50953299.0,
      "step": 20105
    },
    {
      "entropy": 6.234563875198364,
      "epoch": 2.3208309290248126,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004469579366084989,
      "loss": 5.6111,
      "mean_token_accuracy": 0.20312338769435884,
      "num_tokens": 50965550.0,
      "step": 20110
    },
    {
      "entropy": 6.242292261123657,
      "epoch": 2.321407963069821,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044693132187336094,
      "loss": 5.6313,
      "mean_token_accuracy": 0.2061188131570816,
      "num_tokens": 50978454.0,
      "step": 20115
    },
    {
      "entropy": 6.253513383865356,
      "epoch": 2.32198499711483,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004469047013553983,
      "loss": 5.6664,
      "mean_token_accuracy": 0.20614543557167053,
      "num_tokens": 50990473.0,
      "step": 20120
    },
    {
      "entropy": 6.2439676284790036,
      "epoch": 2.3225620311598383,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044687807505550646,
      "loss": 5.7371,
      "mean_token_accuracy": 0.19927602261304855,
      "num_tokens": 51003914.0,
      "step": 20125
    },
    {
      "entropy": 6.314337491989136,
      "epoch": 2.323139065204847,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044685144297458096,
      "loss": 5.7356,
      "mean_token_accuracy": 0.19555247724056243,
      "num_tokens": 51016874.0,
      "step": 20130
    },
    {
      "entropy": 6.242905044555664,
      "epoch": 2.3237160992498556,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044682480511351754,
      "loss": 5.6388,
      "mean_token_accuracy": 0.2058805674314499,
      "num_tokens": 51030568.0,
      "step": 20135
    },
    {
      "entropy": 6.298398113250732,
      "epoch": 2.3242931332948644,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004467981614732121,
      "loss": 5.6662,
      "mean_token_accuracy": 0.20071652084589003,
      "num_tokens": 51043876.0,
      "step": 20140
    },
    {
      "entropy": 6.272222518920898,
      "epoch": 2.3248701673398733,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044677151205456086,
      "loss": 5.7368,
      "mean_token_accuracy": 0.19520047754049302,
      "num_tokens": 51056693.0,
      "step": 20145
    },
    {
      "entropy": 6.313663482666016,
      "epoch": 2.3254472013848817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004467448568584601,
      "loss": 5.7123,
      "mean_token_accuracy": 0.20012226402759553,
      "num_tokens": 51068878.0,
      "step": 20150
    },
    {
      "entropy": 6.242487239837646,
      "epoch": 2.32602423542989,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004467181958858064,
      "loss": 5.7012,
      "mean_token_accuracy": 0.19994593262672425,
      "num_tokens": 51082234.0,
      "step": 20155
    },
    {
      "entropy": 6.217567682266235,
      "epoch": 2.326601269474899,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004466915291374965,
      "loss": 5.6737,
      "mean_token_accuracy": 0.20937056541442872,
      "num_tokens": 51095469.0,
      "step": 20160
    },
    {
      "entropy": 6.342366170883179,
      "epoch": 2.327178303519908,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004466648566144273,
      "loss": 5.6637,
      "mean_token_accuracy": 0.20084816962480545,
      "num_tokens": 51107302.0,
      "step": 20165
    },
    {
      "entropy": 6.266392612457276,
      "epoch": 2.3277553375649163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000446638178317496,
      "loss": 5.7272,
      "mean_token_accuracy": 0.19425797909498216,
      "num_tokens": 51121222.0,
      "step": 20170
    },
    {
      "entropy": 6.279293584823608,
      "epoch": 2.328332371609925,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044661149424759974,
      "loss": 5.7287,
      "mean_token_accuracy": 0.19627294540405274,
      "num_tokens": 51134113.0,
      "step": 20175
    },
    {
      "entropy": 6.2605797290802006,
      "epoch": 2.3289094056549335,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004465848044056361,
      "loss": 5.6407,
      "mean_token_accuracy": 0.20388253331184386,
      "num_tokens": 51145978.0,
      "step": 20180
    },
    {
      "entropy": 6.095645475387573,
      "epoch": 2.3294864396999424,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004465581087925028,
      "loss": 5.5517,
      "mean_token_accuracy": 0.21137773096561432,
      "num_tokens": 51158016.0,
      "step": 20185
    },
    {
      "entropy": 6.243349504470825,
      "epoch": 2.330063473744951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004465314074090978,
      "loss": 5.6553,
      "mean_token_accuracy": 0.2074376255273819,
      "num_tokens": 51170222.0,
      "step": 20190
    },
    {
      "entropy": 6.268354272842407,
      "epoch": 2.3306405077899597,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044650470025631904,
      "loss": 5.6328,
      "mean_token_accuracy": 0.20250659584999084,
      "num_tokens": 51183560.0,
      "step": 20195
    },
    {
      "entropy": 6.191282033920288,
      "epoch": 2.331217541834968,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004464779873350649,
      "loss": 5.6318,
      "mean_token_accuracy": 0.20710641592741014,
      "num_tokens": 51195532.0,
      "step": 20200
    },
    {
      "entropy": 6.277365684509277,
      "epoch": 2.331794575879977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004464512686462339,
      "loss": 5.7021,
      "mean_token_accuracy": 0.19780575335025788,
      "num_tokens": 51207078.0,
      "step": 20205
    },
    {
      "entropy": 6.256212377548218,
      "epoch": 2.3323716099249854,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044642454419072455,
      "loss": 5.7052,
      "mean_token_accuracy": 0.19188258647918702,
      "num_tokens": 51220642.0,
      "step": 20210
    },
    {
      "entropy": 6.265449142456054,
      "epoch": 2.3329486439699942,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004463978139694358,
      "loss": 5.6808,
      "mean_token_accuracy": 0.20078595131635665,
      "num_tokens": 51234127.0,
      "step": 20215
    },
    {
      "entropy": 6.24262113571167,
      "epoch": 2.333525678015003,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044637107798326675,
      "loss": 5.6427,
      "mean_token_accuracy": 0.2048109143972397,
      "num_tokens": 51246645.0,
      "step": 20220
    },
    {
      "entropy": 6.177686405181885,
      "epoch": 2.3341027120600115,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004463443362331166,
      "loss": 5.5785,
      "mean_token_accuracy": 0.21757539361715317,
      "num_tokens": 51259392.0,
      "step": 20225
    },
    {
      "entropy": 6.19996485710144,
      "epoch": 2.3346797461050204,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044631758871988486,
      "loss": 5.6291,
      "mean_token_accuracy": 0.20728014260530472,
      "num_tokens": 51273073.0,
      "step": 20230
    },
    {
      "entropy": 6.250855588912964,
      "epoch": 2.335256780150029,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004462908354444711,
      "loss": 5.6512,
      "mean_token_accuracy": 0.20559770613908768,
      "num_tokens": 51286350.0,
      "step": 20235
    },
    {
      "entropy": 6.172985410690307,
      "epoch": 2.3358338141950377,
      "grad_norm": 1.125,
      "learning_rate": 0.0004462640764077751,
      "loss": 5.5579,
      "mean_token_accuracy": 0.2123764455318451,
      "num_tokens": 51298262.0,
      "step": 20240
    },
    {
      "entropy": 6.28583025932312,
      "epoch": 2.336410848240046,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004462373116106971,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1995835542678833,
      "num_tokens": 51310162.0,
      "step": 20245
    },
    {
      "entropy": 6.208287906646729,
      "epoch": 2.336987882285055,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044621054105413704,
      "loss": 5.5873,
      "mean_token_accuracy": 0.20805788338184356,
      "num_tokens": 51323468.0,
      "step": 20250
    },
    {
      "entropy": 6.291377162933349,
      "epoch": 2.3375649163300634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044618376473899555,
      "loss": 5.6746,
      "mean_token_accuracy": 0.199124713242054,
      "num_tokens": 51336674.0,
      "step": 20255
    },
    {
      "entropy": 6.258267211914062,
      "epoch": 2.338141950375072,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004461569826661732,
      "loss": 5.7142,
      "mean_token_accuracy": 0.19726691991090775,
      "num_tokens": 51349950.0,
      "step": 20260
    },
    {
      "entropy": 6.243049621582031,
      "epoch": 2.3387189844200806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004461301948365707,
      "loss": 5.6959,
      "mean_token_accuracy": 0.2013692229986191,
      "num_tokens": 51362191.0,
      "step": 20265
    },
    {
      "entropy": 6.211537837982178,
      "epoch": 2.3392960184650895,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004461034012510891,
      "loss": 5.6169,
      "mean_token_accuracy": 0.20620788782835006,
      "num_tokens": 51374325.0,
      "step": 20270
    },
    {
      "entropy": 6.120653343200684,
      "epoch": 2.339873052510098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044607660191062963,
      "loss": 5.4956,
      "mean_token_accuracy": 0.21309128701686858,
      "num_tokens": 51386632.0,
      "step": 20275
    },
    {
      "entropy": 6.2541351318359375,
      "epoch": 2.340450086555107,
      "grad_norm": 1.0,
      "learning_rate": 0.00044604979681609364,
      "loss": 5.7402,
      "mean_token_accuracy": 0.19787564724683762,
      "num_tokens": 51397876.0,
      "step": 20280
    },
    {
      "entropy": 6.246520185470581,
      "epoch": 2.341027120600115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044602298596838264,
      "loss": 5.6462,
      "mean_token_accuracy": 0.2068067580461502,
      "num_tokens": 51411466.0,
      "step": 20285
    },
    {
      "entropy": 6.214586973190308,
      "epoch": 2.341604154645124,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044599616936839853,
      "loss": 5.5652,
      "mean_token_accuracy": 0.20829868465662002,
      "num_tokens": 51423241.0,
      "step": 20290
    },
    {
      "entropy": 6.21762523651123,
      "epoch": 2.342181188690133,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004459693470170432,
      "loss": 5.6707,
      "mean_token_accuracy": 0.2072637990117073,
      "num_tokens": 51436056.0,
      "step": 20295
    },
    {
      "entropy": 6.289853858947754,
      "epoch": 2.3427582227351413,
      "grad_norm": 1.125,
      "learning_rate": 0.0004459425189152187,
      "loss": 5.679,
      "mean_token_accuracy": 0.20727627724409103,
      "num_tokens": 51449986.0,
      "step": 20300
    },
    {
      "entropy": 6.056736040115356,
      "epoch": 2.34333525678015,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044591568506382757,
      "loss": 5.4803,
      "mean_token_accuracy": 0.2252707690000534,
      "num_tokens": 51463784.0,
      "step": 20305
    },
    {
      "entropy": 6.146295166015625,
      "epoch": 2.3439122908251586,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044588884546377226,
      "loss": 5.5487,
      "mean_token_accuracy": 0.22001761347055435,
      "num_tokens": 51476772.0,
      "step": 20310
    },
    {
      "entropy": 6.225093841552734,
      "epoch": 2.3444893248701675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004458620001159555,
      "loss": 5.6751,
      "mean_token_accuracy": 0.20063820332288743,
      "num_tokens": 51489889.0,
      "step": 20315
    },
    {
      "entropy": 6.22968168258667,
      "epoch": 2.345066358915176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004458351490212803,
      "loss": 5.6662,
      "mean_token_accuracy": 0.20106541961431504,
      "num_tokens": 51503462.0,
      "step": 20320
    },
    {
      "entropy": 6.220002269744873,
      "epoch": 2.3456433929601848,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044580829218064964,
      "loss": 5.5684,
      "mean_token_accuracy": 0.2107448771595955,
      "num_tokens": 51515338.0,
      "step": 20325
    },
    {
      "entropy": 6.1662609577178955,
      "epoch": 2.346220427005193,
      "grad_norm": 1.046875,
      "learning_rate": 0.000445781429594967,
      "loss": 5.6026,
      "mean_token_accuracy": 0.209906005859375,
      "num_tokens": 51527990.0,
      "step": 20330
    },
    {
      "entropy": 6.224651098251343,
      "epoch": 2.346797461050202,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004457545612651357,
      "loss": 5.6715,
      "mean_token_accuracy": 0.1988444834947586,
      "num_tokens": 51542316.0,
      "step": 20335
    },
    {
      "entropy": 6.214133358001709,
      "epoch": 2.3473744950952105,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044572768719205964,
      "loss": 5.6321,
      "mean_token_accuracy": 0.19868172258138656,
      "num_tokens": 51556268.0,
      "step": 20340
    },
    {
      "entropy": 6.3023186206817625,
      "epoch": 2.3479515291402193,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044570080737664264,
      "loss": 5.7039,
      "mean_token_accuracy": 0.20154277235269547,
      "num_tokens": 51568511.0,
      "step": 20345
    },
    {
      "entropy": 6.2537730693817135,
      "epoch": 2.3485285631852277,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044567392181978874,
      "loss": 5.7158,
      "mean_token_accuracy": 0.20214542150497436,
      "num_tokens": 51581915.0,
      "step": 20350
    },
    {
      "entropy": 6.206254243850708,
      "epoch": 2.3491055972302366,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044564703052240223,
      "loss": 5.5447,
      "mean_token_accuracy": 0.2054579436779022,
      "num_tokens": 51593544.0,
      "step": 20355
    },
    {
      "entropy": 6.19595627784729,
      "epoch": 2.349682631275245,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004456201334853876,
      "loss": 5.642,
      "mean_token_accuracy": 0.20682139992713927,
      "num_tokens": 51605722.0,
      "step": 20360
    },
    {
      "entropy": 6.161370038986206,
      "epoch": 2.350259665320254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044559323070964956,
      "loss": 5.5446,
      "mean_token_accuracy": 0.21595069169998168,
      "num_tokens": 51618277.0,
      "step": 20365
    },
    {
      "entropy": 6.24941234588623,
      "epoch": 2.3508366993652627,
      "grad_norm": 1.015625,
      "learning_rate": 0.000445566322196093,
      "loss": 5.624,
      "mean_token_accuracy": 0.2115999087691307,
      "num_tokens": 51629960.0,
      "step": 20370
    },
    {
      "entropy": 6.118126964569091,
      "epoch": 2.351413733410271,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004455394079456228,
      "loss": 5.5174,
      "mean_token_accuracy": 0.21368240416049958,
      "num_tokens": 51643089.0,
      "step": 20375
    },
    {
      "entropy": 6.28010573387146,
      "epoch": 2.35199076745528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044551248795914446,
      "loss": 5.7344,
      "mean_token_accuracy": 0.19721965193748475,
      "num_tokens": 51655581.0,
      "step": 20380
    },
    {
      "entropy": 6.235591697692871,
      "epoch": 2.3525678015002884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004454855622375632,
      "loss": 5.6152,
      "mean_token_accuracy": 0.2024390920996666,
      "num_tokens": 51667693.0,
      "step": 20385
    },
    {
      "entropy": 6.2534552097320555,
      "epoch": 2.3531448355452973,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004454586307817848,
      "loss": 5.6749,
      "mean_token_accuracy": 0.20312736928462982,
      "num_tokens": 51680479.0,
      "step": 20390
    },
    {
      "entropy": 6.26989631652832,
      "epoch": 2.3537218695903057,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000445431693592715,
      "loss": 5.7448,
      "mean_token_accuracy": 0.1991966634988785,
      "num_tokens": 51693053.0,
      "step": 20395
    },
    {
      "entropy": 6.255859231948852,
      "epoch": 2.3542989036353146,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004454047506712598,
      "loss": 5.6958,
      "mean_token_accuracy": 0.2003437399864197,
      "num_tokens": 51704974.0,
      "step": 20400
    },
    {
      "entropy": 6.248658514022827,
      "epoch": 2.354875937680323,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044537780201832545,
      "loss": 5.604,
      "mean_token_accuracy": 0.2076022893190384,
      "num_tokens": 51717190.0,
      "step": 20405
    },
    {
      "entropy": 6.216630506515503,
      "epoch": 2.355452971725332,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004453508476348184,
      "loss": 5.6495,
      "mean_token_accuracy": 0.2045443907380104,
      "num_tokens": 51730462.0,
      "step": 20410
    },
    {
      "entropy": 6.25622968673706,
      "epoch": 2.3560300057703403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004453238875216452,
      "loss": 5.7166,
      "mean_token_accuracy": 0.1982462599873543,
      "num_tokens": 51743723.0,
      "step": 20415
    },
    {
      "entropy": 6.2455785274505615,
      "epoch": 2.356607039815349,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004452969216797127,
      "loss": 5.6411,
      "mean_token_accuracy": 0.20469288676977157,
      "num_tokens": 51756959.0,
      "step": 20420
    },
    {
      "entropy": 6.208768653869629,
      "epoch": 2.357184073860358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004452699501099277,
      "loss": 5.6845,
      "mean_token_accuracy": 0.19687082171440123,
      "num_tokens": 51768282.0,
      "step": 20425
    },
    {
      "entropy": 6.291310453414917,
      "epoch": 2.3577611079053664,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044524297281319766,
      "loss": 5.695,
      "mean_token_accuracy": 0.20112762302160264,
      "num_tokens": 51780729.0,
      "step": 20430
    },
    {
      "entropy": 6.313424587249756,
      "epoch": 2.358338141950375,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044521598979042963,
      "loss": 5.7606,
      "mean_token_accuracy": 0.19430041015148164,
      "num_tokens": 51793661.0,
      "step": 20435
    },
    {
      "entropy": 6.308770561218262,
      "epoch": 2.3589151759953837,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044518900104253154,
      "loss": 5.7046,
      "mean_token_accuracy": 0.2027783825993538,
      "num_tokens": 51806027.0,
      "step": 20440
    },
    {
      "entropy": 6.269329786300659,
      "epoch": 2.3594922100403926,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004451620065704108,
      "loss": 5.7498,
      "mean_token_accuracy": 0.1959427371621132,
      "num_tokens": 51817495.0,
      "step": 20445
    },
    {
      "entropy": 6.237169694900513,
      "epoch": 2.360069244085401,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044513500637497546,
      "loss": 5.7045,
      "mean_token_accuracy": 0.19854709804058074,
      "num_tokens": 51829948.0,
      "step": 20450
    },
    {
      "entropy": 6.277944564819336,
      "epoch": 2.36064627813041,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044510800045713373,
      "loss": 5.7427,
      "mean_token_accuracy": 0.19941256046295167,
      "num_tokens": 51842615.0,
      "step": 20455
    },
    {
      "entropy": 6.31951003074646,
      "epoch": 2.3612233121754183,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044508098881779396,
      "loss": 5.6378,
      "mean_token_accuracy": 0.21265835613012313,
      "num_tokens": 51853785.0,
      "step": 20460
    },
    {
      "entropy": 6.203953409194947,
      "epoch": 2.361800346220427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004450539714578645,
      "loss": 5.5438,
      "mean_token_accuracy": 0.21673437654972078,
      "num_tokens": 51867427.0,
      "step": 20465
    },
    {
      "entropy": 6.32550082206726,
      "epoch": 2.3623773802654355,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004450269483782543,
      "loss": 5.8172,
      "mean_token_accuracy": 0.18711897730827332,
      "num_tokens": 51880241.0,
      "step": 20470
    },
    {
      "entropy": 6.213586044311524,
      "epoch": 2.3629544143104444,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004449999195798721,
      "loss": 5.6617,
      "mean_token_accuracy": 0.20396423637866973,
      "num_tokens": 51891927.0,
      "step": 20475
    },
    {
      "entropy": 6.237623500823974,
      "epoch": 2.363531448355453,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044497288506362706,
      "loss": 5.683,
      "mean_token_accuracy": 0.2023163065314293,
      "num_tokens": 51905633.0,
      "step": 20480
    },
    {
      "entropy": 6.269557905197144,
      "epoch": 2.3641084824004617,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004449458448304284,
      "loss": 5.6987,
      "mean_token_accuracy": 0.19690544605255128,
      "num_tokens": 51919398.0,
      "step": 20485
    },
    {
      "entropy": 6.119033432006836,
      "epoch": 2.36468551644547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044491879888118574,
      "loss": 5.4803,
      "mean_token_accuracy": 0.22232985496520996,
      "num_tokens": 51932732.0,
      "step": 20490
    },
    {
      "entropy": 6.164668130874634,
      "epoch": 2.365262550490479,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004448917472168087,
      "loss": 5.5814,
      "mean_token_accuracy": 0.20964327901601792,
      "num_tokens": 51945830.0,
      "step": 20495
    },
    {
      "entropy": 6.2759322166442875,
      "epoch": 2.365839584535488,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044486468983820707,
      "loss": 5.603,
      "mean_token_accuracy": 0.20815862715244293,
      "num_tokens": 51957643.0,
      "step": 20500
    },
    {
      "entropy": 6.220407342910766,
      "epoch": 2.3664166185804962,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000444837626746291,
      "loss": 5.5739,
      "mean_token_accuracy": 0.20977915078401566,
      "num_tokens": 51969840.0,
      "step": 20505
    },
    {
      "entropy": 6.240037298202514,
      "epoch": 2.366993652625505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004448105579419707,
      "loss": 5.6981,
      "mean_token_accuracy": 0.19629756063222886,
      "num_tokens": 51981250.0,
      "step": 20510
    },
    {
      "entropy": 6.284100961685181,
      "epoch": 2.3675706866705135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004447834834261567,
      "loss": 5.6845,
      "mean_token_accuracy": 0.20061818212270738,
      "num_tokens": 51993312.0,
      "step": 20515
    },
    {
      "entropy": 6.187756586074829,
      "epoch": 2.3681477207155224,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004447564031997595,
      "loss": 5.6264,
      "mean_token_accuracy": 0.20650881081819533,
      "num_tokens": 52006697.0,
      "step": 20520
    },
    {
      "entropy": 6.207859706878662,
      "epoch": 2.368724754760531,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044472931726369,
      "loss": 5.5848,
      "mean_token_accuracy": 0.20812041014432908,
      "num_tokens": 52018938.0,
      "step": 20525
    },
    {
      "entropy": 6.167420768737793,
      "epoch": 2.3693017888055397,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044470222561885926,
      "loss": 5.5927,
      "mean_token_accuracy": 0.2115847334265709,
      "num_tokens": 52031309.0,
      "step": 20530
    },
    {
      "entropy": 6.251200437545776,
      "epoch": 2.369878822850548,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044467512826617845,
      "loss": 5.6581,
      "mean_token_accuracy": 0.20321357548236846,
      "num_tokens": 52043661.0,
      "step": 20535
    },
    {
      "entropy": 6.225104856491089,
      "epoch": 2.370455856895557,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044464802520655897,
      "loss": 5.6307,
      "mean_token_accuracy": 0.2067578077316284,
      "num_tokens": 52055049.0,
      "step": 20540
    },
    {
      "entropy": 6.1780791759490965,
      "epoch": 2.3710328909405654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004446209164409124,
      "loss": 5.6435,
      "mean_token_accuracy": 0.2026913583278656,
      "num_tokens": 52068188.0,
      "step": 20545
    },
    {
      "entropy": 6.264406728744507,
      "epoch": 2.371609924985574,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004445938019701506,
      "loss": 5.6765,
      "mean_token_accuracy": 0.20910051167011262,
      "num_tokens": 52080662.0,
      "step": 20550
    },
    {
      "entropy": 6.272018766403198,
      "epoch": 2.3721869590305826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004445666817951855,
      "loss": 5.699,
      "mean_token_accuracy": 0.20190092474222182,
      "num_tokens": 52093088.0,
      "step": 20555
    },
    {
      "entropy": 6.19417176246643,
      "epoch": 2.3727639930755915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445395559169293,
      "loss": 5.5948,
      "mean_token_accuracy": 0.21268751323223115,
      "num_tokens": 52105796.0,
      "step": 20560
    },
    {
      "entropy": 6.23984375,
      "epoch": 2.3733410271206,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004445124243362943,
      "loss": 5.6719,
      "mean_token_accuracy": 0.20233068317174913,
      "num_tokens": 52118146.0,
      "step": 20565
    },
    {
      "entropy": 6.199374389648438,
      "epoch": 2.3739180611656088,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004444852870541932,
      "loss": 5.5862,
      "mean_token_accuracy": 0.2131514698266983,
      "num_tokens": 52131033.0,
      "step": 20570
    },
    {
      "entropy": 6.257831621170044,
      "epoch": 2.3744950952106176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004444581440715386,
      "loss": 5.7114,
      "mean_token_accuracy": 0.19805960357189178,
      "num_tokens": 52143742.0,
      "step": 20575
    },
    {
      "entropy": 6.189561367034912,
      "epoch": 2.375072129255626,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044443099538924347,
      "loss": 5.5697,
      "mean_token_accuracy": 0.2104736864566803,
      "num_tokens": 52156091.0,
      "step": 20580
    },
    {
      "entropy": 6.29191780090332,
      "epoch": 2.375649163300635,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004444038410082211,
      "loss": 5.7261,
      "mean_token_accuracy": 0.196769680082798,
      "num_tokens": 52168404.0,
      "step": 20585
    },
    {
      "entropy": 6.194567251205444,
      "epoch": 2.3762261973456433,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004443766809293846,
      "loss": 5.5752,
      "mean_token_accuracy": 0.21289655566215515,
      "num_tokens": 52180660.0,
      "step": 20590
    },
    {
      "entropy": 6.2320091247558596,
      "epoch": 2.376803231390652,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004443495151536475,
      "loss": 5.657,
      "mean_token_accuracy": 0.2089839607477188,
      "num_tokens": 52193752.0,
      "step": 20595
    },
    {
      "entropy": 6.323664283752441,
      "epoch": 2.3773802654356606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004443223436819237,
      "loss": 5.7049,
      "mean_token_accuracy": 0.20083066523075105,
      "num_tokens": 52205686.0,
      "step": 20600
    },
    {
      "entropy": 6.253013944625854,
      "epoch": 2.3779572994806695,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044429516651512687,
      "loss": 5.6981,
      "mean_token_accuracy": 0.1995725080370903,
      "num_tokens": 52218743.0,
      "step": 20605
    },
    {
      "entropy": 6.2568464279174805,
      "epoch": 2.378534333525678,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044426798365417133,
      "loss": 5.5282,
      "mean_token_accuracy": 0.21039217710494995,
      "num_tokens": 52231760.0,
      "step": 20610
    },
    {
      "entropy": 6.269671249389648,
      "epoch": 2.3791113675706868,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044424079509997104,
      "loss": 5.6853,
      "mean_token_accuracy": 0.2015765592455864,
      "num_tokens": 52243146.0,
      "step": 20615
    },
    {
      "entropy": 6.235985946655274,
      "epoch": 2.379688401615695,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004442136008534409,
      "loss": 5.695,
      "mean_token_accuracy": 0.1999841585755348,
      "num_tokens": 52255330.0,
      "step": 20620
    },
    {
      "entropy": 6.24383511543274,
      "epoch": 2.380265435660704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044418640091549525,
      "loss": 5.6935,
      "mean_token_accuracy": 0.20174667984247208,
      "num_tokens": 52266563.0,
      "step": 20625
    },
    {
      "entropy": 6.260625553131104,
      "epoch": 2.3808424697057124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000444159195287049,
      "loss": 5.7374,
      "mean_token_accuracy": 0.1974148765206337,
      "num_tokens": 52279137.0,
      "step": 20630
    },
    {
      "entropy": 6.22347240447998,
      "epoch": 2.3814195037507213,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004441319839690173,
      "loss": 5.6457,
      "mean_token_accuracy": 0.2098652571439743,
      "num_tokens": 52293317.0,
      "step": 20635
    },
    {
      "entropy": 6.316980218887329,
      "epoch": 2.3819965377957297,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004441047669623153,
      "loss": 5.7047,
      "mean_token_accuracy": 0.19738261997699738,
      "num_tokens": 52306383.0,
      "step": 20640
    },
    {
      "entropy": 6.202902603149414,
      "epoch": 2.3825735718407386,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044407754426785845,
      "loss": 5.5882,
      "mean_token_accuracy": 0.20461847633123398,
      "num_tokens": 52318446.0,
      "step": 20645
    },
    {
      "entropy": 6.2510169506072994,
      "epoch": 2.3831506058857475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004440503158865625,
      "loss": 5.6152,
      "mean_token_accuracy": 0.19937023222446443,
      "num_tokens": 52330761.0,
      "step": 20650
    },
    {
      "entropy": 6.330129671096802,
      "epoch": 2.383727639930756,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044402308181934295,
      "loss": 5.695,
      "mean_token_accuracy": 0.19718821942806244,
      "num_tokens": 52342906.0,
      "step": 20655
    },
    {
      "entropy": 6.2026081562042235,
      "epoch": 2.3843046739757647,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004439958420671162,
      "loss": 5.6451,
      "mean_token_accuracy": 0.20673254877328873,
      "num_tokens": 52355928.0,
      "step": 20660
    },
    {
      "entropy": 6.227880191802979,
      "epoch": 2.384881708020773,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044396859663079814,
      "loss": 5.6748,
      "mean_token_accuracy": 0.19866825491189957,
      "num_tokens": 52368370.0,
      "step": 20665
    },
    {
      "entropy": 6.3192863941192625,
      "epoch": 2.385458742065782,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044394134551130533,
      "loss": 5.71,
      "mean_token_accuracy": 0.19565349370241164,
      "num_tokens": 52382636.0,
      "step": 20670
    },
    {
      "entropy": 6.229170417785644,
      "epoch": 2.3860357761107904,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004439140887095542,
      "loss": 5.6299,
      "mean_token_accuracy": 0.2101860299706459,
      "num_tokens": 52395745.0,
      "step": 20675
    },
    {
      "entropy": 6.258525371551514,
      "epoch": 2.3866128101557993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044388682622646165,
      "loss": 5.6103,
      "mean_token_accuracy": 0.20800065100193024,
      "num_tokens": 52408929.0,
      "step": 20680
    },
    {
      "entropy": 6.20223183631897,
      "epoch": 2.3871898442008077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004438595580629446,
      "loss": 5.5955,
      "mean_token_accuracy": 0.2040373846888542,
      "num_tokens": 52421626.0,
      "step": 20685
    },
    {
      "entropy": 6.339115333557129,
      "epoch": 2.3877668782458166,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004438322842199202,
      "loss": 5.7508,
      "mean_token_accuracy": 0.19360657632350922,
      "num_tokens": 52434139.0,
      "step": 20690
    },
    {
      "entropy": 6.249014902114868,
      "epoch": 2.388343912290825,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004438050046983057,
      "loss": 5.6366,
      "mean_token_accuracy": 0.2051589459180832,
      "num_tokens": 52446141.0,
      "step": 20695
    },
    {
      "entropy": 6.251748895645141,
      "epoch": 2.388920946335834,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00044377771949901876,
      "loss": 5.6771,
      "mean_token_accuracy": 0.2008539319038391,
      "num_tokens": 52459110.0,
      "step": 20700
    },
    {
      "entropy": 6.272036123275757,
      "epoch": 2.3894979803808427,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044375042862297703,
      "loss": 5.7793,
      "mean_token_accuracy": 0.19308315664529802,
      "num_tokens": 52471985.0,
      "step": 20705
    },
    {
      "entropy": 6.306207799911499,
      "epoch": 2.390075014425851,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044372313207109856,
      "loss": 5.7078,
      "mean_token_accuracy": 0.2025204688310623,
      "num_tokens": 52484226.0,
      "step": 20710
    },
    {
      "entropy": 6.243333387374878,
      "epoch": 2.3906520484708595,
      "grad_norm": 1.0,
      "learning_rate": 0.00044369582984430127,
      "loss": 5.6596,
      "mean_token_accuracy": 0.20185588598251342,
      "num_tokens": 52496141.0,
      "step": 20715
    },
    {
      "entropy": 6.233908081054688,
      "epoch": 2.3912290825158684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044366852194350354,
      "loss": 5.6774,
      "mean_token_accuracy": 0.2013825923204422,
      "num_tokens": 52509231.0,
      "step": 20720
    },
    {
      "entropy": 6.240170335769653,
      "epoch": 2.3918061165608773,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004436412083696239,
      "loss": 5.6039,
      "mean_token_accuracy": 0.20638748854398728,
      "num_tokens": 52522619.0,
      "step": 20725
    },
    {
      "entropy": 6.306197309494019,
      "epoch": 2.3923831506058857,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044361388912358095,
      "loss": 5.6782,
      "mean_token_accuracy": 0.20261909067630768,
      "num_tokens": 52534241.0,
      "step": 20730
    },
    {
      "entropy": 6.330400180816651,
      "epoch": 2.3929601846508946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004435865642062936,
      "loss": 5.6454,
      "mean_token_accuracy": 0.20136924982070922,
      "num_tokens": 52546306.0,
      "step": 20735
    },
    {
      "entropy": 6.282812929153442,
      "epoch": 2.393537218695903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004435592336186809,
      "loss": 5.7407,
      "mean_token_accuracy": 0.19723884612321854,
      "num_tokens": 52559855.0,
      "step": 20740
    },
    {
      "entropy": 6.135396432876587,
      "epoch": 2.394114252740912,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004435318973616622,
      "loss": 5.5354,
      "mean_token_accuracy": 0.21988216191530227,
      "num_tokens": 52574071.0,
      "step": 20745
    },
    {
      "entropy": 6.247970247268677,
      "epoch": 2.3946912867859202,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044350455543615665,
      "loss": 5.6426,
      "mean_token_accuracy": 0.20348259508609773,
      "num_tokens": 52585983.0,
      "step": 20750
    },
    {
      "entropy": 6.243813371658325,
      "epoch": 2.395268320830929,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004434772078430843,
      "loss": 5.5602,
      "mean_token_accuracy": 0.2071308448910713,
      "num_tokens": 52598399.0,
      "step": 20755
    },
    {
      "entropy": 6.263616895675659,
      "epoch": 2.3958453548759375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004434498545833646,
      "loss": 5.691,
      "mean_token_accuracy": 0.19686609953641893,
      "num_tokens": 52610172.0,
      "step": 20760
    },
    {
      "entropy": 6.23023419380188,
      "epoch": 2.3964223889209464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004434224956579177,
      "loss": 5.6226,
      "mean_token_accuracy": 0.21566786020994186,
      "num_tokens": 52622626.0,
      "step": 20765
    },
    {
      "entropy": 6.266142654418945,
      "epoch": 2.396999422965955,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004433951310676639,
      "loss": 5.6658,
      "mean_token_accuracy": 0.20079080164432525,
      "num_tokens": 52633893.0,
      "step": 20770
    },
    {
      "entropy": 6.2647566318511965,
      "epoch": 2.3975764570109637,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044336776081352347,
      "loss": 5.6694,
      "mean_token_accuracy": 0.19948896318674086,
      "num_tokens": 52645674.0,
      "step": 20775
    },
    {
      "entropy": 6.296689796447754,
      "epoch": 2.3981534910559725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044334038489641706,
      "loss": 5.7212,
      "mean_token_accuracy": 0.20096069425344468,
      "num_tokens": 52659030.0,
      "step": 20780
    },
    {
      "entropy": 6.262747716903687,
      "epoch": 2.398730525100981,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044331300331726544,
      "loss": 5.6424,
      "mean_token_accuracy": 0.20609911233186723,
      "num_tokens": 52670680.0,
      "step": 20785
    },
    {
      "entropy": 6.33933391571045,
      "epoch": 2.3993075591459894,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044328561607698947,
      "loss": 5.7691,
      "mean_token_accuracy": 0.19635584503412246,
      "num_tokens": 52682860.0,
      "step": 20790
    },
    {
      "entropy": 6.189906215667724,
      "epoch": 2.3998845931909982,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004432582231765105,
      "loss": 5.6152,
      "mean_token_accuracy": 0.20150961279869078,
      "num_tokens": 52695566.0,
      "step": 20795
    },
    {
      "entropy": 6.205738258361817,
      "epoch": 2.400461627236007,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044323082461674974,
      "loss": 5.5686,
      "mean_token_accuracy": 0.2133536621928215,
      "num_tokens": 52708409.0,
      "step": 20800
    },
    {
      "entropy": 6.2242083072662355,
      "epoch": 2.4010386612810155,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004432034203986287,
      "loss": 5.6071,
      "mean_token_accuracy": 0.20954781919717788,
      "num_tokens": 52721637.0,
      "step": 20805
    },
    {
      "entropy": 6.263833951950073,
      "epoch": 2.4016156953260244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004431760105230693,
      "loss": 5.6939,
      "mean_token_accuracy": 0.19896161705255508,
      "num_tokens": 52732157.0,
      "step": 20810
    },
    {
      "entropy": 6.238463640213013,
      "epoch": 2.402192729371033,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044314859499099325,
      "loss": 5.678,
      "mean_token_accuracy": 0.2026016116142273,
      "num_tokens": 52745038.0,
      "step": 20815
    },
    {
      "entropy": 6.28553032875061,
      "epoch": 2.4027697634160416,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044312117380332274,
      "loss": 5.6259,
      "mean_token_accuracy": 0.2097138911485672,
      "num_tokens": 52757381.0,
      "step": 20820
    },
    {
      "entropy": 6.160345840454101,
      "epoch": 2.40334679746105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044309374696098,
      "loss": 5.5806,
      "mean_token_accuracy": 0.21553289294242858,
      "num_tokens": 52770604.0,
      "step": 20825
    },
    {
      "entropy": 6.185074281692505,
      "epoch": 2.403923831506059,
      "grad_norm": 1.0,
      "learning_rate": 0.0004430663144648876,
      "loss": 5.5834,
      "mean_token_accuracy": 0.21756615340709687,
      "num_tokens": 52784634.0,
      "step": 20830
    },
    {
      "entropy": 6.2482646942138675,
      "epoch": 2.4045008655510673,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044303887631596823,
      "loss": 5.6334,
      "mean_token_accuracy": 0.20515166968107224,
      "num_tokens": 52798417.0,
      "step": 20835
    },
    {
      "entropy": 6.348049640655518,
      "epoch": 2.405077899596076,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004430114325151447,
      "loss": 5.7919,
      "mean_token_accuracy": 0.19508688896894455,
      "num_tokens": 52810885.0,
      "step": 20840
    },
    {
      "entropy": 6.200775575637818,
      "epoch": 2.4056549336410846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004429839830633401,
      "loss": 5.6318,
      "mean_token_accuracy": 0.2083239808678627,
      "num_tokens": 52824450.0,
      "step": 20845
    },
    {
      "entropy": 6.235306692123413,
      "epoch": 2.4062319676860935,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004429565279614777,
      "loss": 5.6535,
      "mean_token_accuracy": 0.20343738049268723,
      "num_tokens": 52837053.0,
      "step": 20850
    },
    {
      "entropy": 6.217371082305908,
      "epoch": 2.4068090017311023,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044292906721048094,
      "loss": 5.6501,
      "mean_token_accuracy": 0.20233725011348724,
      "num_tokens": 52851695.0,
      "step": 20855
    },
    {
      "entropy": 6.299347639083862,
      "epoch": 2.4073860357761108,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004429016008112733,
      "loss": 5.6653,
      "mean_token_accuracy": 0.19931307584047317,
      "num_tokens": 52862999.0,
      "step": 20860
    },
    {
      "entropy": 6.151570558547974,
      "epoch": 2.4079630698211196,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004428741287647788,
      "loss": 5.5858,
      "mean_token_accuracy": 0.2115140438079834,
      "num_tokens": 52875224.0,
      "step": 20865
    },
    {
      "entropy": 6.239431571960449,
      "epoch": 2.408540103866128,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044284665107192136,
      "loss": 5.6871,
      "mean_token_accuracy": 0.198722206056118,
      "num_tokens": 52887296.0,
      "step": 20870
    },
    {
      "entropy": 6.257601165771485,
      "epoch": 2.409117137911137,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004428191677336251,
      "loss": 5.5971,
      "mean_token_accuracy": 0.20384119749069213,
      "num_tokens": 52900670.0,
      "step": 20875
    },
    {
      "entropy": 6.280495357513428,
      "epoch": 2.4096941719561453,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004427916787508146,
      "loss": 5.7507,
      "mean_token_accuracy": 0.200733482837677,
      "num_tokens": 52914515.0,
      "step": 20880
    },
    {
      "entropy": 6.161097812652588,
      "epoch": 2.410271206001154,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004427641841244144,
      "loss": 5.5656,
      "mean_token_accuracy": 0.21280494928359986,
      "num_tokens": 52928448.0,
      "step": 20885
    },
    {
      "entropy": 6.265078592300415,
      "epoch": 2.4108482400461626,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000442736683855349,
      "loss": 5.6537,
      "mean_token_accuracy": 0.20518819689750672,
      "num_tokens": 52941381.0,
      "step": 20890
    },
    {
      "entropy": 6.198981237411499,
      "epoch": 2.4114252740911715,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004427091779445437,
      "loss": 5.6001,
      "mean_token_accuracy": 0.19855600744485855,
      "num_tokens": 52954959.0,
      "step": 20895
    },
    {
      "entropy": 6.244168472290039,
      "epoch": 2.41200230813618,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004426816663929235,
      "loss": 5.628,
      "mean_token_accuracy": 0.2065839037299156,
      "num_tokens": 52967956.0,
      "step": 20900
    },
    {
      "entropy": 6.271139001846313,
      "epoch": 2.4125793421811887,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044265414920141366,
      "loss": 5.6933,
      "mean_token_accuracy": 0.20381143391132356,
      "num_tokens": 52980997.0,
      "step": 20905
    },
    {
      "entropy": 6.276428890228272,
      "epoch": 2.413156376226197,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044262662637093987,
      "loss": 5.6692,
      "mean_token_accuracy": 0.19934424757957458,
      "num_tokens": 52994636.0,
      "step": 20910
    },
    {
      "entropy": 6.225430488586426,
      "epoch": 2.413733410271206,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044259909790242776,
      "loss": 5.6131,
      "mean_token_accuracy": 0.2003405675292015,
      "num_tokens": 53007065.0,
      "step": 20915
    },
    {
      "entropy": 6.164848756790161,
      "epoch": 2.4143104443162144,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004425715637968032,
      "loss": 5.593,
      "mean_token_accuracy": 0.20153367519378662,
      "num_tokens": 53020479.0,
      "step": 20920
    },
    {
      "entropy": 6.260810327529907,
      "epoch": 2.4148874783612233,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004425440240549923,
      "loss": 5.7039,
      "mean_token_accuracy": 0.20422882884740828,
      "num_tokens": 53036390.0,
      "step": 20925
    },
    {
      "entropy": 6.28423171043396,
      "epoch": 2.415464512406232,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044251647867792147,
      "loss": 5.6153,
      "mean_token_accuracy": 0.20164746642112732,
      "num_tokens": 53048604.0,
      "step": 20930
    },
    {
      "entropy": 6.259721088409424,
      "epoch": 2.4160415464512406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044248892766651706,
      "loss": 5.682,
      "mean_token_accuracy": 0.20140611976385117,
      "num_tokens": 53062774.0,
      "step": 20935
    },
    {
      "entropy": 6.235981512069702,
      "epoch": 2.4166185804962494,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004424613710217057,
      "loss": 5.6427,
      "mean_token_accuracy": 0.2072612687945366,
      "num_tokens": 53076701.0,
      "step": 20940
    },
    {
      "entropy": 6.295486545562744,
      "epoch": 2.417195614541258,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044243380874441437,
      "loss": 5.6961,
      "mean_token_accuracy": 0.20581175684928893,
      "num_tokens": 53089950.0,
      "step": 20945
    },
    {
      "entropy": 6.185495090484619,
      "epoch": 2.4177726485862667,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044240624083557,
      "loss": 5.5392,
      "mean_token_accuracy": 0.20899192988872528,
      "num_tokens": 53103622.0,
      "step": 20950
    },
    {
      "entropy": 6.283469486236572,
      "epoch": 2.418349682631275,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044237866729609994,
      "loss": 5.7511,
      "mean_token_accuracy": 0.20393532365560532,
      "num_tokens": 53116382.0,
      "step": 20955
    },
    {
      "entropy": 6.310854768753051,
      "epoch": 2.418926716676284,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004423510881269315,
      "loss": 5.653,
      "mean_token_accuracy": 0.2082198977470398,
      "num_tokens": 53128759.0,
      "step": 20960
    },
    {
      "entropy": 6.210053777694702,
      "epoch": 2.4195037507212924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004423235033289924,
      "loss": 5.651,
      "mean_token_accuracy": 0.21253834962844848,
      "num_tokens": 53141131.0,
      "step": 20965
    },
    {
      "entropy": 6.24911470413208,
      "epoch": 2.4200807847663013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004422959129032103,
      "loss": 5.7141,
      "mean_token_accuracy": 0.19888682961463927,
      "num_tokens": 53152486.0,
      "step": 20970
    },
    {
      "entropy": 6.1909582138061525,
      "epoch": 2.4206578188113097,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044226831685051333,
      "loss": 5.5714,
      "mean_token_accuracy": 0.20512726604938508,
      "num_tokens": 53164897.0,
      "step": 20975
    },
    {
      "entropy": 6.262216424942016,
      "epoch": 2.4212348528563186,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004422407151718296,
      "loss": 5.7223,
      "mean_token_accuracy": 0.1995513767004013,
      "num_tokens": 53177759.0,
      "step": 20980
    },
    {
      "entropy": 6.337183237075806,
      "epoch": 2.4218118869013274,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044221310786808746,
      "loss": 5.679,
      "mean_token_accuracy": 0.20560426265001297,
      "num_tokens": 53191985.0,
      "step": 20985
    },
    {
      "entropy": 6.225973129272461,
      "epoch": 2.422388920946336,
      "grad_norm": 1.0,
      "learning_rate": 0.0004421854949402155,
      "loss": 5.5518,
      "mean_token_accuracy": 0.20679832249879837,
      "num_tokens": 53203531.0,
      "step": 20990
    },
    {
      "entropy": 6.2301177978515625,
      "epoch": 2.4229659549913443,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044215787638914245,
      "loss": 5.6539,
      "mean_token_accuracy": 0.19672989398241042,
      "num_tokens": 53216440.0,
      "step": 20995
    },
    {
      "entropy": 6.275468921661377,
      "epoch": 2.423542989036353,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004421302522157973,
      "loss": 5.7077,
      "mean_token_accuracy": 0.19501957595348357,
      "num_tokens": 53230674.0,
      "step": 21000
    },
    {
      "epoch": 2.423542989036353,
      "eval_entropy": 6.059461270200281,
      "eval_loss": 5.755512714385986,
      "eval_mean_token_accuracy": 0.20611847640736253,
      "eval_num_tokens": 53230674.0,
      "eval_runtime": 17.5493,
      "eval_samples_per_second": 1603.256,
      "eval_steps_per_second": 200.407,
      "step": 21000
    },
    {
      "entropy": 6.289402961730957,
      "epoch": 2.424120023081362,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004421026224211091,
      "loss": 5.7143,
      "mean_token_accuracy": 0.1973523959517479,
      "num_tokens": 53242069.0,
      "step": 21005
    },
    {
      "entropy": 6.1928526878356935,
      "epoch": 2.4246970571263704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044207498700600724,
      "loss": 5.5707,
      "mean_token_accuracy": 0.2183891698718071,
      "num_tokens": 53254803.0,
      "step": 21010
    },
    {
      "entropy": 6.2736715316772464,
      "epoch": 2.4252740911713793,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044204734597142115,
      "loss": 5.7148,
      "mean_token_accuracy": 0.20200884491205215,
      "num_tokens": 53268001.0,
      "step": 21015
    },
    {
      "entropy": 6.229497671127319,
      "epoch": 2.4258511252163877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044201969931828057,
      "loss": 5.6707,
      "mean_token_accuracy": 0.2032482072710991,
      "num_tokens": 53281013.0,
      "step": 21020
    },
    {
      "entropy": 6.285059976577759,
      "epoch": 2.4264281592613965,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004419920470475154,
      "loss": 5.6804,
      "mean_token_accuracy": 0.2059529021382332,
      "num_tokens": 53295714.0,
      "step": 21025
    },
    {
      "entropy": 6.27531328201294,
      "epoch": 2.427005193306405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004419643891600556,
      "loss": 5.6925,
      "mean_token_accuracy": 0.19999144673347474,
      "num_tokens": 53308473.0,
      "step": 21030
    },
    {
      "entropy": 6.290670537948609,
      "epoch": 2.427582227351414,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044193672565683153,
      "loss": 5.742,
      "mean_token_accuracy": 0.19421708583831787,
      "num_tokens": 53321513.0,
      "step": 21035
    },
    {
      "entropy": 6.231978464126587,
      "epoch": 2.4281592613964222,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004419090565387736,
      "loss": 5.643,
      "mean_token_accuracy": 0.20642639100551605,
      "num_tokens": 53334080.0,
      "step": 21040
    },
    {
      "entropy": 6.257345294952392,
      "epoch": 2.428736295441431,
      "grad_norm": 0.875,
      "learning_rate": 0.00044188138180681255,
      "loss": 5.6236,
      "mean_token_accuracy": 0.20663311034440995,
      "num_tokens": 53346986.0,
      "step": 21045
    },
    {
      "entropy": 6.247120332717896,
      "epoch": 2.4293133294864395,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441853701461879,
      "loss": 5.6781,
      "mean_token_accuracy": 0.1969202771782875,
      "num_tokens": 53359283.0,
      "step": 21050
    },
    {
      "entropy": 6.314809274673462,
      "epoch": 2.4298903635314484,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044182601550490414,
      "loss": 5.6428,
      "mean_token_accuracy": 0.20648427456617355,
      "num_tokens": 53372388.0,
      "step": 21055
    },
    {
      "entropy": 6.220915269851685,
      "epoch": 2.4304673975764572,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004417983239368192,
      "loss": 5.6158,
      "mean_token_accuracy": 0.20480076223611832,
      "num_tokens": 53385181.0,
      "step": 21060
    },
    {
      "entropy": 6.150717639923096,
      "epoch": 2.4310444316214657,
      "grad_norm": 1.0,
      "learning_rate": 0.00044177062675855534,
      "loss": 5.6134,
      "mean_token_accuracy": 0.2072327569127083,
      "num_tokens": 53397293.0,
      "step": 21065
    },
    {
      "entropy": 6.202611589431763,
      "epoch": 2.431621465666474,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004417429239710444,
      "loss": 5.5906,
      "mean_token_accuracy": 0.20629713237285613,
      "num_tokens": 53410999.0,
      "step": 21070
    },
    {
      "entropy": 6.309600687026977,
      "epoch": 2.432198499711483,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004417152155752179,
      "loss": 5.697,
      "mean_token_accuracy": 0.20385901480913163,
      "num_tokens": 53422829.0,
      "step": 21075
    },
    {
      "entropy": 6.248817443847656,
      "epoch": 2.432775533756492,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004416875015720081,
      "loss": 5.6456,
      "mean_token_accuracy": 0.21230741292238237,
      "num_tokens": 53434935.0,
      "step": 21080
    },
    {
      "entropy": 6.260274505615234,
      "epoch": 2.4333525678015,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004416597819623469,
      "loss": 5.6518,
      "mean_token_accuracy": 0.20882325768470764,
      "num_tokens": 53448529.0,
      "step": 21085
    },
    {
      "entropy": 6.343727016448975,
      "epoch": 2.433929601846509,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044163205674716666,
      "loss": 5.7312,
      "mean_token_accuracy": 0.19641571789979934,
      "num_tokens": 53461588.0,
      "step": 21090
    },
    {
      "entropy": 6.179997730255127,
      "epoch": 2.4345066358915175,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044160432592740015,
      "loss": 5.6227,
      "mean_token_accuracy": 0.20335731655359268,
      "num_tokens": 53473637.0,
      "step": 21095
    },
    {
      "entropy": 6.242405796051026,
      "epoch": 2.4350836699365264,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044157658950397965,
      "loss": 5.6746,
      "mean_token_accuracy": 0.20028466880321502,
      "num_tokens": 53486070.0,
      "step": 21100
    },
    {
      "entropy": 6.212961721420288,
      "epoch": 2.435660703981535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044154884747783844,
      "loss": 5.5699,
      "mean_token_accuracy": 0.20874811112880706,
      "num_tokens": 53497224.0,
      "step": 21105
    },
    {
      "entropy": 6.223957586288452,
      "epoch": 2.4362377380265436,
      "grad_norm": 1.0,
      "learning_rate": 0.00044152109984990954,
      "loss": 5.622,
      "mean_token_accuracy": 0.20716502964496614,
      "num_tokens": 53509705.0,
      "step": 21110
    },
    {
      "entropy": 6.329930591583252,
      "epoch": 2.436814772071552,
      "grad_norm": 1.0,
      "learning_rate": 0.000441493346621126,
      "loss": 5.7548,
      "mean_token_accuracy": 0.19628022164106368,
      "num_tokens": 53521421.0,
      "step": 21115
    },
    {
      "entropy": 6.240514326095581,
      "epoch": 2.437391806116561,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044146558779242155,
      "loss": 5.6077,
      "mean_token_accuracy": 0.20815987288951873,
      "num_tokens": 53534571.0,
      "step": 21120
    },
    {
      "entropy": 6.277189636230469,
      "epoch": 2.4379688401615693,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004414378233647298,
      "loss": 5.7761,
      "mean_token_accuracy": 0.19118778109550477,
      "num_tokens": 53547411.0,
      "step": 21125
    },
    {
      "entropy": 6.310594272613526,
      "epoch": 2.438545874206578,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044141005333898437,
      "loss": 5.6757,
      "mean_token_accuracy": 0.2046084851026535,
      "num_tokens": 53559998.0,
      "step": 21130
    },
    {
      "entropy": 6.235295677185059,
      "epoch": 2.439122908251587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004413822777161196,
      "loss": 5.6348,
      "mean_token_accuracy": 0.20892527401447297,
      "num_tokens": 53572588.0,
      "step": 21135
    },
    {
      "entropy": 6.1655701160430905,
      "epoch": 2.4396999422965955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004413544964970695,
      "loss": 5.5236,
      "mean_token_accuracy": 0.21049555093050004,
      "num_tokens": 53586119.0,
      "step": 21140
    },
    {
      "entropy": 6.269611597061157,
      "epoch": 2.4402769763416043,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004413267096827686,
      "loss": 5.618,
      "mean_token_accuracy": 0.2037939265370369,
      "num_tokens": 53600285.0,
      "step": 21145
    },
    {
      "entropy": 6.237957000732422,
      "epoch": 2.4408540103866128,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004412989172741514,
      "loss": 5.6311,
      "mean_token_accuracy": 0.20424163043498994,
      "num_tokens": 53612180.0,
      "step": 21150
    },
    {
      "entropy": 6.2627769947052006,
      "epoch": 2.4414310444316216,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044127111927215267,
      "loss": 5.6572,
      "mean_token_accuracy": 0.2072310119867325,
      "num_tokens": 53624543.0,
      "step": 21155
    },
    {
      "entropy": 6.22073278427124,
      "epoch": 2.44200807847663,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044124331567770746,
      "loss": 5.5805,
      "mean_token_accuracy": 0.2093193456530571,
      "num_tokens": 53636984.0,
      "step": 21160
    },
    {
      "entropy": 6.195828676223755,
      "epoch": 2.442585112521639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004412155064917509,
      "loss": 5.5649,
      "mean_token_accuracy": 0.2027840331196785,
      "num_tokens": 53648844.0,
      "step": 21165
    },
    {
      "entropy": 6.168500375747681,
      "epoch": 2.4431621465666473,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044118769171521836,
      "loss": 5.5723,
      "mean_token_accuracy": 0.20852757394313812,
      "num_tokens": 53662061.0,
      "step": 21170
    },
    {
      "entropy": 6.2353729724884035,
      "epoch": 2.443739180611656,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044115987134904543,
      "loss": 5.5877,
      "mean_token_accuracy": 0.20755985379219055,
      "num_tokens": 53675081.0,
      "step": 21175
    },
    {
      "entropy": 6.2837623119354244,
      "epoch": 2.4443162146566646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044113204539416776,
      "loss": 5.6928,
      "mean_token_accuracy": 0.1996910348534584,
      "num_tokens": 53687825.0,
      "step": 21180
    },
    {
      "entropy": 6.2594469547271725,
      "epoch": 2.4448932487016735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004411042138515212,
      "loss": 5.68,
      "mean_token_accuracy": 0.19770339876413345,
      "num_tokens": 53699564.0,
      "step": 21185
    },
    {
      "entropy": 6.209108591079712,
      "epoch": 2.445470282746682,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004410763767220419,
      "loss": 5.6052,
      "mean_token_accuracy": 0.2066243037581444,
      "num_tokens": 53712657.0,
      "step": 21190
    },
    {
      "entropy": 6.2413112163543705,
      "epoch": 2.4460473167916907,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004410485340066662,
      "loss": 5.6116,
      "mean_token_accuracy": 0.20941103249788284,
      "num_tokens": 53727592.0,
      "step": 21195
    },
    {
      "entropy": 6.3249036312103275,
      "epoch": 2.446624350836699,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004410206857063305,
      "loss": 5.6634,
      "mean_token_accuracy": 0.20049799382686614,
      "num_tokens": 53742133.0,
      "step": 21200
    },
    {
      "entropy": 6.226745080947876,
      "epoch": 2.447201384881708,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004409928318219715,
      "loss": 5.6943,
      "mean_token_accuracy": 0.20384602546691893,
      "num_tokens": 53754984.0,
      "step": 21205
    },
    {
      "entropy": 6.232125520706177,
      "epoch": 2.447778418926717,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004409649723545262,
      "loss": 5.6286,
      "mean_token_accuracy": 0.20479609072208405,
      "num_tokens": 53768466.0,
      "step": 21210
    },
    {
      "entropy": 6.292121648788452,
      "epoch": 2.4483554529717253,
      "grad_norm": 1.0,
      "learning_rate": 0.0004409371073049313,
      "loss": 5.6475,
      "mean_token_accuracy": 0.20150268822908401,
      "num_tokens": 53780837.0,
      "step": 21215
    },
    {
      "entropy": 6.17997179031372,
      "epoch": 2.448932487016734,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004409092366741243,
      "loss": 5.4846,
      "mean_token_accuracy": 0.2143329918384552,
      "num_tokens": 53792839.0,
      "step": 21220
    },
    {
      "entropy": 6.20372633934021,
      "epoch": 2.4495095210617426,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004408813604630425,
      "loss": 5.6837,
      "mean_token_accuracy": 0.20512353926897048,
      "num_tokens": 53806328.0,
      "step": 21225
    },
    {
      "entropy": 6.1446298122406,
      "epoch": 2.4500865551067514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004408534786726236,
      "loss": 5.6045,
      "mean_token_accuracy": 0.20715100616216658,
      "num_tokens": 53818631.0,
      "step": 21230
    },
    {
      "entropy": 6.212185001373291,
      "epoch": 2.45066358915176,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004408255913038053,
      "loss": 5.5586,
      "mean_token_accuracy": 0.21319554895162582,
      "num_tokens": 53831308.0,
      "step": 21235
    },
    {
      "entropy": 6.203452014923096,
      "epoch": 2.4512406231967687,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004407976983575256,
      "loss": 5.6992,
      "mean_token_accuracy": 0.20277448296546935,
      "num_tokens": 53843085.0,
      "step": 21240
    },
    {
      "entropy": 6.292973232269287,
      "epoch": 2.451817657241777,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004407697998347227,
      "loss": 5.7212,
      "mean_token_accuracy": 0.19782721549272536,
      "num_tokens": 53855828.0,
      "step": 21245
    },
    {
      "entropy": 6.19419641494751,
      "epoch": 2.452394691286786,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004407418957363349,
      "loss": 5.5933,
      "mean_token_accuracy": 0.2066192924976349,
      "num_tokens": 53870000.0,
      "step": 21250
    },
    {
      "entropy": 6.20596981048584,
      "epoch": 2.4529717253317944,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044071398606330075,
      "loss": 5.631,
      "mean_token_accuracy": 0.20458243489265443,
      "num_tokens": 53881895.0,
      "step": 21255
    },
    {
      "entropy": 6.30525541305542,
      "epoch": 2.4535487593768033,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000440686070816559,
      "loss": 5.7187,
      "mean_token_accuracy": 0.19853709042072296,
      "num_tokens": 53894971.0,
      "step": 21260
    },
    {
      "entropy": 6.2315673828125,
      "epoch": 2.4541257934218117,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004406581499970486,
      "loss": 5.6489,
      "mean_token_accuracy": 0.1972912833094597,
      "num_tokens": 53906879.0,
      "step": 21265
    },
    {
      "entropy": 6.236013269424438,
      "epoch": 2.4547028274668206,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004406302236057086,
      "loss": 5.6532,
      "mean_token_accuracy": 0.2014811873435974,
      "num_tokens": 53920574.0,
      "step": 21270
    },
    {
      "entropy": 6.315651750564575,
      "epoch": 2.455279861511829,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044060229164347826,
      "loss": 5.6985,
      "mean_token_accuracy": 0.2006470650434494,
      "num_tokens": 53932299.0,
      "step": 21275
    },
    {
      "entropy": 6.348610305786133,
      "epoch": 2.455856895556838,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044057435411129714,
      "loss": 5.6783,
      "mean_token_accuracy": 0.20104454904794694,
      "num_tokens": 53944319.0,
      "step": 21280
    },
    {
      "entropy": 6.245542526245117,
      "epoch": 2.4564339296018467,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044054641101010485,
      "loss": 5.7048,
      "mean_token_accuracy": 0.19634830802679062,
      "num_tokens": 53957246.0,
      "step": 21285
    },
    {
      "entropy": 6.235433673858642,
      "epoch": 2.457010963646855,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044051846234084127,
      "loss": 5.6508,
      "mean_token_accuracy": 0.19950271248817444,
      "num_tokens": 53969525.0,
      "step": 21290
    },
    {
      "entropy": 6.2495574951171875,
      "epoch": 2.457587997691864,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004404905081044464,
      "loss": 5.6754,
      "mean_token_accuracy": 0.20457307994365692,
      "num_tokens": 53982137.0,
      "step": 21295
    },
    {
      "entropy": 6.1984326362609865,
      "epoch": 2.4581650317368724,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004404625483018605,
      "loss": 5.647,
      "mean_token_accuracy": 0.2022179201245308,
      "num_tokens": 53995381.0,
      "step": 21300
    },
    {
      "entropy": 6.31358380317688,
      "epoch": 2.4587420657818813,
      "grad_norm": 0.921875,
      "learning_rate": 0.000440434582934024,
      "loss": 5.7154,
      "mean_token_accuracy": 0.19425119757652282,
      "num_tokens": 54008233.0,
      "step": 21305
    },
    {
      "entropy": 6.334550237655639,
      "epoch": 2.4593190998268897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044040661200187746,
      "loss": 5.6832,
      "mean_token_accuracy": 0.20341849327087402,
      "num_tokens": 54021129.0,
      "step": 21310
    },
    {
      "entropy": 6.290974426269531,
      "epoch": 2.4598961338718985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044037863550636173,
      "loss": 5.6797,
      "mean_token_accuracy": 0.19913502931594848,
      "num_tokens": 54034147.0,
      "step": 21315
    },
    {
      "entropy": 6.247121667861938,
      "epoch": 2.460473167916907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044035065344841766,
      "loss": 5.6208,
      "mean_token_accuracy": 0.20915993601083754,
      "num_tokens": 54046823.0,
      "step": 21320
    },
    {
      "entropy": 6.286140871047974,
      "epoch": 2.461050201961916,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044032266582898655,
      "loss": 5.6896,
      "mean_token_accuracy": 0.19982291907072067,
      "num_tokens": 54059024.0,
      "step": 21325
    },
    {
      "entropy": 6.234236478805542,
      "epoch": 2.4616272360069242,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004402946726490097,
      "loss": 5.6006,
      "mean_token_accuracy": 0.198054338991642,
      "num_tokens": 54070916.0,
      "step": 21330
    },
    {
      "entropy": 6.3291853904724125,
      "epoch": 2.462204270051933,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044026667390942867,
      "loss": 5.7131,
      "mean_token_accuracy": 0.19325656741857528,
      "num_tokens": 54083216.0,
      "step": 21335
    },
    {
      "entropy": 6.2415930271148685,
      "epoch": 2.462781304096942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004402386696111851,
      "loss": 5.6738,
      "mean_token_accuracy": 0.19743142873048783,
      "num_tokens": 54095803.0,
      "step": 21340
    },
    {
      "entropy": 6.268628978729248,
      "epoch": 2.4633583381419504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044021065975522096,
      "loss": 5.6569,
      "mean_token_accuracy": 0.20121200680732726,
      "num_tokens": 54107173.0,
      "step": 21345
    },
    {
      "entropy": 6.251019859313965,
      "epoch": 2.463935372186959,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004401826443424784,
      "loss": 5.6334,
      "mean_token_accuracy": 0.2080500304698944,
      "num_tokens": 54119773.0,
      "step": 21350
    },
    {
      "entropy": 6.258079099655151,
      "epoch": 2.4645124062319677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044015462337389954,
      "loss": 5.6692,
      "mean_token_accuracy": 0.2061997875571251,
      "num_tokens": 54132852.0,
      "step": 21355
    },
    {
      "entropy": 6.299524545669556,
      "epoch": 2.4650894402769765,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044012659685042704,
      "loss": 5.6701,
      "mean_token_accuracy": 0.20350652188062668,
      "num_tokens": 54145537.0,
      "step": 21360
    },
    {
      "entropy": 6.155417537689209,
      "epoch": 2.465666474321985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004400985647730034,
      "loss": 5.561,
      "mean_token_accuracy": 0.20963367819786072,
      "num_tokens": 54157697.0,
      "step": 21365
    },
    {
      "entropy": 6.179322528839111,
      "epoch": 2.466243508366994,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044007052714257165,
      "loss": 5.5836,
      "mean_token_accuracy": 0.20789146274328232,
      "num_tokens": 54170223.0,
      "step": 21370
    },
    {
      "entropy": 6.363755941390991,
      "epoch": 2.466820542412002,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004400424839600747,
      "loss": 5.7763,
      "mean_token_accuracy": 0.1946374997496605,
      "num_tokens": 54182765.0,
      "step": 21375
    },
    {
      "entropy": 6.205576324462891,
      "epoch": 2.467397576457011,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044001443522645575,
      "loss": 5.6252,
      "mean_token_accuracy": 0.20408218502998351,
      "num_tokens": 54195154.0,
      "step": 21380
    },
    {
      "entropy": 6.204359436035157,
      "epoch": 2.4679746105020195,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004399863809426582,
      "loss": 5.722,
      "mean_token_accuracy": 0.19979955852031708,
      "num_tokens": 54207466.0,
      "step": 21385
    },
    {
      "entropy": 6.294766521453857,
      "epoch": 2.4685516445470284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043995832110962576,
      "loss": 5.6341,
      "mean_token_accuracy": 0.20264960080385208,
      "num_tokens": 54218647.0,
      "step": 21390
    },
    {
      "entropy": 6.254158687591553,
      "epoch": 2.4691286785920368,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043993025572830203,
      "loss": 5.5808,
      "mean_token_accuracy": 0.20949572175741196,
      "num_tokens": 54231486.0,
      "step": 21395
    },
    {
      "entropy": 6.204292964935303,
      "epoch": 2.4697057126370456,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043990218479963115,
      "loss": 5.6598,
      "mean_token_accuracy": 0.1996624767780304,
      "num_tokens": 54245186.0,
      "step": 21400
    },
    {
      "entropy": 6.2732048511505125,
      "epoch": 2.470282746682054,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043987410832455715,
      "loss": 5.7393,
      "mean_token_accuracy": 0.1911991685628891,
      "num_tokens": 54257675.0,
      "step": 21405
    },
    {
      "entropy": 6.260737609863281,
      "epoch": 2.470859780727063,
      "grad_norm": 0.890625,
      "learning_rate": 0.00043984602630402447,
      "loss": 5.6251,
      "mean_token_accuracy": 0.20740590244531631,
      "num_tokens": 54270605.0,
      "step": 21410
    },
    {
      "entropy": 6.234555292129516,
      "epoch": 2.4714368147720718,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043981793873897756,
      "loss": 5.6503,
      "mean_token_accuracy": 0.20723131746053697,
      "num_tokens": 54283003.0,
      "step": 21415
    },
    {
      "entropy": 6.307766103744507,
      "epoch": 2.47201384881708,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004397898456303612,
      "loss": 5.7333,
      "mean_token_accuracy": 0.19427064508199693,
      "num_tokens": 54294760.0,
      "step": 21420
    },
    {
      "entropy": 6.144174575805664,
      "epoch": 2.472590882862089,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043976174697912015,
      "loss": 5.6369,
      "mean_token_accuracy": 0.21197833269834518,
      "num_tokens": 54309116.0,
      "step": 21425
    },
    {
      "entropy": 6.1588600158691404,
      "epoch": 2.4731679169070975,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004397336427861996,
      "loss": 5.6279,
      "mean_token_accuracy": 0.20820801258087157,
      "num_tokens": 54322758.0,
      "step": 21430
    },
    {
      "entropy": 6.348609685897827,
      "epoch": 2.4737449509521063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397055330525448,
      "loss": 5.7323,
      "mean_token_accuracy": 0.20001533031463622,
      "num_tokens": 54335009.0,
      "step": 21435
    },
    {
      "entropy": 6.292016410827637,
      "epoch": 2.4743219849971148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004396774177791012,
      "loss": 5.6837,
      "mean_token_accuracy": 0.2042809709906578,
      "num_tokens": 54348035.0,
      "step": 21440
    },
    {
      "entropy": 6.21948823928833,
      "epoch": 2.4748990190421236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043964929696681444,
      "loss": 5.6809,
      "mean_token_accuracy": 0.20219341814517974,
      "num_tokens": 54361261.0,
      "step": 21445
    },
    {
      "entropy": 6.256408262252807,
      "epoch": 2.475476053087132,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004396211706166305,
      "loss": 5.6707,
      "mean_token_accuracy": 0.20397267639636993,
      "num_tokens": 54373454.0,
      "step": 21450
    },
    {
      "entropy": 6.334561967849732,
      "epoch": 2.476053087132141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004395930387294951,
      "loss": 5.7264,
      "mean_token_accuracy": 0.1966416284441948,
      "num_tokens": 54385793.0,
      "step": 21455
    },
    {
      "entropy": 6.232241535186768,
      "epoch": 2.4766301211771493,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004395649013063546,
      "loss": 5.5903,
      "mean_token_accuracy": 0.2120537579059601,
      "num_tokens": 54398723.0,
      "step": 21460
    },
    {
      "entropy": 6.185580253601074,
      "epoch": 2.477207155222158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004395367583481554,
      "loss": 5.5576,
      "mean_token_accuracy": 0.21847614794969558,
      "num_tokens": 54411009.0,
      "step": 21465
    },
    {
      "entropy": 6.231150817871094,
      "epoch": 2.4777841892671666,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043950860985584423,
      "loss": 5.6702,
      "mean_token_accuracy": 0.19831295162439347,
      "num_tokens": 54423161.0,
      "step": 21470
    },
    {
      "entropy": 6.290499019622803,
      "epoch": 2.4783612233121755,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004394804558303675,
      "loss": 5.6808,
      "mean_token_accuracy": 0.198335699737072,
      "num_tokens": 54436594.0,
      "step": 21475
    },
    {
      "entropy": 6.319542598724365,
      "epoch": 2.478938257357184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004394522962726724,
      "loss": 5.7122,
      "mean_token_accuracy": 0.20206487476825713,
      "num_tokens": 54450038.0,
      "step": 21480
    },
    {
      "entropy": 6.186439085006714,
      "epoch": 2.4795152914021927,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043942413118370604,
      "loss": 5.5336,
      "mean_token_accuracy": 0.21414311081171036,
      "num_tokens": 54461475.0,
      "step": 21485
    },
    {
      "entropy": 6.168961763381958,
      "epoch": 2.4800923254472016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004393959605644157,
      "loss": 5.6335,
      "mean_token_accuracy": 0.20764028429985046,
      "num_tokens": 54473460.0,
      "step": 21490
    },
    {
      "entropy": 6.287596273422241,
      "epoch": 2.48066935949221,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004393677844157488,
      "loss": 5.6542,
      "mean_token_accuracy": 0.20022647678852082,
      "num_tokens": 54486629.0,
      "step": 21495
    },
    {
      "entropy": 6.216681480407715,
      "epoch": 2.481246393537219,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004393396027386532,
      "loss": 5.6133,
      "mean_token_accuracy": 0.2118172287940979,
      "num_tokens": 54500091.0,
      "step": 21500
    },
    {
      "entropy": 6.243809223175049,
      "epoch": 2.4818234275822273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043931141553407667,
      "loss": 5.6823,
      "mean_token_accuracy": 0.19680240601301194,
      "num_tokens": 54512303.0,
      "step": 21505
    },
    {
      "entropy": 6.249721622467041,
      "epoch": 2.482400461627236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004392832228029672,
      "loss": 5.6047,
      "mean_token_accuracy": 0.2107479214668274,
      "num_tokens": 54524098.0,
      "step": 21510
    },
    {
      "entropy": 6.2743078708648685,
      "epoch": 2.4829774956722446,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043925502454627327,
      "loss": 5.654,
      "mean_token_accuracy": 0.20403279066085817,
      "num_tokens": 54536473.0,
      "step": 21515
    },
    {
      "entropy": 6.271044445037842,
      "epoch": 2.4835545297172534,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004392268207649431,
      "loss": 5.7024,
      "mean_token_accuracy": 0.1968773201107979,
      "num_tokens": 54549314.0,
      "step": 21520
    },
    {
      "entropy": 6.260916328430175,
      "epoch": 2.484131563762262,
      "grad_norm": 1.0,
      "learning_rate": 0.00043919861145992546,
      "loss": 5.6816,
      "mean_token_accuracy": 0.20433084666728973,
      "num_tokens": 54562647.0,
      "step": 21525
    },
    {
      "entropy": 6.188565492630005,
      "epoch": 2.4847085978072707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000439170396632169,
      "loss": 5.5061,
      "mean_token_accuracy": 0.21750845462083818,
      "num_tokens": 54574821.0,
      "step": 21530
    },
    {
      "entropy": 6.222858715057373,
      "epoch": 2.485285631852279,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004391421762826229,
      "loss": 5.6212,
      "mean_token_accuracy": 0.20149155408143998,
      "num_tokens": 54587563.0,
      "step": 21535
    },
    {
      "entropy": 6.245347499847412,
      "epoch": 2.485862665897288,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004391139504122362,
      "loss": 5.6441,
      "mean_token_accuracy": 0.20286660641431808,
      "num_tokens": 54599645.0,
      "step": 21540
    },
    {
      "entropy": 6.245686388015747,
      "epoch": 2.4864396999422964,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043908571902195827,
      "loss": 5.5833,
      "mean_token_accuracy": 0.21152589917182923,
      "num_tokens": 54611656.0,
      "step": 21545
    },
    {
      "entropy": 6.318813610076904,
      "epoch": 2.4870167339873053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043905748211273864,
      "loss": 5.7314,
      "mean_token_accuracy": 0.19691648334264755,
      "num_tokens": 54624428.0,
      "step": 21550
    },
    {
      "entropy": 6.321946382522583,
      "epoch": 2.4875937680323137,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043902923968552714,
      "loss": 5.7941,
      "mean_token_accuracy": 0.19471813291311263,
      "num_tokens": 54637108.0,
      "step": 21555
    },
    {
      "entropy": 6.195163631439209,
      "epoch": 2.4881708020773226,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004390009917412737,
      "loss": 5.5903,
      "mean_token_accuracy": 0.20831742733716965,
      "num_tokens": 54649887.0,
      "step": 21560
    },
    {
      "entropy": 6.264089870452881,
      "epoch": 2.4887478361223314,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043897273828092824,
      "loss": 5.6759,
      "mean_token_accuracy": 0.20019799172878266,
      "num_tokens": 54661418.0,
      "step": 21565
    },
    {
      "entropy": 6.272325944900513,
      "epoch": 2.48932487016734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004389444793054413,
      "loss": 5.6735,
      "mean_token_accuracy": 0.20333685576915742,
      "num_tokens": 54673932.0,
      "step": 21570
    },
    {
      "entropy": 6.227572441101074,
      "epoch": 2.4899019042123487,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043891621481576324,
      "loss": 5.638,
      "mean_token_accuracy": 0.20631409287452698,
      "num_tokens": 54687062.0,
      "step": 21575
    },
    {
      "entropy": 6.1818595886230465,
      "epoch": 2.490478938257357,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004388879448128446,
      "loss": 5.6134,
      "mean_token_accuracy": 0.2012318879365921,
      "num_tokens": 54700263.0,
      "step": 21580
    },
    {
      "entropy": 6.274971342086792,
      "epoch": 2.491055972302366,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043885966929763635,
      "loss": 5.6464,
      "mean_token_accuracy": 0.19777673333883286,
      "num_tokens": 54712653.0,
      "step": 21585
    },
    {
      "entropy": 6.3075000762939455,
      "epoch": 2.4916330063473744,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043883138827108964,
      "loss": 5.6842,
      "mean_token_accuracy": 0.19628312140703202,
      "num_tokens": 54723832.0,
      "step": 21590
    },
    {
      "entropy": 6.18686466217041,
      "epoch": 2.4922100403923833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004388031017341555,
      "loss": 5.6183,
      "mean_token_accuracy": 0.20528174489736556,
      "num_tokens": 54737233.0,
      "step": 21595
    },
    {
      "entropy": 6.142673301696777,
      "epoch": 2.4927870744373917,
      "grad_norm": 1.125,
      "learning_rate": 0.00043877480968778536,
      "loss": 5.5163,
      "mean_token_accuracy": 0.22360447645187378,
      "num_tokens": 54750535.0,
      "step": 21600
    },
    {
      "entropy": 6.309977388381958,
      "epoch": 2.4933641084824005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004387465121329309,
      "loss": 5.6785,
      "mean_token_accuracy": 0.19566155970096588,
      "num_tokens": 54761989.0,
      "step": 21605
    },
    {
      "entropy": 6.188508796691894,
      "epoch": 2.493941142527409,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043871820907054375,
      "loss": 5.6035,
      "mean_token_accuracy": 0.20941323786973953,
      "num_tokens": 54774562.0,
      "step": 21610
    },
    {
      "entropy": 6.335770225524902,
      "epoch": 2.494518176572418,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043868990050157617,
      "loss": 5.7675,
      "mean_token_accuracy": 0.1910898357629776,
      "num_tokens": 54787177.0,
      "step": 21615
    },
    {
      "entropy": 6.089748954772949,
      "epoch": 2.4950952106174267,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004386615864269799,
      "loss": 5.4441,
      "mean_token_accuracy": 0.23035222440958023,
      "num_tokens": 54802449.0,
      "step": 21620
    },
    {
      "entropy": 6.283981132507324,
      "epoch": 2.495672244662435,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043863326684770746,
      "loss": 5.6466,
      "mean_token_accuracy": 0.20307061970233917,
      "num_tokens": 54815169.0,
      "step": 21625
    },
    {
      "entropy": 6.236248445510864,
      "epoch": 2.4962492787074435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004386049417647115,
      "loss": 5.6287,
      "mean_token_accuracy": 0.2055135875940323,
      "num_tokens": 54826559.0,
      "step": 21630
    },
    {
      "entropy": 6.170205068588257,
      "epoch": 2.4968263127524524,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004385766111789445,
      "loss": 5.6237,
      "mean_token_accuracy": 0.20537068098783492,
      "num_tokens": 54839388.0,
      "step": 21635
    },
    {
      "entropy": 6.26614899635315,
      "epoch": 2.4974033467974612,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004385482750913595,
      "loss": 5.6966,
      "mean_token_accuracy": 0.1966714695096016,
      "num_tokens": 54852717.0,
      "step": 21640
    },
    {
      "entropy": 6.2910566329956055,
      "epoch": 2.4979803808424696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004385199335029094,
      "loss": 5.7267,
      "mean_token_accuracy": 0.20120503306388854,
      "num_tokens": 54865683.0,
      "step": 21645
    },
    {
      "entropy": 6.257294511795044,
      "epoch": 2.4985574148874785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043849158641454767,
      "loss": 5.6345,
      "mean_token_accuracy": 0.20715747624635697,
      "num_tokens": 54878494.0,
      "step": 21650
    },
    {
      "entropy": 6.275032472610474,
      "epoch": 2.499134448932487,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043846323382722765,
      "loss": 5.7168,
      "mean_token_accuracy": 0.19551513493061065,
      "num_tokens": 54891001.0,
      "step": 21655
    },
    {
      "entropy": 6.2365562915802,
      "epoch": 2.499711482977496,
      "grad_norm": 1.0,
      "learning_rate": 0.00043843487574190285,
      "loss": 5.5823,
      "mean_token_accuracy": 0.20593899339437485,
      "num_tokens": 54902513.0,
      "step": 21660
    },
    {
      "entropy": 6.205258560180664,
      "epoch": 2.500288517022504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043840651215952726,
      "loss": 5.5876,
      "mean_token_accuracy": 0.21272094249725343,
      "num_tokens": 54914072.0,
      "step": 21665
    },
    {
      "entropy": 6.2462512969970705,
      "epoch": 2.500865551067513,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043837814308105477,
      "loss": 5.6464,
      "mean_token_accuracy": 0.20234570801258087,
      "num_tokens": 54926674.0,
      "step": 21670
    },
    {
      "entropy": 6.252710056304932,
      "epoch": 2.5014425851125215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043834976850743964,
      "loss": 5.6374,
      "mean_token_accuracy": 0.20573928207159042,
      "num_tokens": 54939271.0,
      "step": 21675
    },
    {
      "entropy": 6.247317266464234,
      "epoch": 2.5020196191575304,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004383213884396361,
      "loss": 5.6639,
      "mean_token_accuracy": 0.19915210604667663,
      "num_tokens": 54950359.0,
      "step": 21680
    },
    {
      "entropy": 6.179173564910888,
      "epoch": 2.5025966532025388,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004382930028785989,
      "loss": 5.6019,
      "mean_token_accuracy": 0.20802728235721588,
      "num_tokens": 54963094.0,
      "step": 21685
    },
    {
      "entropy": 6.219745206832886,
      "epoch": 2.5031736872475476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004382646118252826,
      "loss": 5.5629,
      "mean_token_accuracy": 0.20071256011724473,
      "num_tokens": 54975063.0,
      "step": 21690
    },
    {
      "entropy": 6.175311136245727,
      "epoch": 2.5037507212925565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043823621528064216,
      "loss": 5.5704,
      "mean_token_accuracy": 0.2057197645306587,
      "num_tokens": 54986738.0,
      "step": 21695
    },
    {
      "entropy": 6.2306952476501465,
      "epoch": 2.504327755337565,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043820781324563276,
      "loss": 5.6714,
      "mean_token_accuracy": 0.19630372673273086,
      "num_tokens": 54999181.0,
      "step": 21700
    },
    {
      "entropy": 6.34551043510437,
      "epoch": 2.5049047893825733,
      "grad_norm": 0.90625,
      "learning_rate": 0.00043817940572120963,
      "loss": 5.7259,
      "mean_token_accuracy": 0.1981436252593994,
      "num_tokens": 55012003.0,
      "step": 21705
    },
    {
      "entropy": 6.22825288772583,
      "epoch": 2.505481823427582,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043815099270832814,
      "loss": 5.6367,
      "mean_token_accuracy": 0.20454103648662567,
      "num_tokens": 55025224.0,
      "step": 21710
    },
    {
      "entropy": 6.281681966781616,
      "epoch": 2.506058857472591,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043812257420794415,
      "loss": 5.6732,
      "mean_token_accuracy": 0.19923289716243744,
      "num_tokens": 55038196.0,
      "step": 21715
    },
    {
      "entropy": 6.264958047866822,
      "epoch": 2.5066358915175995,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043809415022101335,
      "loss": 5.6379,
      "mean_token_accuracy": 0.20355327129364015,
      "num_tokens": 55049866.0,
      "step": 21720
    },
    {
      "entropy": 6.214623975753784,
      "epoch": 2.5072129255626083,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004380657207484919,
      "loss": 5.6331,
      "mean_token_accuracy": 0.21095803678035735,
      "num_tokens": 55062821.0,
      "step": 21725
    },
    {
      "entropy": 6.221945905685425,
      "epoch": 2.5077899596076167,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043803728579133586,
      "loss": 5.6391,
      "mean_token_accuracy": 0.2057766616344452,
      "num_tokens": 55076010.0,
      "step": 21730
    },
    {
      "entropy": 6.180741786956787,
      "epoch": 2.5083669936526256,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004380088453505017,
      "loss": 5.5807,
      "mean_token_accuracy": 0.21112370789051055,
      "num_tokens": 55089460.0,
      "step": 21735
    },
    {
      "entropy": 6.2625044822692875,
      "epoch": 2.508944027697634,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043798039942694603,
      "loss": 5.6641,
      "mean_token_accuracy": 0.2064279541373253,
      "num_tokens": 55101320.0,
      "step": 21740
    },
    {
      "entropy": 6.24510760307312,
      "epoch": 2.509521061742643,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00043795194802162557,
      "loss": 5.6302,
      "mean_token_accuracy": 0.2082613229751587,
      "num_tokens": 55114534.0,
      "step": 21745
    },
    {
      "entropy": 6.288366413116455,
      "epoch": 2.5100980957876513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004379234911354973,
      "loss": 5.6448,
      "mean_token_accuracy": 0.20710760056972505,
      "num_tokens": 55126727.0,
      "step": 21750
    },
    {
      "entropy": 6.202338123321534,
      "epoch": 2.51067512983266,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043789502876951834,
      "loss": 5.5918,
      "mean_token_accuracy": 0.21053524762392045,
      "num_tokens": 55139513.0,
      "step": 21755
    },
    {
      "entropy": 6.256877660751343,
      "epoch": 2.5112521638776686,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004378665609246459,
      "loss": 5.6604,
      "mean_token_accuracy": 0.20573174059391022,
      "num_tokens": 55153245.0,
      "step": 21760
    },
    {
      "entropy": 6.306287097930908,
      "epoch": 2.5118291979226774,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043783808760183764,
      "loss": 5.6507,
      "mean_token_accuracy": 0.19992405474185942,
      "num_tokens": 55165850.0,
      "step": 21765
    },
    {
      "entropy": 6.326633882522583,
      "epoch": 2.5124062319676863,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043780960880205125,
      "loss": 5.6904,
      "mean_token_accuracy": 0.20125101059675216,
      "num_tokens": 55177347.0,
      "step": 21770
    },
    {
      "entropy": 6.301204347610474,
      "epoch": 2.5129832660126947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043778112452624445,
      "loss": 5.7156,
      "mean_token_accuracy": 0.20001438856124878,
      "num_tokens": 55190568.0,
      "step": 21775
    },
    {
      "entropy": 6.1845564365386965,
      "epoch": 2.513560300057703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043775263477537546,
      "loss": 5.6315,
      "mean_token_accuracy": 0.2058517098426819,
      "num_tokens": 55203088.0,
      "step": 21780
    },
    {
      "entropy": 6.2149893283844,
      "epoch": 2.514137334102712,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004377241395504024,
      "loss": 5.5707,
      "mean_token_accuracy": 0.20881011337041855,
      "num_tokens": 55215388.0,
      "step": 21785
    },
    {
      "entropy": 6.179731225967407,
      "epoch": 2.514714368147721,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004376956388522837,
      "loss": 5.6014,
      "mean_token_accuracy": 0.20760366171598435,
      "num_tokens": 55228627.0,
      "step": 21790
    },
    {
      "entropy": 6.277399730682373,
      "epoch": 2.5152914021927293,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004376671326819781,
      "loss": 5.7117,
      "mean_token_accuracy": 0.20504764467477798,
      "num_tokens": 55240719.0,
      "step": 21795
    },
    {
      "entropy": 6.259164476394654,
      "epoch": 2.515868436237738,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00043763862104044424,
      "loss": 5.6223,
      "mean_token_accuracy": 0.20535571575164796,
      "num_tokens": 55253417.0,
      "step": 21800
    },
    {
      "entropy": 6.24641695022583,
      "epoch": 2.5164454702827466,
      "grad_norm": 1.0,
      "learning_rate": 0.00043761010392864114,
      "loss": 5.6232,
      "mean_token_accuracy": 0.20752396434545517,
      "num_tokens": 55264459.0,
      "step": 21805
    },
    {
      "entropy": 6.295002269744873,
      "epoch": 2.5170225043277554,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043758158134752793,
      "loss": 5.7344,
      "mean_token_accuracy": 0.19896973967552184,
      "num_tokens": 55277825.0,
      "step": 21810
    },
    {
      "entropy": 6.232796764373779,
      "epoch": 2.517599538372764,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004375530532980642,
      "loss": 5.7019,
      "mean_token_accuracy": 0.19707966595888138,
      "num_tokens": 55291445.0,
      "step": 21815
    },
    {
      "entropy": 6.259813928604126,
      "epoch": 2.5181765724177727,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000437524519781209,
      "loss": 5.6484,
      "mean_token_accuracy": 0.20565554946660997,
      "num_tokens": 55303896.0,
      "step": 21820
    },
    {
      "entropy": 6.3100916862487795,
      "epoch": 2.5187536064627816,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004374959807979224,
      "loss": 5.7202,
      "mean_token_accuracy": 0.19861781746149063,
      "num_tokens": 55316694.0,
      "step": 21825
    },
    {
      "entropy": 6.241120958328247,
      "epoch": 2.51933064050779,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004374674363491642,
      "loss": 5.648,
      "mean_token_accuracy": 0.2060450494289398,
      "num_tokens": 55329910.0,
      "step": 21830
    },
    {
      "entropy": 6.240363740921021,
      "epoch": 2.5199076745527984,
      "grad_norm": 0.90625,
      "learning_rate": 0.00043743888643589453,
      "loss": 5.6786,
      "mean_token_accuracy": 0.19961225986480713,
      "num_tokens": 55343502.0,
      "step": 21835
    },
    {
      "entropy": 6.298206520080567,
      "epoch": 2.5204847085978073,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043741033105907355,
      "loss": 5.66,
      "mean_token_accuracy": 0.19992417842149734,
      "num_tokens": 55355213.0,
      "step": 21840
    },
    {
      "entropy": 6.216961431503296,
      "epoch": 2.521061742642816,
      "grad_norm": 1.0,
      "learning_rate": 0.0004373817702196618,
      "loss": 5.6254,
      "mean_token_accuracy": 0.20176700502634048,
      "num_tokens": 55368127.0,
      "step": 21845
    },
    {
      "entropy": 6.29473557472229,
      "epoch": 2.5216387766878245,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043735320391861986,
      "loss": 5.6759,
      "mean_token_accuracy": 0.2040083661675453,
      "num_tokens": 55380492.0,
      "step": 21850
    },
    {
      "entropy": 6.323779582977295,
      "epoch": 2.522215810732833,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004373246321569085,
      "loss": 5.6961,
      "mean_token_accuracy": 0.20227411836385728,
      "num_tokens": 55393731.0,
      "step": 21855
    },
    {
      "entropy": 6.202720022201538,
      "epoch": 2.522792844777842,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004372960549354888,
      "loss": 5.6012,
      "mean_token_accuracy": 0.2033860132098198,
      "num_tokens": 55405687.0,
      "step": 21860
    },
    {
      "entropy": 6.289368391036987,
      "epoch": 2.5233698788228507,
      "grad_norm": 1.0,
      "learning_rate": 0.00043726747225532195,
      "loss": 5.6629,
      "mean_token_accuracy": 0.19990501701831817,
      "num_tokens": 55418951.0,
      "step": 21865
    },
    {
      "entropy": 6.237732410430908,
      "epoch": 2.523946912867859,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004372388841173692,
      "loss": 5.6825,
      "mean_token_accuracy": 0.19736928641796112,
      "num_tokens": 55431791.0,
      "step": 21870
    },
    {
      "entropy": 6.305558204650879,
      "epoch": 2.524523946912868,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004372102905225922,
      "loss": 5.6091,
      "mean_token_accuracy": 0.20365576297044755,
      "num_tokens": 55442611.0,
      "step": 21875
    },
    {
      "entropy": 6.239367866516114,
      "epoch": 2.5251009809578764,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043718169147195275,
      "loss": 5.6591,
      "mean_token_accuracy": 0.20660190433263778,
      "num_tokens": 55455303.0,
      "step": 21880
    },
    {
      "entropy": 6.280702018737793,
      "epoch": 2.5256780150028852,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043715308696641255,
      "loss": 5.6774,
      "mean_token_accuracy": 0.2025445058941841,
      "num_tokens": 55467183.0,
      "step": 21885
    },
    {
      "entropy": 6.27821569442749,
      "epoch": 2.5262550490478937,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004371244770069338,
      "loss": 5.6915,
      "mean_token_accuracy": 0.20096874833106995,
      "num_tokens": 55479917.0,
      "step": 21890
    },
    {
      "entropy": 6.222524833679199,
      "epoch": 2.5268320830929025,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043709586159447884,
      "loss": 5.598,
      "mean_token_accuracy": 0.21309640556573867,
      "num_tokens": 55493325.0,
      "step": 21895
    },
    {
      "entropy": 6.270376491546631,
      "epoch": 2.5274091171379114,
      "grad_norm": 1.0,
      "learning_rate": 0.00043706724073001017,
      "loss": 5.622,
      "mean_token_accuracy": 0.2006096363067627,
      "num_tokens": 55505727.0,
      "step": 21900
    },
    {
      "entropy": 6.242608594894409,
      "epoch": 2.52798615118292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004370386144144902,
      "loss": 5.6431,
      "mean_token_accuracy": 0.21060980409383773,
      "num_tokens": 55518076.0,
      "step": 21905
    },
    {
      "entropy": 6.257837390899658,
      "epoch": 2.5285631852279282,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004370099826488821,
      "loss": 5.6672,
      "mean_token_accuracy": 0.19669223129749297,
      "num_tokens": 55531141.0,
      "step": 21910
    },
    {
      "entropy": 6.234814643859863,
      "epoch": 2.529140219272937,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004369813454341486,
      "loss": 5.6551,
      "mean_token_accuracy": 0.20589874684810638,
      "num_tokens": 55543365.0,
      "step": 21915
    },
    {
      "entropy": 6.278917121887207,
      "epoch": 2.529717253317946,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043695270277125303,
      "loss": 5.5616,
      "mean_token_accuracy": 0.20613712817430496,
      "num_tokens": 55555378.0,
      "step": 21920
    },
    {
      "entropy": 6.214874172210694,
      "epoch": 2.5302942873629544,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004369240546611587,
      "loss": 5.584,
      "mean_token_accuracy": 0.21410153508186341,
      "num_tokens": 55567260.0,
      "step": 21925
    },
    {
      "entropy": 6.1591695785522464,
      "epoch": 2.5308713214079632,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043689540110482933,
      "loss": 5.5879,
      "mean_token_accuracy": 0.20682580918073654,
      "num_tokens": 55580177.0,
      "step": 21930
    },
    {
      "entropy": 6.250010013580322,
      "epoch": 2.5314483554529716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004368667421032285,
      "loss": 5.6412,
      "mean_token_accuracy": 0.2048771411180496,
      "num_tokens": 55591878.0,
      "step": 21935
    },
    {
      "entropy": 6.2520678520202635,
      "epoch": 2.5320253894979805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043683807765732013,
      "loss": 5.657,
      "mean_token_accuracy": 0.20342988520860672,
      "num_tokens": 55603465.0,
      "step": 21940
    },
    {
      "entropy": 6.268637847900391,
      "epoch": 2.532602423542989,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043680940776806847,
      "loss": 5.6572,
      "mean_token_accuracy": 0.20088669955730437,
      "num_tokens": 55615555.0,
      "step": 21945
    },
    {
      "entropy": 6.2877027034759525,
      "epoch": 2.533179457587998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004367807324364378,
      "loss": 5.6875,
      "mean_token_accuracy": 0.20010219812393187,
      "num_tokens": 55628680.0,
      "step": 21950
    },
    {
      "entropy": 6.148497819900513,
      "epoch": 2.533756491633006,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043675205166339247,
      "loss": 5.4444,
      "mean_token_accuracy": 0.22404915541410447,
      "num_tokens": 55640415.0,
      "step": 21955
    },
    {
      "entropy": 6.198409843444824,
      "epoch": 2.534333525678015,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004367233654498973,
      "loss": 5.6127,
      "mean_token_accuracy": 0.20940684974193574,
      "num_tokens": 55653551.0,
      "step": 21960
    },
    {
      "entropy": 6.2874843120574955,
      "epoch": 2.5349105597230235,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004366946737969171,
      "loss": 5.6705,
      "mean_token_accuracy": 0.19906880110502242,
      "num_tokens": 55665699.0,
      "step": 21965
    },
    {
      "entropy": 6.241062879562378,
      "epoch": 2.5354875937680323,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004366659767054168,
      "loss": 5.6742,
      "mean_token_accuracy": 0.20272763967514038,
      "num_tokens": 55677949.0,
      "step": 21970
    },
    {
      "entropy": 6.202496480941773,
      "epoch": 2.536064627813041,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043663727417636166,
      "loss": 5.5943,
      "mean_token_accuracy": 0.20167382210493087,
      "num_tokens": 55690055.0,
      "step": 21975
    },
    {
      "entropy": 6.229793548583984,
      "epoch": 2.5366416618580496,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004366085662107171,
      "loss": 5.5817,
      "mean_token_accuracy": 0.212956203520298,
      "num_tokens": 55703053.0,
      "step": 21980
    },
    {
      "entropy": 6.326654052734375,
      "epoch": 2.537218695903058,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043657985280944875,
      "loss": 5.8064,
      "mean_token_accuracy": 0.19366192519664766,
      "num_tokens": 55716052.0,
      "step": 21985
    },
    {
      "entropy": 6.25507607460022,
      "epoch": 2.537795729948067,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004365511339735223,
      "loss": 5.6405,
      "mean_token_accuracy": 0.20398181229829787,
      "num_tokens": 55728762.0,
      "step": 21990
    },
    {
      "entropy": 6.23028335571289,
      "epoch": 2.5383727639930758,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004365224097039036,
      "loss": 5.6641,
      "mean_token_accuracy": 0.20436502397060394,
      "num_tokens": 55742029.0,
      "step": 21995
    },
    {
      "entropy": 6.245124292373657,
      "epoch": 2.538949798038084,
      "grad_norm": 1.140625,
      "learning_rate": 0.000436493680001559,
      "loss": 5.6392,
      "mean_token_accuracy": 0.20721605718135833,
      "num_tokens": 55754790.0,
      "step": 22000
    },
    {
      "entropy": 6.2365966796875,
      "epoch": 2.539526832083093,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043646494486745464,
      "loss": 5.6537,
      "mean_token_accuracy": 0.20898378789424896,
      "num_tokens": 55767651.0,
      "step": 22005
    },
    {
      "entropy": 6.267479419708252,
      "epoch": 2.5401038661281015,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004364362043025571,
      "loss": 5.6484,
      "mean_token_accuracy": 0.20620125383138657,
      "num_tokens": 55779919.0,
      "step": 22010
    },
    {
      "entropy": 6.2517163276672365,
      "epoch": 2.5406809001731103,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043640745830783297,
      "loss": 5.5831,
      "mean_token_accuracy": 0.2093014508485794,
      "num_tokens": 55792171.0,
      "step": 22015
    },
    {
      "entropy": 6.333767318725586,
      "epoch": 2.5412579342181187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004363787068842491,
      "loss": 5.7399,
      "mean_token_accuracy": 0.1960183262825012,
      "num_tokens": 55804292.0,
      "step": 22020
    },
    {
      "entropy": 6.206714248657226,
      "epoch": 2.5418349682631276,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004363499500327727,
      "loss": 5.5699,
      "mean_token_accuracy": 0.21040678918361663,
      "num_tokens": 55816520.0,
      "step": 22025
    },
    {
      "entropy": 6.222457695007324,
      "epoch": 2.542412002308136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043632118775437085,
      "loss": 5.5915,
      "mean_token_accuracy": 0.2091295212507248,
      "num_tokens": 55827774.0,
      "step": 22030
    },
    {
      "entropy": 6.203572511672974,
      "epoch": 2.542989036353145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043629242005001096,
      "loss": 5.651,
      "mean_token_accuracy": 0.20442977994680406,
      "num_tokens": 55841201.0,
      "step": 22035
    },
    {
      "entropy": 6.283616638183593,
      "epoch": 2.5435660703981533,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004362636469206607,
      "loss": 5.6812,
      "mean_token_accuracy": 0.2044379949569702,
      "num_tokens": 55853581.0,
      "step": 22040
    },
    {
      "entropy": 6.245355319976807,
      "epoch": 2.544143104443162,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004362348683672877,
      "loss": 5.6739,
      "mean_token_accuracy": 0.20200098901987076,
      "num_tokens": 55866216.0,
      "step": 22045
    },
    {
      "entropy": 6.210954904556274,
      "epoch": 2.544720138488171,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004362060843908601,
      "loss": 5.6739,
      "mean_token_accuracy": 0.20694909393787383,
      "num_tokens": 55880618.0,
      "step": 22050
    },
    {
      "entropy": 6.2821251392364506,
      "epoch": 2.5452971725331794,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004361772949923458,
      "loss": 5.6901,
      "mean_token_accuracy": 0.2027858778834343,
      "num_tokens": 55894888.0,
      "step": 22055
    },
    {
      "entropy": 6.279632234573365,
      "epoch": 2.545874206578188,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004361485001727132,
      "loss": 5.66,
      "mean_token_accuracy": 0.20160701274871826,
      "num_tokens": 55907273.0,
      "step": 22060
    },
    {
      "entropy": 6.2687067031860355,
      "epoch": 2.5464512406231967,
      "grad_norm": 1.03125,
      "learning_rate": 0.000436119699932931,
      "loss": 5.6523,
      "mean_token_accuracy": 0.20281035006046294,
      "num_tokens": 55919341.0,
      "step": 22065
    },
    {
      "entropy": 6.295963954925537,
      "epoch": 2.5470282746682056,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043609089427396763,
      "loss": 5.6999,
      "mean_token_accuracy": 0.20186266005039216,
      "num_tokens": 55932054.0,
      "step": 22070
    },
    {
      "entropy": 6.268542814254761,
      "epoch": 2.547605308713214,
      "grad_norm": 1.0,
      "learning_rate": 0.00043606208319679215,
      "loss": 5.646,
      "mean_token_accuracy": 0.2055470824241638,
      "num_tokens": 55945134.0,
      "step": 22075
    },
    {
      "entropy": 6.240560007095337,
      "epoch": 2.548182342758223,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004360332667023734,
      "loss": 5.6492,
      "mean_token_accuracy": 0.20508747100830077,
      "num_tokens": 55957386.0,
      "step": 22080
    },
    {
      "entropy": 6.2559479713439945,
      "epoch": 2.5487593768032313,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043600444479168084,
      "loss": 5.6144,
      "mean_token_accuracy": 0.20969599187374116,
      "num_tokens": 55971356.0,
      "step": 22085
    },
    {
      "entropy": 6.244159507751465,
      "epoch": 2.54933641084824,
      "grad_norm": 1.0,
      "learning_rate": 0.0004359756174656836,
      "loss": 5.6286,
      "mean_token_accuracy": 0.20803606510162354,
      "num_tokens": 55982827.0,
      "step": 22090
    },
    {
      "entropy": 6.236189222335815,
      "epoch": 2.5499134448932486,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043594678472535153,
      "loss": 5.6552,
      "mean_token_accuracy": 0.19993047267198563,
      "num_tokens": 55994358.0,
      "step": 22095
    },
    {
      "entropy": 6.3259705066680905,
      "epoch": 2.5504904789382574,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043591794657165425,
      "loss": 5.6789,
      "mean_token_accuracy": 0.19830653220415115,
      "num_tokens": 56006757.0,
      "step": 22100
    },
    {
      "entropy": 6.299615383148193,
      "epoch": 2.5510675129832663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004358891030055617,
      "loss": 5.7058,
      "mean_token_accuracy": 0.19988015443086624,
      "num_tokens": 56018813.0,
      "step": 22105
    },
    {
      "entropy": 6.329316473007202,
      "epoch": 2.5516445470282747,
      "grad_norm": 1.0,
      "learning_rate": 0.0004358602540280443,
      "loss": 5.688,
      "mean_token_accuracy": 0.20038418024778365,
      "num_tokens": 56030970.0,
      "step": 22110
    },
    {
      "entropy": 6.345421409606933,
      "epoch": 2.552221581073283,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043583139964007203,
      "loss": 5.7021,
      "mean_token_accuracy": 0.20111321061849594,
      "num_tokens": 56042672.0,
      "step": 22115
    },
    {
      "entropy": 6.243420886993408,
      "epoch": 2.552798615118292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043580253984261554,
      "loss": 5.6419,
      "mean_token_accuracy": 0.20535562336444854,
      "num_tokens": 56056729.0,
      "step": 22120
    },
    {
      "entropy": 6.165840768814087,
      "epoch": 2.553375649163301,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043577367463664546,
      "loss": 5.5809,
      "mean_token_accuracy": 0.21117308139801025,
      "num_tokens": 56069717.0,
      "step": 22125
    },
    {
      "entropy": 6.157013416290283,
      "epoch": 2.5539526832083093,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004357448040231327,
      "loss": 5.5437,
      "mean_token_accuracy": 0.22166015058755875,
      "num_tokens": 56084500.0,
      "step": 22130
    },
    {
      "entropy": 6.273871040344238,
      "epoch": 2.5545297172533177,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004357159280030483,
      "loss": 5.6511,
      "mean_token_accuracy": 0.1983496904373169,
      "num_tokens": 56096574.0,
      "step": 22135
    },
    {
      "entropy": 6.250105094909668,
      "epoch": 2.5551067512983265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004356870465773635,
      "loss": 5.6654,
      "mean_token_accuracy": 0.19923075288534164,
      "num_tokens": 56109897.0,
      "step": 22140
    },
    {
      "entropy": 6.192885398864746,
      "epoch": 2.5556837853433354,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043565815974704974,
      "loss": 5.6968,
      "mean_token_accuracy": 0.20882801711559296,
      "num_tokens": 56123623.0,
      "step": 22145
    },
    {
      "entropy": 6.306439971923828,
      "epoch": 2.556260819388344,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00043562926751307857,
      "loss": 5.6678,
      "mean_token_accuracy": 0.20548348426818847,
      "num_tokens": 56135870.0,
      "step": 22150
    },
    {
      "entropy": 6.189848756790161,
      "epoch": 2.5568378534333527,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043560036987642177,
      "loss": 5.5715,
      "mean_token_accuracy": 0.20849827229976653,
      "num_tokens": 56149024.0,
      "step": 22155
    },
    {
      "entropy": 6.266059017181396,
      "epoch": 2.557414887478361,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043557146683805134,
      "loss": 5.6871,
      "mean_token_accuracy": 0.19633420258760453,
      "num_tokens": 56161621.0,
      "step": 22160
    },
    {
      "entropy": 6.2153017044067385,
      "epoch": 2.55799192152337,
      "grad_norm": 1.0,
      "learning_rate": 0.0004355425583989393,
      "loss": 5.6281,
      "mean_token_accuracy": 0.20757195204496384,
      "num_tokens": 56173770.0,
      "step": 22165
    },
    {
      "entropy": 6.189179468154907,
      "epoch": 2.5585689555683784,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043551364456005816,
      "loss": 5.5817,
      "mean_token_accuracy": 0.20870570838451385,
      "num_tokens": 56186362.0,
      "step": 22170
    },
    {
      "entropy": 6.178240537643433,
      "epoch": 2.5591459896133872,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043548472532238015,
      "loss": 5.6576,
      "mean_token_accuracy": 0.20631828010082245,
      "num_tokens": 56199288.0,
      "step": 22175
    },
    {
      "entropy": 6.280341482162475,
      "epoch": 2.559723023658396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043545580068687836,
      "loss": 5.604,
      "mean_token_accuracy": 0.20955042243003846,
      "num_tokens": 56212810.0,
      "step": 22180
    },
    {
      "entropy": 6.219345188140869,
      "epoch": 2.5603000577034045,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004354268706545252,
      "loss": 5.6965,
      "mean_token_accuracy": 0.1968012198805809,
      "num_tokens": 56224936.0,
      "step": 22185
    },
    {
      "entropy": 6.236649894714356,
      "epoch": 2.560877091748413,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043539793522629405,
      "loss": 5.6117,
      "mean_token_accuracy": 0.21067962646484376,
      "num_tokens": 56238191.0,
      "step": 22190
    },
    {
      "entropy": 6.205950450897217,
      "epoch": 2.561454125793422,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043536899440315796,
      "loss": 5.5786,
      "mean_token_accuracy": 0.21169401109218597,
      "num_tokens": 56251184.0,
      "step": 22195
    },
    {
      "entropy": 6.223605918884277,
      "epoch": 2.5620311598384307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004353400481860904,
      "loss": 5.5794,
      "mean_token_accuracy": 0.21389643102884293,
      "num_tokens": 56263949.0,
      "step": 22200
    },
    {
      "entropy": 6.190797185897827,
      "epoch": 2.562608193883439,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00043531109657606504,
      "loss": 5.6306,
      "mean_token_accuracy": 0.214384987950325,
      "num_tokens": 56277861.0,
      "step": 22205
    },
    {
      "entropy": 6.247290420532226,
      "epoch": 2.563185227928448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004352821395740555,
      "loss": 5.646,
      "mean_token_accuracy": 0.20477901101112367,
      "num_tokens": 56291494.0,
      "step": 22210
    },
    {
      "entropy": 6.259798192977906,
      "epoch": 2.5637622619734564,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004352531771810359,
      "loss": 5.6337,
      "mean_token_accuracy": 0.20467002987861632,
      "num_tokens": 56302961.0,
      "step": 22215
    },
    {
      "entropy": 6.275150728225708,
      "epoch": 2.564339296018465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004352242093979802,
      "loss": 5.6984,
      "mean_token_accuracy": 0.1947944715619087,
      "num_tokens": 56314714.0,
      "step": 22220
    },
    {
      "entropy": 6.216671514511108,
      "epoch": 2.5649163300634736,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004351952362258628,
      "loss": 5.6491,
      "mean_token_accuracy": 0.21160974949598313,
      "num_tokens": 56328947.0,
      "step": 22225
    },
    {
      "entropy": 6.262277412414551,
      "epoch": 2.5654933641084825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043516625766565825,
      "loss": 5.6496,
      "mean_token_accuracy": 0.20615430325269699,
      "num_tokens": 56341835.0,
      "step": 22230
    },
    {
      "entropy": 6.311953496932984,
      "epoch": 2.566070398153491,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004351372737183412,
      "loss": 5.678,
      "mean_token_accuracy": 0.20848768800497056,
      "num_tokens": 56355102.0,
      "step": 22235
    },
    {
      "entropy": 6.256657695770263,
      "epoch": 2.5666474321985,
      "grad_norm": 1.0,
      "learning_rate": 0.0004351082843848865,
      "loss": 5.6619,
      "mean_token_accuracy": 0.2062780648469925,
      "num_tokens": 56367731.0,
      "step": 22240
    },
    {
      "entropy": 6.213457250595093,
      "epoch": 2.567224466243508,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043507928966626916,
      "loss": 5.5719,
      "mean_token_accuracy": 0.21271344721317292,
      "num_tokens": 56380641.0,
      "step": 22245
    },
    {
      "entropy": 6.247823858261109,
      "epoch": 2.567801500288517,
      "grad_norm": 1.0,
      "learning_rate": 0.0004350502895634644,
      "loss": 5.6808,
      "mean_token_accuracy": 0.2024888053536415,
      "num_tokens": 56392733.0,
      "step": 22250
    },
    {
      "entropy": 6.271483755111694,
      "epoch": 2.568378534333526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004350212840774476,
      "loss": 5.6833,
      "mean_token_accuracy": 0.1996483772993088,
      "num_tokens": 56404838.0,
      "step": 22255
    },
    {
      "entropy": 6.279106712341308,
      "epoch": 2.5689555683785343,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043499227320919456,
      "loss": 5.645,
      "mean_token_accuracy": 0.1998480811715126,
      "num_tokens": 56417657.0,
      "step": 22260
    },
    {
      "entropy": 6.312775468826294,
      "epoch": 2.5695326024235428,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043496325695968076,
      "loss": 5.719,
      "mean_token_accuracy": 0.19552648663520814,
      "num_tokens": 56430651.0,
      "step": 22265
    },
    {
      "entropy": 6.275038194656372,
      "epoch": 2.5701096364685516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004349342353298823,
      "loss": 5.6961,
      "mean_token_accuracy": 0.20148481726646422,
      "num_tokens": 56443379.0,
      "step": 22270
    },
    {
      "entropy": 6.2527069568634035,
      "epoch": 2.5706866705135605,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004349052083207753,
      "loss": 5.6123,
      "mean_token_accuracy": 0.20248958915472032,
      "num_tokens": 56455400.0,
      "step": 22275
    },
    {
      "entropy": 6.292047595977783,
      "epoch": 2.571263704558569,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000434876175933336,
      "loss": 5.6845,
      "mean_token_accuracy": 0.20215694159269332,
      "num_tokens": 56467653.0,
      "step": 22280
    },
    {
      "entropy": 6.261963987350464,
      "epoch": 2.5718407386035778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043484713816854093,
      "loss": 5.6175,
      "mean_token_accuracy": 0.20584796667098998,
      "num_tokens": 56480352.0,
      "step": 22285
    },
    {
      "entropy": 6.29503493309021,
      "epoch": 2.572417772648586,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004348180950273668,
      "loss": 5.5996,
      "mean_token_accuracy": 0.20520972311496735,
      "num_tokens": 56492642.0,
      "step": 22290
    },
    {
      "entropy": 6.302676200866699,
      "epoch": 2.572994806693595,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004347890465107905,
      "loss": 5.718,
      "mean_token_accuracy": 0.19347016811370848,
      "num_tokens": 56504777.0,
      "step": 22295
    },
    {
      "entropy": 6.247596216201782,
      "epoch": 2.5735718407386035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043475999261978894,
      "loss": 5.5644,
      "mean_token_accuracy": 0.21321454346179963,
      "num_tokens": 56516281.0,
      "step": 22300
    },
    {
      "entropy": 6.215832328796386,
      "epoch": 2.5741488747836123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004347309333553393,
      "loss": 5.6361,
      "mean_token_accuracy": 0.21067800223827363,
      "num_tokens": 56527928.0,
      "step": 22305
    },
    {
      "entropy": 6.266404533386231,
      "epoch": 2.5747259088286207,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00043470186871841915,
      "loss": 5.6754,
      "mean_token_accuracy": 0.19688660353422166,
      "num_tokens": 56542033.0,
      "step": 22310
    },
    {
      "entropy": 6.264832782745361,
      "epoch": 2.5753029428736296,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000434672798710006,
      "loss": 5.6504,
      "mean_token_accuracy": 0.19580039530992507,
      "num_tokens": 56555371.0,
      "step": 22315
    },
    {
      "entropy": 6.264954233169556,
      "epoch": 2.575879976918638,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043464372333107756,
      "loss": 5.6534,
      "mean_token_accuracy": 0.20534077882766724,
      "num_tokens": 56568382.0,
      "step": 22320
    },
    {
      "entropy": 6.1607378959655765,
      "epoch": 2.576457010963647,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043461464258261174,
      "loss": 5.5684,
      "mean_token_accuracy": 0.22072719484567643,
      "num_tokens": 56580419.0,
      "step": 22325
    },
    {
      "entropy": 6.283749198913574,
      "epoch": 2.5770340450086557,
      "grad_norm": 1.0,
      "learning_rate": 0.0004345855564655867,
      "loss": 5.7132,
      "mean_token_accuracy": 0.20023158341646194,
      "num_tokens": 56592224.0,
      "step": 22330
    },
    {
      "entropy": 6.289744329452515,
      "epoch": 2.577611079053664,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043455646498098075,
      "loss": 5.6053,
      "mean_token_accuracy": 0.20618803948163986,
      "num_tokens": 56605497.0,
      "step": 22335
    },
    {
      "entropy": 6.299744939804077,
      "epoch": 2.5781881130986726,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043452736812977236,
      "loss": 5.7629,
      "mean_token_accuracy": 0.1959328144788742,
      "num_tokens": 56617969.0,
      "step": 22340
    },
    {
      "entropy": 6.127130842208862,
      "epoch": 2.5787651471436814,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004344982659129402,
      "loss": 5.4899,
      "mean_token_accuracy": 0.21704300194978715,
      "num_tokens": 56631243.0,
      "step": 22345
    },
    {
      "entropy": 6.294405794143676,
      "epoch": 2.5793421811886903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004344691583314631,
      "loss": 5.6711,
      "mean_token_accuracy": 0.2023612767457962,
      "num_tokens": 56643176.0,
      "step": 22350
    },
    {
      "entropy": 6.188527536392212,
      "epoch": 2.5799192152336987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043444004538632005,
      "loss": 5.5853,
      "mean_token_accuracy": 0.20244961529970168,
      "num_tokens": 56655651.0,
      "step": 22355
    },
    {
      "entropy": 6.289286947250366,
      "epoch": 2.5804962492787076,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043441092707849024,
      "loss": 5.7098,
      "mean_token_accuracy": 0.19944323599338531,
      "num_tokens": 56668102.0,
      "step": 22360
    },
    {
      "entropy": 6.335380220413208,
      "epoch": 2.581073283323716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043438180340895315,
      "loss": 5.6245,
      "mean_token_accuracy": 0.20636338144540786,
      "num_tokens": 56679275.0,
      "step": 22365
    },
    {
      "entropy": 6.2415016174316404,
      "epoch": 2.581650317368725,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043435267437868825,
      "loss": 5.6922,
      "mean_token_accuracy": 0.1979640930891037,
      "num_tokens": 56691056.0,
      "step": 22370
    },
    {
      "entropy": 6.308930206298828,
      "epoch": 2.5822273514137333,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004343235399886753,
      "loss": 5.7331,
      "mean_token_accuracy": 0.19567639678716658,
      "num_tokens": 56703639.0,
      "step": 22375
    },
    {
      "entropy": 6.331395673751831,
      "epoch": 2.582804385458742,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004342944002398942,
      "loss": 5.7351,
      "mean_token_accuracy": 0.19284417927265168,
      "num_tokens": 56715976.0,
      "step": 22380
    },
    {
      "entropy": 6.208407640457153,
      "epoch": 2.583381419503751,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004342652551333251,
      "loss": 5.6369,
      "mean_token_accuracy": 0.20475158244371414,
      "num_tokens": 56729629.0,
      "step": 22385
    },
    {
      "entropy": 6.280992078781128,
      "epoch": 2.5839584535487594,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004342361046699482,
      "loss": 5.6693,
      "mean_token_accuracy": 0.197301284968853,
      "num_tokens": 56741903.0,
      "step": 22390
    },
    {
      "entropy": 6.276900053024292,
      "epoch": 2.584535487593768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004342069488507441,
      "loss": 5.6903,
      "mean_token_accuracy": 0.20061750262975692,
      "num_tokens": 56753968.0,
      "step": 22395
    },
    {
      "entropy": 6.204394054412842,
      "epoch": 2.5851125216387767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004341777876766933,
      "loss": 5.5948,
      "mean_token_accuracy": 0.21044307500123977,
      "num_tokens": 56767539.0,
      "step": 22400
    },
    {
      "entropy": 6.260430002212525,
      "epoch": 2.5856895556837856,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004341486211487767,
      "loss": 5.6342,
      "mean_token_accuracy": 0.20906252413988113,
      "num_tokens": 56780449.0,
      "step": 22405
    },
    {
      "entropy": 6.2412656307220455,
      "epoch": 2.586266589728794,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004341194492679753,
      "loss": 5.6394,
      "mean_token_accuracy": 0.20530153810977936,
      "num_tokens": 56793939.0,
      "step": 22410
    },
    {
      "entropy": 6.30711407661438,
      "epoch": 2.5868436237738024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043409027203527016,
      "loss": 5.6748,
      "mean_token_accuracy": 0.2046242892742157,
      "num_tokens": 56806093.0,
      "step": 22415
    },
    {
      "entropy": 6.277175712585449,
      "epoch": 2.5874206578188113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043406108945164283,
      "loss": 5.715,
      "mean_token_accuracy": 0.20068129301071166,
      "num_tokens": 56818727.0,
      "step": 22420
    },
    {
      "entropy": 6.265715646743774,
      "epoch": 2.58799769186382,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043403190151807474,
      "loss": 5.6682,
      "mean_token_accuracy": 0.20796943306922913,
      "num_tokens": 56830777.0,
      "step": 22425
    },
    {
      "entropy": 6.258620548248291,
      "epoch": 2.5885747259088285,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004340027082355477,
      "loss": 5.6379,
      "mean_token_accuracy": 0.203522227704525,
      "num_tokens": 56843183.0,
      "step": 22430
    },
    {
      "entropy": 6.2951267719268795,
      "epoch": 2.5891517599538374,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004339735096050434,
      "loss": 5.6294,
      "mean_token_accuracy": 0.21290096044540405,
      "num_tokens": 56856607.0,
      "step": 22435
    },
    {
      "entropy": 6.271540594100952,
      "epoch": 2.589728793998846,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004339443056275443,
      "loss": 5.6363,
      "mean_token_accuracy": 0.2073008105158806,
      "num_tokens": 56869568.0,
      "step": 22440
    },
    {
      "entropy": 6.254832220077515,
      "epoch": 2.5903058280438547,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043391509630403224,
      "loss": 5.6823,
      "mean_token_accuracy": 0.20042108595371247,
      "num_tokens": 56881885.0,
      "step": 22445
    },
    {
      "entropy": 6.219962501525879,
      "epoch": 2.590882862088863,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043388588163549,
      "loss": 5.5852,
      "mean_token_accuracy": 0.20773003101348878,
      "num_tokens": 56893872.0,
      "step": 22450
    },
    {
      "entropy": 6.2351349830627445,
      "epoch": 2.591459896133872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043385666162289995,
      "loss": 5.6451,
      "mean_token_accuracy": 0.20903323888778685,
      "num_tokens": 56908376.0,
      "step": 22455
    },
    {
      "entropy": 6.15577449798584,
      "epoch": 2.592036930178881,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043382743626724506,
      "loss": 5.5986,
      "mean_token_accuracy": 0.21275263726711274,
      "num_tokens": 56920899.0,
      "step": 22460
    },
    {
      "entropy": 6.233894729614258,
      "epoch": 2.5926139642238892,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004337982055695082,
      "loss": 5.6289,
      "mean_token_accuracy": 0.20799754709005355,
      "num_tokens": 56934456.0,
      "step": 22465
    },
    {
      "entropy": 6.170180797576904,
      "epoch": 2.5931909982688977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004337689695306726,
      "loss": 5.563,
      "mean_token_accuracy": 0.22034967690706253,
      "num_tokens": 56948621.0,
      "step": 22470
    },
    {
      "entropy": 6.270507907867431,
      "epoch": 2.5937680323139065,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004337397281517215,
      "loss": 5.6619,
      "mean_token_accuracy": 0.2025494620203972,
      "num_tokens": 56960272.0,
      "step": 22475
    },
    {
      "entropy": 6.2543412208557125,
      "epoch": 2.5943450663589154,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004337104814336386,
      "loss": 5.6599,
      "mean_token_accuracy": 0.19882901310920714,
      "num_tokens": 56973158.0,
      "step": 22480
    },
    {
      "entropy": 6.285262775421143,
      "epoch": 2.594922100403924,
      "grad_norm": 1.0,
      "learning_rate": 0.0004336812293774075,
      "loss": 5.6919,
      "mean_token_accuracy": 0.19968826174736024,
      "num_tokens": 56985064.0,
      "step": 22485
    },
    {
      "entropy": 6.244166469573974,
      "epoch": 2.595499134448932,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043365197198401195,
      "loss": 5.694,
      "mean_token_accuracy": 0.20402184128761292,
      "num_tokens": 56998952.0,
      "step": 22490
    },
    {
      "entropy": 6.28422737121582,
      "epoch": 2.596076168493941,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004336227092544362,
      "loss": 5.6276,
      "mean_token_accuracy": 0.20938894897699356,
      "num_tokens": 57011102.0,
      "step": 22495
    },
    {
      "entropy": 6.351907110214233,
      "epoch": 2.59665320253895,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004335934411896644,
      "loss": 5.7569,
      "mean_token_accuracy": 0.19748670607805252,
      "num_tokens": 57022125.0,
      "step": 22500
    },
    {
      "entropy": 6.173913383483887,
      "epoch": 2.5972302365839584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000433564167790681,
      "loss": 5.5678,
      "mean_token_accuracy": 0.21557908803224562,
      "num_tokens": 57034687.0,
      "step": 22505
    },
    {
      "entropy": 6.292545747756958,
      "epoch": 2.597807270628967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004335348890584706,
      "loss": 5.7372,
      "mean_token_accuracy": 0.1979536697268486,
      "num_tokens": 57048238.0,
      "step": 22510
    },
    {
      "entropy": 6.32321457862854,
      "epoch": 2.5983843046739756,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043350560499401793,
      "loss": 5.7502,
      "mean_token_accuracy": 0.19906389564275742,
      "num_tokens": 57060599.0,
      "step": 22515
    },
    {
      "entropy": 6.121924257278442,
      "epoch": 2.5989613387189845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004334763155983079,
      "loss": 5.5309,
      "mean_token_accuracy": 0.21142698526382447,
      "num_tokens": 57073042.0,
      "step": 22520
    },
    {
      "entropy": 6.258372735977173,
      "epoch": 2.599538372763993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043344702087232584,
      "loss": 5.7216,
      "mean_token_accuracy": 0.19925712496042253,
      "num_tokens": 57084795.0,
      "step": 22525
    },
    {
      "entropy": 6.250919485092163,
      "epoch": 2.6001154068090018,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004334177208170569,
      "loss": 5.6211,
      "mean_token_accuracy": 0.20900704264640807,
      "num_tokens": 57096175.0,
      "step": 22530
    },
    {
      "entropy": 6.207519817352295,
      "epoch": 2.6006924408540106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043338841543348654,
      "loss": 5.5652,
      "mean_token_accuracy": 0.20881575345993042,
      "num_tokens": 57109881.0,
      "step": 22535
    },
    {
      "entropy": 6.2671510696411135,
      "epoch": 2.601269474899019,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043335910472260057,
      "loss": 5.649,
      "mean_token_accuracy": 0.20215400904417039,
      "num_tokens": 57121392.0,
      "step": 22540
    },
    {
      "entropy": 6.293084573745728,
      "epoch": 2.6018465089440275,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043332978868538483,
      "loss": 5.605,
      "mean_token_accuracy": 0.20692111700773239,
      "num_tokens": 57133699.0,
      "step": 22545
    },
    {
      "entropy": 6.203144598007202,
      "epoch": 2.6024235429890363,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043330046732282515,
      "loss": 5.5788,
      "mean_token_accuracy": 0.2078753098845482,
      "num_tokens": 57146926.0,
      "step": 22550
    },
    {
      "entropy": 6.322693538665772,
      "epoch": 2.603000577034045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043327114063590797,
      "loss": 5.7365,
      "mean_token_accuracy": 0.20374195873737336,
      "num_tokens": 57159752.0,
      "step": 22555
    },
    {
      "entropy": 6.308671665191651,
      "epoch": 2.6035776110790536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004332418086256196,
      "loss": 5.7391,
      "mean_token_accuracy": 0.20282414108514785,
      "num_tokens": 57171337.0,
      "step": 22560
    },
    {
      "entropy": 6.249724102020264,
      "epoch": 2.6041546451240625,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004332124712929466,
      "loss": 5.6422,
      "mean_token_accuracy": 0.2095663145184517,
      "num_tokens": 57183914.0,
      "step": 22565
    },
    {
      "entropy": 6.24452223777771,
      "epoch": 2.604731679169071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043318312863887566,
      "loss": 5.6835,
      "mean_token_accuracy": 0.1987679347395897,
      "num_tokens": 57196149.0,
      "step": 22570
    },
    {
      "entropy": 6.220917797088623,
      "epoch": 2.6053087132140798,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043315378066439384,
      "loss": 5.5003,
      "mean_token_accuracy": 0.21464720070362092,
      "num_tokens": 57208929.0,
      "step": 22575
    },
    {
      "entropy": 6.240729093551636,
      "epoch": 2.605885747259088,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004331244273704881,
      "loss": 5.6083,
      "mean_token_accuracy": 0.21305236965417862,
      "num_tokens": 57222078.0,
      "step": 22580
    },
    {
      "entropy": 6.244909238815308,
      "epoch": 2.606462781304097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004330950687581458,
      "loss": 5.7144,
      "mean_token_accuracy": 0.19999710768461226,
      "num_tokens": 57235073.0,
      "step": 22585
    },
    {
      "entropy": 6.190898513793945,
      "epoch": 2.6070398153491054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043306570482835443,
      "loss": 5.539,
      "mean_token_accuracy": 0.2074820026755333,
      "num_tokens": 57246829.0,
      "step": 22590
    },
    {
      "entropy": 6.26214747428894,
      "epoch": 2.6076168493941143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004330363355821016,
      "loss": 5.6528,
      "mean_token_accuracy": 0.20356080383062364,
      "num_tokens": 57259645.0,
      "step": 22595
    },
    {
      "entropy": 6.205470323562622,
      "epoch": 2.6081938834391227,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004330069610203751,
      "loss": 5.5714,
      "mean_token_accuracy": 0.21345322877168654,
      "num_tokens": 57273053.0,
      "step": 22600
    },
    {
      "entropy": 6.2901115894317625,
      "epoch": 2.6087709174841316,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004329775811441629,
      "loss": 5.7251,
      "mean_token_accuracy": 0.200989331305027,
      "num_tokens": 57286987.0,
      "step": 22605
    },
    {
      "entropy": 6.3358758926391605,
      "epoch": 2.6093479515291405,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043294819595445324,
      "loss": 5.7163,
      "mean_token_accuracy": 0.1954937309026718,
      "num_tokens": 57299358.0,
      "step": 22610
    },
    {
      "entropy": 6.326014709472656,
      "epoch": 2.609924985574149,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004329188054522345,
      "loss": 5.7232,
      "mean_token_accuracy": 0.20004579722881316,
      "num_tokens": 57311623.0,
      "step": 22615
    },
    {
      "entropy": 6.256593227386475,
      "epoch": 2.6105020196191573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004328894096384952,
      "loss": 5.6774,
      "mean_token_accuracy": 0.20402936935424804,
      "num_tokens": 57323947.0,
      "step": 22620
    },
    {
      "entropy": 6.258550977706909,
      "epoch": 2.611079053664166,
      "grad_norm": 0.9375,
      "learning_rate": 0.000432860008514224,
      "loss": 5.7027,
      "mean_token_accuracy": 0.19979041516780854,
      "num_tokens": 57337274.0,
      "step": 22625
    },
    {
      "entropy": 6.317932176589966,
      "epoch": 2.611656087709175,
      "grad_norm": 1.0,
      "learning_rate": 0.0004328306020804099,
      "loss": 5.6905,
      "mean_token_accuracy": 0.199943907558918,
      "num_tokens": 57349827.0,
      "step": 22630
    },
    {
      "entropy": 6.28595290184021,
      "epoch": 2.6122331217541834,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043280119033804184,
      "loss": 5.6862,
      "mean_token_accuracy": 0.1980353906750679,
      "num_tokens": 57361729.0,
      "step": 22635
    },
    {
      "entropy": 6.172456693649292,
      "epoch": 2.6128101557991923,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043277177328810914,
      "loss": 5.4209,
      "mean_token_accuracy": 0.2269761934876442,
      "num_tokens": 57375241.0,
      "step": 22640
    },
    {
      "entropy": 6.213561868667602,
      "epoch": 2.6133871898442007,
      "grad_norm": 1.0,
      "learning_rate": 0.0004327423509316012,
      "loss": 5.5671,
      "mean_token_accuracy": 0.21761482954025269,
      "num_tokens": 57388283.0,
      "step": 22645
    },
    {
      "entropy": 6.239810514450073,
      "epoch": 2.6139642238892096,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043271292326950764,
      "loss": 5.6019,
      "mean_token_accuracy": 0.2028694435954094,
      "num_tokens": 57400149.0,
      "step": 22650
    },
    {
      "entropy": 6.21050443649292,
      "epoch": 2.614541257934218,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004326834903028182,
      "loss": 5.5866,
      "mean_token_accuracy": 0.2107081741094589,
      "num_tokens": 57413543.0,
      "step": 22655
    },
    {
      "entropy": 6.275677442550659,
      "epoch": 2.615118291979227,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000432654052032523,
      "loss": 5.6509,
      "mean_token_accuracy": 0.19843944013118744,
      "num_tokens": 57425698.0,
      "step": 22660
    },
    {
      "entropy": 6.262874746322632,
      "epoch": 2.6156953260242353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000432624608459612,
      "loss": 5.6697,
      "mean_token_accuracy": 0.19979893118143083,
      "num_tokens": 57439243.0,
      "step": 22665
    },
    {
      "entropy": 6.221421432495117,
      "epoch": 2.616272360069244,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004325951595850756,
      "loss": 5.5589,
      "mean_token_accuracy": 0.21233202517032623,
      "num_tokens": 57451361.0,
      "step": 22670
    },
    {
      "entropy": 6.265252351760864,
      "epoch": 2.6168493941142525,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043256570540990425,
      "loss": 5.6569,
      "mean_token_accuracy": 0.20471830517053605,
      "num_tokens": 57464373.0,
      "step": 22675
    },
    {
      "entropy": 6.254950094223022,
      "epoch": 2.6174264281592614,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004325362459350888,
      "loss": 5.6141,
      "mean_token_accuracy": 0.2071797013282776,
      "num_tokens": 57477686.0,
      "step": 22680
    },
    {
      "entropy": 6.2188409805297855,
      "epoch": 2.6180034622042703,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043250678116161985,
      "loss": 5.5845,
      "mean_token_accuracy": 0.20984272360801698,
      "num_tokens": 57490632.0,
      "step": 22685
    },
    {
      "entropy": 6.198676013946534,
      "epoch": 2.6185804962492787,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043247731109048856,
      "loss": 5.6341,
      "mean_token_accuracy": 0.21136113703250886,
      "num_tokens": 57503042.0,
      "step": 22690
    },
    {
      "entropy": 6.291097640991211,
      "epoch": 2.619157530294287,
      "grad_norm": 0.875,
      "learning_rate": 0.0004324478357226861,
      "loss": 5.6452,
      "mean_token_accuracy": 0.2045815706253052,
      "num_tokens": 57516064.0,
      "step": 22695
    },
    {
      "entropy": 6.286149501800537,
      "epoch": 2.619734564339296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043241835505920405,
      "loss": 5.6841,
      "mean_token_accuracy": 0.20191462635993956,
      "num_tokens": 57528890.0,
      "step": 22700
    },
    {
      "entropy": 6.265228176116944,
      "epoch": 2.620311598384305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004323888691010337,
      "loss": 5.662,
      "mean_token_accuracy": 0.20395620316267013,
      "num_tokens": 57540130.0,
      "step": 22705
    },
    {
      "entropy": 6.192247629165649,
      "epoch": 2.6208886324293132,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043235937784916694,
      "loss": 5.6114,
      "mean_token_accuracy": 0.20545071363449097,
      "num_tokens": 57553438.0,
      "step": 22710
    },
    {
      "entropy": 6.207682180404663,
      "epoch": 2.621465666474322,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043232988130459567,
      "loss": 5.5266,
      "mean_token_accuracy": 0.21662035435438157,
      "num_tokens": 57566330.0,
      "step": 22715
    },
    {
      "entropy": 6.32170524597168,
      "epoch": 2.6220427005193305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000432300379468312,
      "loss": 5.7189,
      "mean_token_accuracy": 0.20327601879835128,
      "num_tokens": 57579390.0,
      "step": 22720
    },
    {
      "entropy": 6.200490951538086,
      "epoch": 2.6226197345643394,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004322708723413082,
      "loss": 5.6384,
      "mean_token_accuracy": 0.19681931436061859,
      "num_tokens": 57592285.0,
      "step": 22725
    },
    {
      "entropy": 6.296734476089478,
      "epoch": 2.623196768609348,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043224135992457667,
      "loss": 5.6087,
      "mean_token_accuracy": 0.21197021901607513,
      "num_tokens": 57604389.0,
      "step": 22730
    },
    {
      "entropy": 6.256216144561767,
      "epoch": 2.6237738026543567,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004322118422191102,
      "loss": 5.6165,
      "mean_token_accuracy": 0.20218596160411834,
      "num_tokens": 57616490.0,
      "step": 22735
    },
    {
      "entropy": 6.216982507705689,
      "epoch": 2.6243508366993655,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043218231922590144,
      "loss": 5.6482,
      "mean_token_accuracy": 0.20208438783884047,
      "num_tokens": 57629282.0,
      "step": 22740
    },
    {
      "entropy": 6.307484102249146,
      "epoch": 2.624927870744374,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004321527909459435,
      "loss": 5.6836,
      "mean_token_accuracy": 0.21159323155879975,
      "num_tokens": 57642338.0,
      "step": 22745
    },
    {
      "entropy": 6.284286212921143,
      "epoch": 2.6255049047893824,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043212325738022946,
      "loss": 5.6936,
      "mean_token_accuracy": 0.19829303473234178,
      "num_tokens": 57655649.0,
      "step": 22750
    },
    {
      "entropy": 6.255371761322022,
      "epoch": 2.6260819388343912,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004320937185297527,
      "loss": 5.7152,
      "mean_token_accuracy": 0.19068589061498642,
      "num_tokens": 57667788.0,
      "step": 22755
    },
    {
      "entropy": 6.219714164733887,
      "epoch": 2.6266589728794,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043206417439550684,
      "loss": 5.6083,
      "mean_token_accuracy": 0.21188358664512635,
      "num_tokens": 57681251.0,
      "step": 22760
    },
    {
      "entropy": 6.230580425262451,
      "epoch": 2.6272360069244085,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043203462497848534,
      "loss": 5.6799,
      "mean_token_accuracy": 0.20108882784843446,
      "num_tokens": 57694146.0,
      "step": 22765
    },
    {
      "entropy": 6.222435808181762,
      "epoch": 2.627813040969417,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043200507027968236,
      "loss": 5.6121,
      "mean_token_accuracy": 0.21273526251316072,
      "num_tokens": 57708411.0,
      "step": 22770
    },
    {
      "entropy": 6.322132968902588,
      "epoch": 2.628390075014426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004319755103000918,
      "loss": 5.6822,
      "mean_token_accuracy": 0.20687335729599,
      "num_tokens": 57721886.0,
      "step": 22775
    },
    {
      "entropy": 6.1889448165893555,
      "epoch": 2.6289671090594346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004319459450407079,
      "loss": 5.5456,
      "mean_token_accuracy": 0.21268752366304397,
      "num_tokens": 57734725.0,
      "step": 22780
    },
    {
      "entropy": 6.336852073669434,
      "epoch": 2.629544143104443,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043191637450252514,
      "loss": 5.765,
      "mean_token_accuracy": 0.20380858927965165,
      "num_tokens": 57747710.0,
      "step": 22785
    },
    {
      "entropy": 6.29661283493042,
      "epoch": 2.630121177149452,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000431886798686538,
      "loss": 5.6369,
      "mean_token_accuracy": 0.2027934432029724,
      "num_tokens": 57759281.0,
      "step": 22790
    },
    {
      "entropy": 6.2617600440979,
      "epoch": 2.6306982111944603,
      "grad_norm": 1.0,
      "learning_rate": 0.0004318572175937414,
      "loss": 5.6001,
      "mean_token_accuracy": 0.20640976577997208,
      "num_tokens": 57771304.0,
      "step": 22795
    },
    {
      "entropy": 6.326433992385864,
      "epoch": 2.631275245239469,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004318276312251301,
      "loss": 5.7466,
      "mean_token_accuracy": 0.19476875066757202,
      "num_tokens": 57783997.0,
      "step": 22800
    },
    {
      "entropy": 6.30865797996521,
      "epoch": 2.6318522792844776,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004317980395816994,
      "loss": 5.7147,
      "mean_token_accuracy": 0.19747118651866913,
      "num_tokens": 57797203.0,
      "step": 22805
    },
    {
      "entropy": 6.303142023086548,
      "epoch": 2.6324293133294865,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004317684426644445,
      "loss": 5.6151,
      "mean_token_accuracy": 0.20279213637113572,
      "num_tokens": 57809118.0,
      "step": 22810
    },
    {
      "entropy": 6.212245941162109,
      "epoch": 2.6330063473744953,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043173884047436093,
      "loss": 5.6364,
      "mean_token_accuracy": 0.20650760382413863,
      "num_tokens": 57823511.0,
      "step": 22815
    },
    {
      "entropy": 6.278103160858154,
      "epoch": 2.6335833814195038,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004317092330124443,
      "loss": 5.648,
      "mean_token_accuracy": 0.2045970305800438,
      "num_tokens": 57835906.0,
      "step": 22820
    },
    {
      "entropy": 6.247374629974365,
      "epoch": 2.634160415464512,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043167962027969043,
      "loss": 5.6599,
      "mean_token_accuracy": 0.20138547867536544,
      "num_tokens": 57848621.0,
      "step": 22825
    },
    {
      "entropy": 6.216165685653687,
      "epoch": 2.634737449509521,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004316500022770954,
      "loss": 5.6461,
      "mean_token_accuracy": 0.19816780984401702,
      "num_tokens": 57861123.0,
      "step": 22830
    },
    {
      "entropy": 6.271229839324951,
      "epoch": 2.63531448355453,
      "grad_norm": 1.703125,
      "learning_rate": 0.00043162037900565533,
      "loss": 5.696,
      "mean_token_accuracy": 0.20933477133512496,
      "num_tokens": 57873575.0,
      "step": 22835
    },
    {
      "entropy": 6.239863348007202,
      "epoch": 2.6358915175995383,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043159075046636666,
      "loss": 5.6133,
      "mean_token_accuracy": 0.20068905353546143,
      "num_tokens": 57885305.0,
      "step": 22840
    },
    {
      "entropy": 6.314450407028199,
      "epoch": 2.636468551644547,
      "grad_norm": 1.0,
      "learning_rate": 0.0004315611166602258,
      "loss": 5.6416,
      "mean_token_accuracy": 0.2039047434926033,
      "num_tokens": 57896906.0,
      "step": 22845
    },
    {
      "entropy": 6.208703517913818,
      "epoch": 2.6370455856895556,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043153147758822957,
      "loss": 5.6098,
      "mean_token_accuracy": 0.205648972094059,
      "num_tokens": 57909417.0,
      "step": 22850
    },
    {
      "entropy": 6.267827081680298,
      "epoch": 2.6376226197345645,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004315018332513749,
      "loss": 5.6585,
      "mean_token_accuracy": 0.20405511409044266,
      "num_tokens": 57922926.0,
      "step": 22855
    },
    {
      "entropy": 6.326779127120972,
      "epoch": 2.638199653779573,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004314721836506587,
      "loss": 5.8263,
      "mean_token_accuracy": 0.1899907797574997,
      "num_tokens": 57936481.0,
      "step": 22860
    },
    {
      "entropy": 6.304150390625,
      "epoch": 2.6387766878245817,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043144252878707845,
      "loss": 5.6557,
      "mean_token_accuracy": 0.20460156202316285,
      "num_tokens": 57949624.0,
      "step": 22865
    },
    {
      "entropy": 6.222836589813232,
      "epoch": 2.63935372186959,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004314128686616314,
      "loss": 5.6159,
      "mean_token_accuracy": 0.2080774039030075,
      "num_tokens": 57961775.0,
      "step": 22870
    },
    {
      "entropy": 6.281342601776123,
      "epoch": 2.639930755914599,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004313832032753152,
      "loss": 5.6949,
      "mean_token_accuracy": 0.20048809945583343,
      "num_tokens": 57974429.0,
      "step": 22875
    },
    {
      "entropy": 6.277792835235596,
      "epoch": 2.6405077899596074,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043135353262912764,
      "loss": 5.6221,
      "mean_token_accuracy": 0.20561502128839493,
      "num_tokens": 57986657.0,
      "step": 22880
    },
    {
      "entropy": 6.2450769424438475,
      "epoch": 2.6410848240046163,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043132385672406664,
      "loss": 5.576,
      "mean_token_accuracy": 0.20748951733112336,
      "num_tokens": 57999016.0,
      "step": 22885
    },
    {
      "entropy": 6.312080192565918,
      "epoch": 2.641661858049625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004312941755611305,
      "loss": 5.6543,
      "mean_token_accuracy": 0.20521320402622223,
      "num_tokens": 58009946.0,
      "step": 22890
    },
    {
      "entropy": 6.239800071716308,
      "epoch": 2.6422388920946336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043126448914131724,
      "loss": 5.6414,
      "mean_token_accuracy": 0.21291528791189193,
      "num_tokens": 58021475.0,
      "step": 22895
    },
    {
      "entropy": 6.208470630645752,
      "epoch": 2.642815926139642,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004312347974656256,
      "loss": 5.61,
      "mean_token_accuracy": 0.21001324504613877,
      "num_tokens": 58035781.0,
      "step": 22900
    },
    {
      "entropy": 6.219689226150512,
      "epoch": 2.643392960184651,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004312051005350541,
      "loss": 5.6524,
      "mean_token_accuracy": 0.20219698995351792,
      "num_tokens": 58049083.0,
      "step": 22905
    },
    {
      "entropy": 6.306261682510376,
      "epoch": 2.6439699942296597,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043117539835060164,
      "loss": 5.698,
      "mean_token_accuracy": 0.19924696385860444,
      "num_tokens": 58061931.0,
      "step": 22910
    },
    {
      "entropy": 6.273612833023071,
      "epoch": 2.644547028274668,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004311456909132673,
      "loss": 5.6533,
      "mean_token_accuracy": 0.2051278159022331,
      "num_tokens": 58075037.0,
      "step": 22915
    },
    {
      "entropy": 6.299002981185913,
      "epoch": 2.645124062319677,
      "grad_norm": 0.875,
      "learning_rate": 0.0004311159782240502,
      "loss": 5.6834,
      "mean_token_accuracy": 0.20195793658494948,
      "num_tokens": 58087593.0,
      "step": 22920
    },
    {
      "entropy": 6.294560098648072,
      "epoch": 2.6457010963646854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004310862602839496,
      "loss": 5.6181,
      "mean_token_accuracy": 0.2106034278869629,
      "num_tokens": 58101719.0,
      "step": 22925
    },
    {
      "entropy": 6.276077222824097,
      "epoch": 2.6462781304096943,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004310565370939653,
      "loss": 5.6402,
      "mean_token_accuracy": 0.2053774908185005,
      "num_tokens": 58114384.0,
      "step": 22930
    },
    {
      "entropy": 6.306130123138428,
      "epoch": 2.6468551644547027,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043102680865509687,
      "loss": 5.734,
      "mean_token_accuracy": 0.1923515811562538,
      "num_tokens": 58126365.0,
      "step": 22935
    },
    {
      "entropy": 6.3651237964630125,
      "epoch": 2.6474321984997116,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004309970749683442,
      "loss": 5.723,
      "mean_token_accuracy": 0.20485451221466064,
      "num_tokens": 58140024.0,
      "step": 22940
    },
    {
      "entropy": 6.236081886291504,
      "epoch": 2.64800923254472,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043096733603470737,
      "loss": 5.6089,
      "mean_token_accuracy": 0.2095950722694397,
      "num_tokens": 58152828.0,
      "step": 22945
    },
    {
      "entropy": 6.296378517150879,
      "epoch": 2.648586266589729,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043093759185518677,
      "loss": 5.6742,
      "mean_token_accuracy": 0.199651075899601,
      "num_tokens": 58166579.0,
      "step": 22950
    },
    {
      "entropy": 6.251994466781616,
      "epoch": 2.6491633006347373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043090784243078264,
      "loss": 5.6426,
      "mean_token_accuracy": 0.2080523431301117,
      "num_tokens": 58180788.0,
      "step": 22955
    },
    {
      "entropy": 6.289929056167603,
      "epoch": 2.649740334679746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004308780877624957,
      "loss": 5.667,
      "mean_token_accuracy": 0.19789770990610123,
      "num_tokens": 58192385.0,
      "step": 22960
    },
    {
      "entropy": 6.257386207580566,
      "epoch": 2.650317368724755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004308483278513267,
      "loss": 5.6514,
      "mean_token_accuracy": 0.20227849036455153,
      "num_tokens": 58203739.0,
      "step": 22965
    },
    {
      "entropy": 6.267434358596802,
      "epoch": 2.6508944027697634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043081856269827656,
      "loss": 5.6288,
      "mean_token_accuracy": 0.2049793615937233,
      "num_tokens": 58215576.0,
      "step": 22970
    },
    {
      "entropy": 6.307793855667114,
      "epoch": 2.651471436814772,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004307887923043465,
      "loss": 5.7507,
      "mean_token_accuracy": 0.2015164703130722,
      "num_tokens": 58228242.0,
      "step": 22975
    },
    {
      "entropy": 6.286945009231568,
      "epoch": 2.6520484708597807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004307590166705379,
      "loss": 5.665,
      "mean_token_accuracy": 0.20053587406873702,
      "num_tokens": 58241568.0,
      "step": 22980
    },
    {
      "entropy": 6.179308080673218,
      "epoch": 2.6526255049047895,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043072923579785203,
      "loss": 5.5474,
      "mean_token_accuracy": 0.21867496222257615,
      "num_tokens": 58254018.0,
      "step": 22985
    },
    {
      "entropy": 6.2373041152954105,
      "epoch": 2.653202538949798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004306994496872907,
      "loss": 5.7233,
      "mean_token_accuracy": 0.20389633923768996,
      "num_tokens": 58266617.0,
      "step": 22990
    },
    {
      "entropy": 6.240955209732055,
      "epoch": 2.653779572994807,
      "grad_norm": 1.421875,
      "learning_rate": 0.00043066965833985567,
      "loss": 5.5827,
      "mean_token_accuracy": 0.21605982035398483,
      "num_tokens": 58279546.0,
      "step": 22995
    },
    {
      "entropy": 6.30221962928772,
      "epoch": 2.6543566070398152,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043063986175654906,
      "loss": 5.7036,
      "mean_token_accuracy": 0.19921424686908723,
      "num_tokens": 58293121.0,
      "step": 23000
    },
    {
      "entropy": 6.270164680480957,
      "epoch": 2.654933641084824,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043061005993837306,
      "loss": 5.707,
      "mean_token_accuracy": 0.20191250294446944,
      "num_tokens": 58306950.0,
      "step": 23005
    },
    {
      "entropy": 6.16685380935669,
      "epoch": 2.6555106751298325,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043058025288632995,
      "loss": 5.5441,
      "mean_token_accuracy": 0.2155737340450287,
      "num_tokens": 58320439.0,
      "step": 23010
    },
    {
      "entropy": 6.301377058029175,
      "epoch": 2.6560877091748414,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043055044060142224,
      "loss": 5.6857,
      "mean_token_accuracy": 0.19749298989772796,
      "num_tokens": 58333113.0,
      "step": 23015
    },
    {
      "entropy": 6.3457215309143065,
      "epoch": 2.6566647432198502,
      "grad_norm": 1.0,
      "learning_rate": 0.00043052062308465277,
      "loss": 5.7087,
      "mean_token_accuracy": 0.1984311044216156,
      "num_tokens": 58345564.0,
      "step": 23020
    },
    {
      "entropy": 6.264771604537964,
      "epoch": 2.6572417772648587,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004304908003370244,
      "loss": 5.655,
      "mean_token_accuracy": 0.20039392113685608,
      "num_tokens": 58358004.0,
      "step": 23025
    },
    {
      "entropy": 6.241031312942505,
      "epoch": 2.657818811309867,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043046097235954024,
      "loss": 5.6581,
      "mean_token_accuracy": 0.20061101019382477,
      "num_tokens": 58370784.0,
      "step": 23030
    },
    {
      "entropy": 6.322096443176269,
      "epoch": 2.658395845354876,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043043113915320345,
      "loss": 5.6619,
      "mean_token_accuracy": 0.20258571356534957,
      "num_tokens": 58382785.0,
      "step": 23035
    },
    {
      "entropy": 6.188664102554322,
      "epoch": 2.658972879399885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043040130071901747,
      "loss": 5.6471,
      "mean_token_accuracy": 0.20759778022766112,
      "num_tokens": 58395674.0,
      "step": 23040
    },
    {
      "entropy": 6.2338052749633786,
      "epoch": 2.659549913444893,
      "grad_norm": 1.0,
      "learning_rate": 0.00043037145705798605,
      "loss": 5.683,
      "mean_token_accuracy": 0.20500317811965943,
      "num_tokens": 58408867.0,
      "step": 23045
    },
    {
      "entropy": 6.32915997505188,
      "epoch": 2.6601269474899016,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043034160817111275,
      "loss": 5.6841,
      "mean_token_accuracy": 0.19876789897680283,
      "num_tokens": 58421082.0,
      "step": 23050
    },
    {
      "entropy": 6.248715400695801,
      "epoch": 2.6607039815349105,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043031175405940157,
      "loss": 5.6164,
      "mean_token_accuracy": 0.20372868925333024,
      "num_tokens": 58435325.0,
      "step": 23055
    },
    {
      "entropy": 6.299409532546997,
      "epoch": 2.6612810155799194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004302818947238568,
      "loss": 5.6962,
      "mean_token_accuracy": 0.2053624540567398,
      "num_tokens": 58447640.0,
      "step": 23060
    },
    {
      "entropy": 6.274459362030029,
      "epoch": 2.661858049624928,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043025203016548263,
      "loss": 5.6678,
      "mean_token_accuracy": 0.2044256642460823,
      "num_tokens": 58460047.0,
      "step": 23065
    },
    {
      "entropy": 6.282318258285523,
      "epoch": 2.6624350836699366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043022216038528347,
      "loss": 5.6728,
      "mean_token_accuracy": 0.19391516745090484,
      "num_tokens": 58472130.0,
      "step": 23070
    },
    {
      "entropy": 6.193539142608643,
      "epoch": 2.663012117714945,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043019228538426416,
      "loss": 5.6171,
      "mean_token_accuracy": 0.20757988840341568,
      "num_tokens": 58486046.0,
      "step": 23075
    },
    {
      "entropy": 6.189791107177735,
      "epoch": 2.663589151759954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043016240516342933,
      "loss": 5.5756,
      "mean_token_accuracy": 0.21357823610305787,
      "num_tokens": 58499503.0,
      "step": 23080
    },
    {
      "entropy": 6.322407627105713,
      "epoch": 2.6641661858049623,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043013251972378404,
      "loss": 5.6905,
      "mean_token_accuracy": 0.20160626471042634,
      "num_tokens": 58511342.0,
      "step": 23085
    },
    {
      "entropy": 6.1234307289123535,
      "epoch": 2.664743219849971,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043010262906633355,
      "loss": 5.5142,
      "mean_token_accuracy": 0.2231653317809105,
      "num_tokens": 58524340.0,
      "step": 23090
    },
    {
      "entropy": 6.227146434783935,
      "epoch": 2.66532025389498,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043007273319208326,
      "loss": 5.6288,
      "mean_token_accuracy": 0.20923641622066497,
      "num_tokens": 58536787.0,
      "step": 23095
    },
    {
      "entropy": 6.199090957641602,
      "epoch": 2.6658972879399885,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004300428321020385,
      "loss": 5.5791,
      "mean_token_accuracy": 0.21574901789426804,
      "num_tokens": 58549983.0,
      "step": 23100
    },
    {
      "entropy": 6.33907036781311,
      "epoch": 2.666474321984997,
      "grad_norm": 1.0,
      "learning_rate": 0.00043001292579720514,
      "loss": 5.7218,
      "mean_token_accuracy": 0.19605542719364166,
      "num_tokens": 58561162.0,
      "step": 23105
    },
    {
      "entropy": 6.280099153518677,
      "epoch": 2.6670513560300058,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000429983014278589,
      "loss": 5.6631,
      "mean_token_accuracy": 0.20181227326393128,
      "num_tokens": 58574296.0,
      "step": 23110
    },
    {
      "entropy": 6.222842597961426,
      "epoch": 2.6676283900750146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004299530975471962,
      "loss": 5.6999,
      "mean_token_accuracy": 0.198213329911232,
      "num_tokens": 58587888.0,
      "step": 23115
    },
    {
      "entropy": 6.340534639358521,
      "epoch": 2.668205424120023,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004299231756040329,
      "loss": 5.7921,
      "mean_token_accuracy": 0.19888650327920915,
      "num_tokens": 58600253.0,
      "step": 23120
    },
    {
      "entropy": 6.2918689250946045,
      "epoch": 2.668782458165032,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004298932484501055,
      "loss": 5.6585,
      "mean_token_accuracy": 0.20706347823143006,
      "num_tokens": 58613006.0,
      "step": 23125
    },
    {
      "entropy": 6.246803283691406,
      "epoch": 2.6693594922100403,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042986331608642076,
      "loss": 5.6618,
      "mean_token_accuracy": 0.20935687869787217,
      "num_tokens": 58625785.0,
      "step": 23130
    },
    {
      "entropy": 6.2685047626495365,
      "epoch": 2.669936526255049,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042983337851398525,
      "loss": 5.7075,
      "mean_token_accuracy": 0.20124064534902572,
      "num_tokens": 58638937.0,
      "step": 23135
    },
    {
      "entropy": 6.320172357559204,
      "epoch": 2.6705135603000576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042980343573380595,
      "loss": 5.6531,
      "mean_token_accuracy": 0.20898209363222123,
      "num_tokens": 58651845.0,
      "step": 23140
    },
    {
      "entropy": 6.265689373016357,
      "epoch": 2.6710905943450665,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042977348774688996,
      "loss": 5.6168,
      "mean_token_accuracy": 0.2005947232246399,
      "num_tokens": 58665213.0,
      "step": 23145
    },
    {
      "entropy": 6.2076269626617435,
      "epoch": 2.671667628390075,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004297435345542445,
      "loss": 5.6536,
      "mean_token_accuracy": 0.2096429631114006,
      "num_tokens": 58677785.0,
      "step": 23150
    },
    {
      "entropy": 6.1867907524108885,
      "epoch": 2.6722446624350837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004297135761568773,
      "loss": 5.6316,
      "mean_token_accuracy": 0.2027116596698761,
      "num_tokens": 58691650.0,
      "step": 23155
    },
    {
      "entropy": 6.16974892616272,
      "epoch": 2.672821696480092,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042968361255579573,
      "loss": 5.5197,
      "mean_token_accuracy": 0.21940902322530748,
      "num_tokens": 58703790.0,
      "step": 23160
    },
    {
      "entropy": 6.279945325851441,
      "epoch": 2.673398730525101,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042965364375200765,
      "loss": 5.6845,
      "mean_token_accuracy": 0.1994474485516548,
      "num_tokens": 58715980.0,
      "step": 23165
    },
    {
      "entropy": 6.299324035644531,
      "epoch": 2.67397576457011,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004296236697465212,
      "loss": 5.6591,
      "mean_token_accuracy": 0.20208367556333542,
      "num_tokens": 58728834.0,
      "step": 23170
    },
    {
      "entropy": 6.29770245552063,
      "epoch": 2.6745527986151183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004295936905403444,
      "loss": 5.6715,
      "mean_token_accuracy": 0.20090619176626207,
      "num_tokens": 58741265.0,
      "step": 23175
    },
    {
      "entropy": 6.210208559036255,
      "epoch": 2.6751298326601267,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004295637061344855,
      "loss": 5.6076,
      "mean_token_accuracy": 0.20512702614068984,
      "num_tokens": 58754173.0,
      "step": 23180
    },
    {
      "entropy": 6.243854141235351,
      "epoch": 2.6757068667051356,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004295337165299532,
      "loss": 5.6796,
      "mean_token_accuracy": 0.19910730719566344,
      "num_tokens": 58766849.0,
      "step": 23185
    },
    {
      "entropy": 6.192908573150635,
      "epoch": 2.6762839007501444,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042950372172775614,
      "loss": 5.5556,
      "mean_token_accuracy": 0.22330635339021682,
      "num_tokens": 58779789.0,
      "step": 23190
    },
    {
      "entropy": 6.197258377075196,
      "epoch": 2.676860934795153,
      "grad_norm": 1.03125,
      "learning_rate": 0.000429473721728903,
      "loss": 5.5998,
      "mean_token_accuracy": 0.20602594017982484,
      "num_tokens": 58792797.0,
      "step": 23195
    },
    {
      "entropy": 6.24486813545227,
      "epoch": 2.6774379688401617,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004294437165344031,
      "loss": 5.5617,
      "mean_token_accuracy": 0.21087709963321685,
      "num_tokens": 58804673.0,
      "step": 23200
    },
    {
      "entropy": 6.234611225128174,
      "epoch": 2.67801500288517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042941370614526554,
      "loss": 5.6178,
      "mean_token_accuracy": 0.20376714318990707,
      "num_tokens": 58816422.0,
      "step": 23205
    },
    {
      "entropy": 6.257846975326538,
      "epoch": 2.678592036930179,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004293836905624996,
      "loss": 5.6501,
      "mean_token_accuracy": 0.20497774332761765,
      "num_tokens": 58828759.0,
      "step": 23210
    },
    {
      "entropy": 6.229918479919434,
      "epoch": 2.6791690709751874,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042935366978711487,
      "loss": 5.6843,
      "mean_token_accuracy": 0.19918799698352813,
      "num_tokens": 58841385.0,
      "step": 23215
    },
    {
      "entropy": 6.247800970077515,
      "epoch": 2.6797461050201963,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042932364382012125,
      "loss": 5.5736,
      "mean_token_accuracy": 0.21655896306037903,
      "num_tokens": 58854215.0,
      "step": 23220
    },
    {
      "entropy": 6.237952184677124,
      "epoch": 2.6803231390652047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042929361266252843,
      "loss": 5.6628,
      "mean_token_accuracy": 0.20530138313770294,
      "num_tokens": 58865464.0,
      "step": 23225
    },
    {
      "entropy": 6.130632781982422,
      "epoch": 2.6809001731102136,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004292635763153466,
      "loss": 5.5416,
      "mean_token_accuracy": 0.21721777468919753,
      "num_tokens": 58877820.0,
      "step": 23230
    },
    {
      "entropy": 6.230659055709839,
      "epoch": 2.681477207155222,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042923353477958607,
      "loss": 5.5863,
      "mean_token_accuracy": 0.20971031934022905,
      "num_tokens": 58889440.0,
      "step": 23235
    },
    {
      "entropy": 6.174967288970947,
      "epoch": 2.682054241200231,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042920348805625716,
      "loss": 5.6184,
      "mean_token_accuracy": 0.2122014805674553,
      "num_tokens": 58903369.0,
      "step": 23240
    },
    {
      "entropy": 6.2576392650604244,
      "epoch": 2.6826312752452397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042917343614637057,
      "loss": 5.6391,
      "mean_token_accuracy": 0.19939488619565965,
      "num_tokens": 58916315.0,
      "step": 23245
    },
    {
      "entropy": 6.293243026733398,
      "epoch": 2.683208309290248,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042914337905093695,
      "loss": 5.652,
      "mean_token_accuracy": 0.20569158345460892,
      "num_tokens": 58931042.0,
      "step": 23250
    },
    {
      "entropy": 6.3168741226196286,
      "epoch": 2.6837853433352565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004291133167709674,
      "loss": 5.6981,
      "mean_token_accuracy": 0.20053772628307343,
      "num_tokens": 58944455.0,
      "step": 23255
    },
    {
      "entropy": 6.2635424613952635,
      "epoch": 2.6843623773802654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042908324930747297,
      "loss": 5.6496,
      "mean_token_accuracy": 0.20859711170196532,
      "num_tokens": 58958666.0,
      "step": 23260
    },
    {
      "entropy": 6.230237579345703,
      "epoch": 2.6849394114252743,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042905317666146496,
      "loss": 5.6389,
      "mean_token_accuracy": 0.2071678876876831,
      "num_tokens": 58971550.0,
      "step": 23265
    },
    {
      "entropy": 6.230517625808716,
      "epoch": 2.6855164454702827,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042902309883395497,
      "loss": 5.6013,
      "mean_token_accuracy": 0.2085379034280777,
      "num_tokens": 58984744.0,
      "step": 23270
    },
    {
      "entropy": 6.320356798171997,
      "epoch": 2.6860934795152915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004289930158259545,
      "loss": 5.7177,
      "mean_token_accuracy": 0.19658586084842683,
      "num_tokens": 58996711.0,
      "step": 23275
    },
    {
      "entropy": 6.3604302406311035,
      "epoch": 2.6866705135603,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042896292763847544,
      "loss": 5.7437,
      "mean_token_accuracy": 0.19284067898988724,
      "num_tokens": 59010525.0,
      "step": 23280
    },
    {
      "entropy": 6.277186918258667,
      "epoch": 2.687247547605309,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042893283427252984,
      "loss": 5.6151,
      "mean_token_accuracy": 0.20588470697402955,
      "num_tokens": 59023004.0,
      "step": 23285
    },
    {
      "entropy": 6.290698146820068,
      "epoch": 2.6878245816503172,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004289027357291297,
      "loss": 5.7242,
      "mean_token_accuracy": 0.19824464172124862,
      "num_tokens": 59036664.0,
      "step": 23290
    },
    {
      "entropy": 6.2314235210418705,
      "epoch": 2.688401615695326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004288726320092876,
      "loss": 5.5897,
      "mean_token_accuracy": 0.21747675240039827,
      "num_tokens": 59049246.0,
      "step": 23295
    },
    {
      "entropy": 6.281346178054809,
      "epoch": 2.6889786497403345,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004288425231140159,
      "loss": 5.5725,
      "mean_token_accuracy": 0.21464629620313644,
      "num_tokens": 59061835.0,
      "step": 23300
    },
    {
      "entropy": 6.311240196228027,
      "epoch": 2.6895556837853434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042881240904432737,
      "loss": 5.7062,
      "mean_token_accuracy": 0.20226139426231385,
      "num_tokens": 59074863.0,
      "step": 23305
    },
    {
      "entropy": 6.252490997314453,
      "epoch": 2.690132717830352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004287822898012348,
      "loss": 5.6394,
      "mean_token_accuracy": 0.201644666492939,
      "num_tokens": 59087924.0,
      "step": 23310
    },
    {
      "entropy": 6.1347860336303714,
      "epoch": 2.6907097518753607,
      "grad_norm": 1.125,
      "learning_rate": 0.0004287521653857514,
      "loss": 5.5709,
      "mean_token_accuracy": 0.21733736842870713,
      "num_tokens": 59100725.0,
      "step": 23315
    },
    {
      "entropy": 6.229749441146851,
      "epoch": 2.6912867859203695,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042872203579889025,
      "loss": 5.5007,
      "mean_token_accuracy": 0.21703790873289108,
      "num_tokens": 59113490.0,
      "step": 23320
    },
    {
      "entropy": 6.327892208099366,
      "epoch": 2.691863819965378,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042869190104166476,
      "loss": 5.7208,
      "mean_token_accuracy": 0.1980925366282463,
      "num_tokens": 59126323.0,
      "step": 23325
    },
    {
      "entropy": 6.16375699043274,
      "epoch": 2.6924408540103864,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004286617611150886,
      "loss": 5.4718,
      "mean_token_accuracy": 0.21402792036533355,
      "num_tokens": 59138415.0,
      "step": 23330
    },
    {
      "entropy": 6.226261711120605,
      "epoch": 2.693017888055395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042863161602017536,
      "loss": 5.6154,
      "mean_token_accuracy": 0.20910578072071076,
      "num_tokens": 59150071.0,
      "step": 23335
    },
    {
      "entropy": 6.2020410060882565,
      "epoch": 2.693594922100404,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004286014657579391,
      "loss": 5.5836,
      "mean_token_accuracy": 0.21050347834825517,
      "num_tokens": 59164006.0,
      "step": 23340
    },
    {
      "entropy": 6.199285697937012,
      "epoch": 2.6941719561454125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004285713103293938,
      "loss": 5.5539,
      "mean_token_accuracy": 0.21524437963962556,
      "num_tokens": 59175316.0,
      "step": 23345
    },
    {
      "entropy": 6.189577436447143,
      "epoch": 2.6947489901904214,
      "grad_norm": 1.0,
      "learning_rate": 0.0004285411497355537,
      "loss": 5.6026,
      "mean_token_accuracy": 0.21412765830755234,
      "num_tokens": 59187802.0,
      "step": 23350
    },
    {
      "entropy": 6.307049989700317,
      "epoch": 2.6953260242354298,
      "grad_norm": 1.0,
      "learning_rate": 0.00042851098397743334,
      "loss": 5.7371,
      "mean_token_accuracy": 0.19934887140989305,
      "num_tokens": 59200681.0,
      "step": 23355
    },
    {
      "entropy": 6.316560363769531,
      "epoch": 2.6959030582804386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004284808130560473,
      "loss": 5.6585,
      "mean_token_accuracy": 0.20682025849819183,
      "num_tokens": 59213947.0,
      "step": 23360
    },
    {
      "entropy": 6.336492729187012,
      "epoch": 2.696480092325447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042845063697241034,
      "loss": 5.7509,
      "mean_token_accuracy": 0.1935090571641922,
      "num_tokens": 59226526.0,
      "step": 23365
    },
    {
      "entropy": 6.291465950012207,
      "epoch": 2.697057126370456,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004284204557275374,
      "loss": 5.6723,
      "mean_token_accuracy": 0.20713513046503068,
      "num_tokens": 59239196.0,
      "step": 23370
    },
    {
      "entropy": 6.250028657913208,
      "epoch": 2.6976341604154648,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004283902693224437,
      "loss": 5.6466,
      "mean_token_accuracy": 0.20417911410331727,
      "num_tokens": 59252824.0,
      "step": 23375
    },
    {
      "entropy": 6.32530517578125,
      "epoch": 2.698211194460473,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042836007775814446,
      "loss": 5.7223,
      "mean_token_accuracy": 0.20550840348005295,
      "num_tokens": 59265348.0,
      "step": 23380
    },
    {
      "entropy": 6.386385774612426,
      "epoch": 2.6987882285054816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004283298810356551,
      "loss": 5.7494,
      "mean_token_accuracy": 0.20172426700592042,
      "num_tokens": 59277882.0,
      "step": 23385
    },
    {
      "entropy": 6.338903141021729,
      "epoch": 2.6993652625504905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004282996791559914,
      "loss": 5.6797,
      "mean_token_accuracy": 0.2042370468378067,
      "num_tokens": 59290979.0,
      "step": 23390
    },
    {
      "entropy": 6.175904273986816,
      "epoch": 2.6999422965954993,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042826947212016924,
      "loss": 5.6385,
      "mean_token_accuracy": 0.20863338112831115,
      "num_tokens": 59305043.0,
      "step": 23395
    },
    {
      "entropy": 6.243907403945923,
      "epoch": 2.7005193306405078,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004282392599292045,
      "loss": 5.5865,
      "mean_token_accuracy": 0.21015459150075913,
      "num_tokens": 59318981.0,
      "step": 23400
    },
    {
      "entropy": 6.17451491355896,
      "epoch": 2.701096364685516,
      "grad_norm": 1.0,
      "learning_rate": 0.0004282090425841133,
      "loss": 5.575,
      "mean_token_accuracy": 0.21415877491235732,
      "num_tokens": 59331627.0,
      "step": 23405
    },
    {
      "entropy": 6.263724422454834,
      "epoch": 2.701673398730525,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004281788200859121,
      "loss": 5.5861,
      "mean_token_accuracy": 0.20921627283096314,
      "num_tokens": 59343652.0,
      "step": 23410
    },
    {
      "entropy": 6.350774717330933,
      "epoch": 2.702250432775534,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042814859243561727,
      "loss": 5.7141,
      "mean_token_accuracy": 0.19527125656604766,
      "num_tokens": 59355667.0,
      "step": 23415
    },
    {
      "entropy": 6.25461630821228,
      "epoch": 2.7028274668205423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042811835963424573,
      "loss": 5.6628,
      "mean_token_accuracy": 0.20457807630300523,
      "num_tokens": 59369021.0,
      "step": 23420
    },
    {
      "entropy": 6.304637718200683,
      "epoch": 2.703404500865551,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042808812168281413,
      "loss": 5.6478,
      "mean_token_accuracy": 0.20571892261505126,
      "num_tokens": 59381052.0,
      "step": 23425
    },
    {
      "entropy": 6.331337118148804,
      "epoch": 2.7039815349105596,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042805787858233966,
      "loss": 5.6708,
      "mean_token_accuracy": 0.20368467420339584,
      "num_tokens": 59395564.0,
      "step": 23430
    },
    {
      "entropy": 6.295053339004516,
      "epoch": 2.7045585689555685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004280276303338395,
      "loss": 5.7383,
      "mean_token_accuracy": 0.1965961441397667,
      "num_tokens": 59409061.0,
      "step": 23435
    },
    {
      "entropy": 6.226743125915528,
      "epoch": 2.705135603000577,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000427997376938331,
      "loss": 5.5965,
      "mean_token_accuracy": 0.2119791626930237,
      "num_tokens": 59422965.0,
      "step": 23440
    },
    {
      "entropy": 6.320301675796509,
      "epoch": 2.7057126370455857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004279671183968317,
      "loss": 5.7582,
      "mean_token_accuracy": 0.19646479338407516,
      "num_tokens": 59434282.0,
      "step": 23445
    },
    {
      "entropy": 6.275979518890381,
      "epoch": 2.7062896710905946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004279368547103593,
      "loss": 5.6963,
      "mean_token_accuracy": 0.19754130244255066,
      "num_tokens": 59447466.0,
      "step": 23450
    },
    {
      "entropy": 6.252089977264404,
      "epoch": 2.706866705135603,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004279065858799318,
      "loss": 5.6155,
      "mean_token_accuracy": 0.20897899568080902,
      "num_tokens": 59460916.0,
      "step": 23455
    },
    {
      "entropy": 6.295191287994385,
      "epoch": 2.7074437391806114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042787631190656725,
      "loss": 5.6778,
      "mean_token_accuracy": 0.20215128511190414,
      "num_tokens": 59473515.0,
      "step": 23460
    },
    {
      "entropy": 6.271441555023193,
      "epoch": 2.7080207732256203,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042784603279128386,
      "loss": 5.6345,
      "mean_token_accuracy": 0.20626669973134995,
      "num_tokens": 59486143.0,
      "step": 23465
    },
    {
      "entropy": 6.266502475738525,
      "epoch": 2.708597807270629,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004278157485351001,
      "loss": 5.6305,
      "mean_token_accuracy": 0.20225792676210402,
      "num_tokens": 59498206.0,
      "step": 23470
    },
    {
      "entropy": 6.271490097045898,
      "epoch": 2.7091748413156376,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042778545913903454,
      "loss": 5.5869,
      "mean_token_accuracy": 0.21201830804347993,
      "num_tokens": 59510863.0,
      "step": 23475
    },
    {
      "entropy": 6.328682231903076,
      "epoch": 2.7097518753606464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042775516460410594,
      "loss": 5.6603,
      "mean_token_accuracy": 0.20773557722568511,
      "num_tokens": 59523539.0,
      "step": 23480
    },
    {
      "entropy": 6.282889318466187,
      "epoch": 2.710328909405655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004277248649313333,
      "loss": 5.645,
      "mean_token_accuracy": 0.2074839100241661,
      "num_tokens": 59535150.0,
      "step": 23485
    },
    {
      "entropy": 6.200600004196167,
      "epoch": 2.7109059434506637,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004276945601217357,
      "loss": 5.5657,
      "mean_token_accuracy": 0.21318063735961915,
      "num_tokens": 59547985.0,
      "step": 23490
    },
    {
      "entropy": 6.312023019790649,
      "epoch": 2.711482977495672,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042766425017633236,
      "loss": 5.6602,
      "mean_token_accuracy": 0.20149388164281845,
      "num_tokens": 59561378.0,
      "step": 23495
    },
    {
      "entropy": 6.300187301635742,
      "epoch": 2.712060011540681,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042763393509614284,
      "loss": 5.6534,
      "mean_token_accuracy": 0.2078022837638855,
      "num_tokens": 59574168.0,
      "step": 23500
    },
    {
      "entropy": 6.204970979690552,
      "epoch": 2.7126370455856894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004276036148821867,
      "loss": 5.5754,
      "mean_token_accuracy": 0.2109737366437912,
      "num_tokens": 59587084.0,
      "step": 23505
    },
    {
      "entropy": 6.177952575683594,
      "epoch": 2.7132140796306983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042757328953548385,
      "loss": 5.5589,
      "mean_token_accuracy": 0.21338616758584977,
      "num_tokens": 59601272.0,
      "step": 23510
    },
    {
      "entropy": 6.217683601379394,
      "epoch": 2.7137911136757067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004275429590570541,
      "loss": 5.6256,
      "mean_token_accuracy": 0.20464665144681932,
      "num_tokens": 59613525.0,
      "step": 23515
    },
    {
      "entropy": 6.163472127914429,
      "epoch": 2.7143681477207156,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00042751262344791776,
      "loss": 5.5276,
      "mean_token_accuracy": 0.2282908082008362,
      "num_tokens": 59627269.0,
      "step": 23520
    },
    {
      "entropy": 6.238416004180908,
      "epoch": 2.7149451817657244,
      "grad_norm": 1.0,
      "learning_rate": 0.000427482282709095,
      "loss": 5.675,
      "mean_token_accuracy": 0.20566678345203399,
      "num_tokens": 59639682.0,
      "step": 23525
    },
    {
      "entropy": 6.231796503067017,
      "epoch": 2.715522215810733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004274519368416065,
      "loss": 5.6663,
      "mean_token_accuracy": 0.20691551864147187,
      "num_tokens": 59653358.0,
      "step": 23530
    },
    {
      "entropy": 6.264365005493164,
      "epoch": 2.7160992498557412,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004274215858464727,
      "loss": 5.5771,
      "mean_token_accuracy": 0.2075978323817253,
      "num_tokens": 59665313.0,
      "step": 23535
    },
    {
      "entropy": 6.277977800369262,
      "epoch": 2.71667628390075,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042739122972471473,
      "loss": 5.6382,
      "mean_token_accuracy": 0.2045588433742523,
      "num_tokens": 59678626.0,
      "step": 23540
    },
    {
      "entropy": 6.3153321743011475,
      "epoch": 2.717253317945759,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004273608684773534,
      "loss": 5.6904,
      "mean_token_accuracy": 0.2029748186469078,
      "num_tokens": 59690839.0,
      "step": 23545
    },
    {
      "entropy": 6.168160486221313,
      "epoch": 2.7178303519907674,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042733050210540984,
      "loss": 5.557,
      "mean_token_accuracy": 0.2026698485016823,
      "num_tokens": 59704418.0,
      "step": 23550
    },
    {
      "entropy": 6.306282997131348,
      "epoch": 2.7184073860357763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004273001306099056,
      "loss": 5.6219,
      "mean_token_accuracy": 0.20672982484102248,
      "num_tokens": 59716293.0,
      "step": 23555
    },
    {
      "entropy": 6.2866480350494385,
      "epoch": 2.7189844200807847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004272697539918621,
      "loss": 5.7353,
      "mean_token_accuracy": 0.19560708105564117,
      "num_tokens": 59727917.0,
      "step": 23560
    },
    {
      "entropy": 6.244938755035401,
      "epoch": 2.7195614541257935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004272393722523011,
      "loss": 5.6818,
      "mean_token_accuracy": 0.20307622104883194,
      "num_tokens": 59740132.0,
      "step": 23565
    },
    {
      "entropy": 6.284699869155884,
      "epoch": 2.720138488170802,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042720898539224436,
      "loss": 5.7006,
      "mean_token_accuracy": 0.2046368360519409,
      "num_tokens": 59753183.0,
      "step": 23570
    },
    {
      "entropy": 6.352438640594483,
      "epoch": 2.720715522215811,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004271785934127141,
      "loss": 5.7025,
      "mean_token_accuracy": 0.19485576152801515,
      "num_tokens": 59765848.0,
      "step": 23575
    },
    {
      "entropy": 6.299515581130981,
      "epoch": 2.7212925562608192,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042714819631473234,
      "loss": 5.6725,
      "mean_token_accuracy": 0.1992581903934479,
      "num_tokens": 59777836.0,
      "step": 23580
    },
    {
      "entropy": 6.292465782165527,
      "epoch": 2.721869590305828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042711779409932164,
      "loss": 5.7068,
      "mean_token_accuracy": 0.20443301945924758,
      "num_tokens": 59790186.0,
      "step": 23585
    },
    {
      "entropy": 6.21800651550293,
      "epoch": 2.7224466243508365,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042708738676750446,
      "loss": 5.6074,
      "mean_token_accuracy": 0.2115076869726181,
      "num_tokens": 59802527.0,
      "step": 23590
    },
    {
      "entropy": 6.192450428009034,
      "epoch": 2.7230236583958454,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004270569743203036,
      "loss": 5.6073,
      "mean_token_accuracy": 0.21003096848726271,
      "num_tokens": 59814543.0,
      "step": 23595
    },
    {
      "entropy": 6.310233879089355,
      "epoch": 2.7236006924408542,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004270265567587419,
      "loss": 5.6662,
      "mean_token_accuracy": 0.20210474878549575,
      "num_tokens": 59826587.0,
      "step": 23600
    },
    {
      "entropy": 6.329673004150391,
      "epoch": 2.7241777264858626,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042699613408384246,
      "loss": 5.7031,
      "mean_token_accuracy": 0.19868629425764084,
      "num_tokens": 59839854.0,
      "step": 23605
    },
    {
      "entropy": 6.260150194168091,
      "epoch": 2.724754760530871,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004269657062966286,
      "loss": 5.6717,
      "mean_token_accuracy": 0.20644620954990386,
      "num_tokens": 59852682.0,
      "step": 23610
    },
    {
      "entropy": 6.177538681030273,
      "epoch": 2.72533179457588,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004269352733981238,
      "loss": 5.5738,
      "mean_token_accuracy": 0.2094959571957588,
      "num_tokens": 59864426.0,
      "step": 23615
    },
    {
      "entropy": 6.202883958816528,
      "epoch": 2.725908828620889,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042690483538935137,
      "loss": 5.5689,
      "mean_token_accuracy": 0.21735895574092864,
      "num_tokens": 59877349.0,
      "step": 23620
    },
    {
      "entropy": 6.319093561172485,
      "epoch": 2.726485862665897,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004268743922713354,
      "loss": 5.6734,
      "mean_token_accuracy": 0.1969464287161827,
      "num_tokens": 59889850.0,
      "step": 23625
    },
    {
      "entropy": 6.278713655471802,
      "epoch": 2.727062896710906,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042684394404509957,
      "loss": 5.6684,
      "mean_token_accuracy": 0.20274453610181808,
      "num_tokens": 59902745.0,
      "step": 23630
    },
    {
      "entropy": 6.202781248092651,
      "epoch": 2.7276399307559145,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042681349071166814,
      "loss": 5.5934,
      "mean_token_accuracy": 0.20545035600662231,
      "num_tokens": 59915557.0,
      "step": 23635
    },
    {
      "entropy": 6.290793228149414,
      "epoch": 2.7282169648009233,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042678303227206536,
      "loss": 5.6586,
      "mean_token_accuracy": 0.2095290333032608,
      "num_tokens": 59928222.0,
      "step": 23640
    },
    {
      "entropy": 6.2483546257019045,
      "epoch": 2.7287939988459318,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004267525687273157,
      "loss": 5.6141,
      "mean_token_accuracy": 0.20852120071649552,
      "num_tokens": 59940417.0,
      "step": 23645
    },
    {
      "entropy": 6.301809883117675,
      "epoch": 2.7293710328909406,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042672210007844383,
      "loss": 5.7104,
      "mean_token_accuracy": 0.2000259593129158,
      "num_tokens": 59952214.0,
      "step": 23650
    },
    {
      "entropy": 6.315805673599243,
      "epoch": 2.7299480669359495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042669162632647434,
      "loss": 5.6635,
      "mean_token_accuracy": 0.20488863736391066,
      "num_tokens": 59964625.0,
      "step": 23655
    },
    {
      "entropy": 6.255316305160522,
      "epoch": 2.730525100980958,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042666114747243243,
      "loss": 5.6634,
      "mean_token_accuracy": 0.20532287657260895,
      "num_tokens": 59978628.0,
      "step": 23660
    },
    {
      "entropy": 6.286081314086914,
      "epoch": 2.7311021350259663,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004266306635173432,
      "loss": 5.6231,
      "mean_token_accuracy": 0.20551324933767318,
      "num_tokens": 59990727.0,
      "step": 23665
    },
    {
      "entropy": 6.24440565109253,
      "epoch": 2.731679169070975,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004266001744622319,
      "loss": 5.6072,
      "mean_token_accuracy": 0.21150770038366318,
      "num_tokens": 60003038.0,
      "step": 23670
    },
    {
      "entropy": 6.130133008956909,
      "epoch": 2.732256203115984,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000426569680308124,
      "loss": 5.485,
      "mean_token_accuracy": 0.21695681512355805,
      "num_tokens": 60015895.0,
      "step": 23675
    },
    {
      "entropy": 6.330006122589111,
      "epoch": 2.7328332371609925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042653918105604524,
      "loss": 5.7776,
      "mean_token_accuracy": 0.1929958388209343,
      "num_tokens": 60027537.0,
      "step": 23680
    },
    {
      "entropy": 6.284506893157959,
      "epoch": 2.733410271206001,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042650867670702123,
      "loss": 5.6165,
      "mean_token_accuracy": 0.21411238461732865,
      "num_tokens": 60040665.0,
      "step": 23685
    },
    {
      "entropy": 6.2476255893707275,
      "epoch": 2.7339873052510097,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004264781672620783,
      "loss": 5.6383,
      "mean_token_accuracy": 0.20618118345737457,
      "num_tokens": 60052750.0,
      "step": 23690
    },
    {
      "entropy": 6.272592973709107,
      "epoch": 2.7345643392960186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004264476527222425,
      "loss": 5.6594,
      "mean_token_accuracy": 0.20323154628276824,
      "num_tokens": 60066126.0,
      "step": 23695
    },
    {
      "entropy": 6.2352776527404785,
      "epoch": 2.735141373341027,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004264171330885401,
      "loss": 5.5758,
      "mean_token_accuracy": 0.20815852880477906,
      "num_tokens": 60079325.0,
      "step": 23700
    },
    {
      "entropy": 6.246604537963867,
      "epoch": 2.735718407386036,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042638660836199756,
      "loss": 5.6021,
      "mean_token_accuracy": 0.20965514034032823,
      "num_tokens": 60092332.0,
      "step": 23705
    },
    {
      "entropy": 6.289112186431884,
      "epoch": 2.7362954414310443,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042635607854364166,
      "loss": 5.713,
      "mean_token_accuracy": 0.19646572321653366,
      "num_tokens": 60105478.0,
      "step": 23710
    },
    {
      "entropy": 6.300439262390137,
      "epoch": 2.736872475476053,
      "grad_norm": 1.0,
      "learning_rate": 0.00042632554363449926,
      "loss": 5.7156,
      "mean_token_accuracy": 0.19542830288410187,
      "num_tokens": 60117831.0,
      "step": 23715
    },
    {
      "entropy": 6.28731336593628,
      "epoch": 2.7374495095210616,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042629500363559734,
      "loss": 5.638,
      "mean_token_accuracy": 0.2045476421713829,
      "num_tokens": 60131255.0,
      "step": 23720
    },
    {
      "entropy": 6.177079200744629,
      "epoch": 2.7380265435660704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042626445854796326,
      "loss": 5.469,
      "mean_token_accuracy": 0.21846752017736434,
      "num_tokens": 60143510.0,
      "step": 23725
    },
    {
      "entropy": 6.2662615299224855,
      "epoch": 2.7386035776110793,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004262339083726241,
      "loss": 5.664,
      "mean_token_accuracy": 0.20243075489997864,
      "num_tokens": 60156620.0,
      "step": 23730
    },
    {
      "entropy": 6.2304675579071045,
      "epoch": 2.7391806116560877,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004262033531106076,
      "loss": 5.6048,
      "mean_token_accuracy": 0.20815187096595764,
      "num_tokens": 60169303.0,
      "step": 23735
    },
    {
      "entropy": 6.2367026805877686,
      "epoch": 2.739757645701096,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042617279276294145,
      "loss": 5.5989,
      "mean_token_accuracy": 0.20491737723350525,
      "num_tokens": 60181562.0,
      "step": 23740
    },
    {
      "entropy": 6.294973516464234,
      "epoch": 2.740334679746105,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004261422273306535,
      "loss": 5.6343,
      "mean_token_accuracy": 0.20499057024717332,
      "num_tokens": 60193393.0,
      "step": 23745
    },
    {
      "entropy": 6.285598945617676,
      "epoch": 2.740911713791114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004261116568147718,
      "loss": 5.6771,
      "mean_token_accuracy": 0.20535802692174912,
      "num_tokens": 60206617.0,
      "step": 23750
    },
    {
      "entropy": 6.175400018692017,
      "epoch": 2.7414887478361223,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004260810812163246,
      "loss": 5.5154,
      "mean_token_accuracy": 0.21458836495876313,
      "num_tokens": 60219909.0,
      "step": 23755
    },
    {
      "entropy": 6.24835524559021,
      "epoch": 2.742065781881131,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042605050053634027,
      "loss": 5.5644,
      "mean_token_accuracy": 0.22158487737178803,
      "num_tokens": 60234078.0,
      "step": 23760
    },
    {
      "entropy": 6.282807731628418,
      "epoch": 2.7426428159261396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004260199147758474,
      "loss": 5.6482,
      "mean_token_accuracy": 0.20004922300577163,
      "num_tokens": 60246053.0,
      "step": 23765
    },
    {
      "entropy": 6.295361471176148,
      "epoch": 2.7432198499711484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004259893239358747,
      "loss": 5.6546,
      "mean_token_accuracy": 0.20343555510044098,
      "num_tokens": 60258700.0,
      "step": 23770
    },
    {
      "entropy": 6.319237995147705,
      "epoch": 2.743796884016157,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042595872801745106,
      "loss": 5.6639,
      "mean_token_accuracy": 0.19805942475795746,
      "num_tokens": 60271989.0,
      "step": 23775
    },
    {
      "entropy": 6.322692584991455,
      "epoch": 2.7443739180611657,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004259281270216056,
      "loss": 5.6753,
      "mean_token_accuracy": 0.201534441113472,
      "num_tokens": 60283270.0,
      "step": 23780
    },
    {
      "entropy": 6.255676126480102,
      "epoch": 2.744950952106174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042589752094936763,
      "loss": 5.6432,
      "mean_token_accuracy": 0.21052137166261672,
      "num_tokens": 60295898.0,
      "step": 23785
    },
    {
      "entropy": 6.282988882064819,
      "epoch": 2.745527986151183,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004258669098017664,
      "loss": 5.6566,
      "mean_token_accuracy": 0.20503535866737366,
      "num_tokens": 60309599.0,
      "step": 23790
    },
    {
      "entropy": 6.270287370681762,
      "epoch": 2.7461050201961914,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042583629357983164,
      "loss": 5.6153,
      "mean_token_accuracy": 0.20949311852455138,
      "num_tokens": 60322126.0,
      "step": 23795
    },
    {
      "entropy": 6.291428136825561,
      "epoch": 2.7466820542412003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042580567228459303,
      "loss": 5.6119,
      "mean_token_accuracy": 0.21190867722034454,
      "num_tokens": 60334540.0,
      "step": 23800
    },
    {
      "entropy": 6.143309640884399,
      "epoch": 2.747259088286209,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004257750459170806,
      "loss": 5.536,
      "mean_token_accuracy": 0.21453500986099244,
      "num_tokens": 60346082.0,
      "step": 23805
    },
    {
      "entropy": 6.256065034866333,
      "epoch": 2.7478361223312175,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042574441447832436,
      "loss": 5.6615,
      "mean_token_accuracy": 0.20734379887580873,
      "num_tokens": 60358876.0,
      "step": 23810
    },
    {
      "entropy": 6.221592855453491,
      "epoch": 2.748413156376226,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004257137779693546,
      "loss": 5.5394,
      "mean_token_accuracy": 0.21070758253335953,
      "num_tokens": 60372398.0,
      "step": 23815
    },
    {
      "entropy": 6.312419128417969,
      "epoch": 2.748990190421235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004256831363912018,
      "loss": 5.6983,
      "mean_token_accuracy": 0.19776445478200913,
      "num_tokens": 60384332.0,
      "step": 23820
    },
    {
      "entropy": 6.291463041305542,
      "epoch": 2.7495672244662437,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042565248974489654,
      "loss": 5.6343,
      "mean_token_accuracy": 0.20623216778039932,
      "num_tokens": 60396459.0,
      "step": 23825
    },
    {
      "entropy": 6.281405925750732,
      "epoch": 2.750144258511252,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004256218380314697,
      "loss": 5.7858,
      "mean_token_accuracy": 0.19714560806751252,
      "num_tokens": 60408952.0,
      "step": 23830
    },
    {
      "entropy": 6.2261137008667,
      "epoch": 2.750721292556261,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004255911812519521,
      "loss": 5.6293,
      "mean_token_accuracy": 0.21033905595541,
      "num_tokens": 60422090.0,
      "step": 23835
    },
    {
      "entropy": 6.275562286376953,
      "epoch": 2.7512983266012694,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004255605194073749,
      "loss": 5.624,
      "mean_token_accuracy": 0.2030717983841896,
      "num_tokens": 60434397.0,
      "step": 23840
    },
    {
      "entropy": 6.16921591758728,
      "epoch": 2.7518753606462782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004255298524987695,
      "loss": 5.6319,
      "mean_token_accuracy": 0.2095465674996376,
      "num_tokens": 60447459.0,
      "step": 23845
    },
    {
      "entropy": 6.186427307128906,
      "epoch": 2.7524523946912867,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004254991805271672,
      "loss": 5.5345,
      "mean_token_accuracy": 0.2215730905532837,
      "num_tokens": 60459737.0,
      "step": 23850
    },
    {
      "entropy": 6.252324771881104,
      "epoch": 2.7530294287362955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042546850349359976,
      "loss": 5.6571,
      "mean_token_accuracy": 0.2056069016456604,
      "num_tokens": 60472721.0,
      "step": 23855
    },
    {
      "entropy": 6.278623104095459,
      "epoch": 2.753606462781304,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042543782139909894,
      "loss": 5.6608,
      "mean_token_accuracy": 0.20077745020389556,
      "num_tokens": 60486357.0,
      "step": 23860
    },
    {
      "entropy": 6.307438325881958,
      "epoch": 2.754183496826313,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042540713424469667,
      "loss": 5.752,
      "mean_token_accuracy": 0.1953136533498764,
      "num_tokens": 60499011.0,
      "step": 23865
    },
    {
      "entropy": 6.286103677749634,
      "epoch": 2.7547605308713212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042537644203142523,
      "loss": 5.6175,
      "mean_token_accuracy": 0.21314542293548583,
      "num_tokens": 60511754.0,
      "step": 23870
    },
    {
      "entropy": 6.296555280685425,
      "epoch": 2.75533756491633,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042534574476031683,
      "loss": 5.7142,
      "mean_token_accuracy": 0.19575420618057252,
      "num_tokens": 60524665.0,
      "step": 23875
    },
    {
      "entropy": 6.31353645324707,
      "epoch": 2.755914598961339,
      "grad_norm": 1.03125,
      "learning_rate": 0.000425315042432404,
      "loss": 5.654,
      "mean_token_accuracy": 0.20257467925548553,
      "num_tokens": 60536808.0,
      "step": 23880
    },
    {
      "entropy": 6.300534725189209,
      "epoch": 2.7564916330063474,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042528433504871935,
      "loss": 5.647,
      "mean_token_accuracy": 0.19970911294221877,
      "num_tokens": 60547709.0,
      "step": 23885
    },
    {
      "entropy": 6.264348077774048,
      "epoch": 2.757068667051356,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004252536226102958,
      "loss": 5.5861,
      "mean_token_accuracy": 0.20604963153600692,
      "num_tokens": 60560708.0,
      "step": 23890
    },
    {
      "entropy": 6.352797222137451,
      "epoch": 2.7576457010963646,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004252229051181662,
      "loss": 5.7261,
      "mean_token_accuracy": 0.19741826206445695,
      "num_tokens": 60573249.0,
      "step": 23895
    },
    {
      "entropy": 6.376231336593628,
      "epoch": 2.7582227351413735,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042519218257336383,
      "loss": 5.7309,
      "mean_token_accuracy": 0.19557021409273148,
      "num_tokens": 60586705.0,
      "step": 23900
    },
    {
      "entropy": 6.212790393829346,
      "epoch": 2.758799769186382,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042516145497692204,
      "loss": 5.6062,
      "mean_token_accuracy": 0.20527324974536895,
      "num_tokens": 60599489.0,
      "step": 23905
    },
    {
      "entropy": 6.245911931991577,
      "epoch": 2.759376803231391,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042513072232987426,
      "loss": 5.6316,
      "mean_token_accuracy": 0.207517372071743,
      "num_tokens": 60612551.0,
      "step": 23910
    },
    {
      "entropy": 6.2348579406738285,
      "epoch": 2.759953837276399,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004250999846332543,
      "loss": 5.5805,
      "mean_token_accuracy": 0.21807863861322402,
      "num_tokens": 60624727.0,
      "step": 23915
    },
    {
      "entropy": 6.1382183074951175,
      "epoch": 2.760530871321408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042506924188809574,
      "loss": 5.5599,
      "mean_token_accuracy": 0.2242341086268425,
      "num_tokens": 60638890.0,
      "step": 23920
    },
    {
      "entropy": 6.268954467773438,
      "epoch": 2.7611079053664165,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000425038494095433,
      "loss": 5.6329,
      "mean_token_accuracy": 0.20799211859703065,
      "num_tokens": 60650402.0,
      "step": 23925
    },
    {
      "entropy": 6.1550966739654545,
      "epoch": 2.7616849394114253,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042500774125629986,
      "loss": 5.533,
      "mean_token_accuracy": 0.2159646973013878,
      "num_tokens": 60662779.0,
      "step": 23930
    },
    {
      "entropy": 6.224617910385132,
      "epoch": 2.762261973456434,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004249769833717309,
      "loss": 5.6577,
      "mean_token_accuracy": 0.21060146391391754,
      "num_tokens": 60675158.0,
      "step": 23935
    },
    {
      "entropy": 6.262550926208496,
      "epoch": 2.7628390075014426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042494622044276066,
      "loss": 5.6645,
      "mean_token_accuracy": 0.202784064412117,
      "num_tokens": 60686872.0,
      "step": 23940
    },
    {
      "entropy": 6.223547744750976,
      "epoch": 2.763416041546451,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004249154524704237,
      "loss": 5.6204,
      "mean_token_accuracy": 0.2043646201491356,
      "num_tokens": 60699163.0,
      "step": 23945
    },
    {
      "entropy": 6.333133554458618,
      "epoch": 2.76399307559146,
      "grad_norm": 0.984375,
      "learning_rate": 0.000424884679455755,
      "loss": 5.7018,
      "mean_token_accuracy": 0.20082663744688034,
      "num_tokens": 60712331.0,
      "step": 23950
    },
    {
      "entropy": 6.289513111114502,
      "epoch": 2.7645701096364688,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042485390139978955,
      "loss": 5.6318,
      "mean_token_accuracy": 0.2053723379969597,
      "num_tokens": 60723994.0,
      "step": 23955
    },
    {
      "entropy": 6.275305271148682,
      "epoch": 2.765147143681477,
      "grad_norm": 1.0,
      "learning_rate": 0.0004248231183035626,
      "loss": 5.7095,
      "mean_token_accuracy": 0.19526817053556442,
      "num_tokens": 60736098.0,
      "step": 23960
    },
    {
      "entropy": 6.274647426605225,
      "epoch": 2.7657241777264856,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004247923301681095,
      "loss": 5.5938,
      "mean_token_accuracy": 0.2056427463889122,
      "num_tokens": 60749165.0,
      "step": 23965
    },
    {
      "entropy": 6.120019054412841,
      "epoch": 2.7663012117714945,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042476153699446567,
      "loss": 5.5305,
      "mean_token_accuracy": 0.2106404960155487,
      "num_tokens": 60761204.0,
      "step": 23970
    },
    {
      "entropy": 6.260897779464722,
      "epoch": 2.7668782458165033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042473073878366695,
      "loss": 5.6517,
      "mean_token_accuracy": 0.20802636742591857,
      "num_tokens": 60773226.0,
      "step": 23975
    },
    {
      "entropy": 6.282143211364746,
      "epoch": 2.7674552798615117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004246999355367493,
      "loss": 5.726,
      "mean_token_accuracy": 0.19036734700202942,
      "num_tokens": 60785917.0,
      "step": 23980
    },
    {
      "entropy": 6.217475366592407,
      "epoch": 2.7680323139065206,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042466912725474865,
      "loss": 5.555,
      "mean_token_accuracy": 0.21284398883581163,
      "num_tokens": 60799133.0,
      "step": 23985
    },
    {
      "entropy": 6.3027019023895265,
      "epoch": 2.768609347951529,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042463831393870123,
      "loss": 5.6906,
      "mean_token_accuracy": 0.20140222311019898,
      "num_tokens": 60810582.0,
      "step": 23990
    },
    {
      "entropy": 6.270441293716431,
      "epoch": 2.769186381996538,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042460749558964346,
      "loss": 5.6961,
      "mean_token_accuracy": 0.19686801135540008,
      "num_tokens": 60822509.0,
      "step": 23995
    },
    {
      "entropy": 6.163192844390869,
      "epoch": 2.7697634160415463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004245766722086118,
      "loss": 5.5423,
      "mean_token_accuracy": 0.21343539357185365,
      "num_tokens": 60835992.0,
      "step": 24000
    },
    {
      "epoch": 2.7697634160415463,
      "eval_entropy": 6.106378685794231,
      "eval_loss": 5.729976654052734,
      "eval_mean_token_accuracy": 0.2084515194663074,
      "eval_num_tokens": 60835992.0,
      "eval_runtime": 17.7134,
      "eval_samples_per_second": 1588.405,
      "eval_steps_per_second": 198.551,
      "step": 24000
    }
  ],
  "logging_steps": 5,
  "max_steps": 86650,
  "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": 1.329922651521024e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}