{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.390857527938451,
  "eval_steps": 3000,
  "global_step": 48000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.055009984970093,
      "epoch": 0.0005615769079575448,
      "grad_norm": 1.046875,
      "learning_rate": 2e-06,
      "loss": 6.7107,
      "mean_token_accuracy": 0.11852641776204109,
      "num_tokens": 11659.0,
      "step": 5
    },
    {
      "entropy": 7.089489412307739,
      "epoch": 0.0011231538159150896,
      "grad_norm": 1.125,
      "learning_rate": 4.5e-06,
      "loss": 6.6891,
      "mean_token_accuracy": 0.1258482627570629,
      "num_tokens": 21794.0,
      "step": 10
    },
    {
      "entropy": 7.030844259262085,
      "epoch": 0.0016847307238726343,
      "grad_norm": 1.109375,
      "learning_rate": 7e-06,
      "loss": 6.692,
      "mean_token_accuracy": 0.12024398297071456,
      "num_tokens": 33206.0,
      "step": 15
    },
    {
      "entropy": 7.072840785980224,
      "epoch": 0.0022463076318301792,
      "grad_norm": 1.0546875,
      "learning_rate": 9.5e-06,
      "loss": 6.7396,
      "mean_token_accuracy": 0.11460101678967476,
      "num_tokens": 44487.0,
      "step": 20
    },
    {
      "entropy": 7.070654821395874,
      "epoch": 0.002807884539787724,
      "grad_norm": 0.890625,
      "learning_rate": 1.2e-05,
      "loss": 6.7794,
      "mean_token_accuracy": 0.12300792410969734,
      "num_tokens": 56062.0,
      "step": 25
    },
    {
      "entropy": 7.022868919372558,
      "epoch": 0.0033694614477452686,
      "grad_norm": 1.0546875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.6531,
      "mean_token_accuracy": 0.12243497297167778,
      "num_tokens": 67453.0,
      "step": 30
    },
    {
      "entropy": 7.1308962345123295,
      "epoch": 0.003931038355702813,
      "grad_norm": 0.9140625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.7718,
      "mean_token_accuracy": 0.12407388612627983,
      "num_tokens": 78527.0,
      "step": 35
    },
    {
      "entropy": 7.098523569107056,
      "epoch": 0.0044926152636603585,
      "grad_norm": 0.9609375,
      "learning_rate": 1.95e-05,
      "loss": 6.6361,
      "mean_token_accuracy": 0.12737443149089814,
      "num_tokens": 89267.0,
      "step": 40
    },
    {
      "entropy": 7.05347261428833,
      "epoch": 0.005054192171617903,
      "grad_norm": 0.99609375,
      "learning_rate": 2.2e-05,
      "loss": 6.6533,
      "mean_token_accuracy": 0.12288121953606605,
      "num_tokens": 102035.0,
      "step": 45
    },
    {
      "entropy": 7.032813835144043,
      "epoch": 0.005615769079575448,
      "grad_norm": 1.0625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.691,
      "mean_token_accuracy": 0.12443716675043107,
      "num_tokens": 112734.0,
      "step": 50
    },
    {
      "entropy": 7.025327157974243,
      "epoch": 0.006177345987532993,
      "grad_norm": 0.83984375,
      "learning_rate": 2.7e-05,
      "loss": 6.605,
      "mean_token_accuracy": 0.12701525092124938,
      "num_tokens": 123854.0,
      "step": 55
    },
    {
      "entropy": 7.010390710830689,
      "epoch": 0.006738922895490537,
      "grad_norm": 1.0546875,
      "learning_rate": 2.95e-05,
      "loss": 6.6598,
      "mean_token_accuracy": 0.12369369864463806,
      "num_tokens": 133444.0,
      "step": 60
    },
    {
      "entropy": 7.015805435180664,
      "epoch": 0.007300499803448083,
      "grad_norm": 0.86328125,
      "learning_rate": 3.2e-05,
      "loss": 6.5399,
      "mean_token_accuracy": 0.13149982020258905,
      "num_tokens": 144678.0,
      "step": 65
    },
    {
      "entropy": 7.051552820205688,
      "epoch": 0.007862076711405626,
      "grad_norm": 1.015625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.6398,
      "mean_token_accuracy": 0.12550251558423042,
      "num_tokens": 155866.0,
      "step": 70
    },
    {
      "entropy": 7.051578426361084,
      "epoch": 0.008423653619363172,
      "grad_norm": 0.9375,
      "learning_rate": 3.7e-05,
      "loss": 6.6549,
      "mean_token_accuracy": 0.12591486871242524,
      "num_tokens": 166965.0,
      "step": 75
    },
    {
      "entropy": 7.005019807815552,
      "epoch": 0.008985230527320717,
      "grad_norm": 1.046875,
      "learning_rate": 3.95e-05,
      "loss": 6.6264,
      "mean_token_accuracy": 0.12710528150200845,
      "num_tokens": 177423.0,
      "step": 80
    },
    {
      "entropy": 7.094327354431153,
      "epoch": 0.00954680743527826,
      "grad_norm": 0.87109375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.6799,
      "mean_token_accuracy": 0.13134793564677238,
      "num_tokens": 188687.0,
      "step": 85
    },
    {
      "entropy": 7.02366008758545,
      "epoch": 0.010108384343235806,
      "grad_norm": 0.921875,
      "learning_rate": 4.45e-05,
      "loss": 6.7402,
      "mean_token_accuracy": 0.12708025872707368,
      "num_tokens": 200173.0,
      "step": 90
    },
    {
      "entropy": 7.101563358306885,
      "epoch": 0.010669961251193351,
      "grad_norm": 0.88671875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.6927,
      "mean_token_accuracy": 0.1279763586819172,
      "num_tokens": 210877.0,
      "step": 95
    },
    {
      "entropy": 7.070672464370728,
      "epoch": 0.011231538159150897,
      "grad_norm": 0.96875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.5966,
      "mean_token_accuracy": 0.12734237387776376,
      "num_tokens": 221770.0,
      "step": 100
    },
    {
      "entropy": 7.042833042144776,
      "epoch": 0.01179311506710844,
      "grad_norm": 1.0234375,
      "learning_rate": 5.2e-05,
      "loss": 6.7053,
      "mean_token_accuracy": 0.13242030143737793,
      "num_tokens": 232140.0,
      "step": 105
    },
    {
      "entropy": 7.078888416290283,
      "epoch": 0.012354691975065986,
      "grad_norm": 0.96875,
      "learning_rate": 5.45e-05,
      "loss": 6.7883,
      "mean_token_accuracy": 0.12373976334929467,
      "num_tokens": 244186.0,
      "step": 110
    },
    {
      "entropy": 7.088221263885498,
      "epoch": 0.012916268883023531,
      "grad_norm": 1.0546875,
      "learning_rate": 5.7e-05,
      "loss": 6.6365,
      "mean_token_accuracy": 0.12594474777579306,
      "num_tokens": 254805.0,
      "step": 115
    },
    {
      "entropy": 6.958014631271363,
      "epoch": 0.013477845790981074,
      "grad_norm": 0.96875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.6024,
      "mean_token_accuracy": 0.13206158876419066,
      "num_tokens": 265055.0,
      "step": 120
    },
    {
      "entropy": 6.960734224319458,
      "epoch": 0.01403942269893862,
      "grad_norm": 0.90234375,
      "learning_rate": 6.2e-05,
      "loss": 6.5299,
      "mean_token_accuracy": 0.13171669095754623,
      "num_tokens": 275850.0,
      "step": 125
    },
    {
      "entropy": 6.957069635391235,
      "epoch": 0.014600999606896165,
      "grad_norm": 0.96484375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.573,
      "mean_token_accuracy": 0.12500080540776254,
      "num_tokens": 286802.0,
      "step": 130
    },
    {
      "entropy": 7.044544267654419,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.03125,
      "learning_rate": 6.7e-05,
      "loss": 6.6832,
      "mean_token_accuracy": 0.12705129086971284,
      "num_tokens": 297070.0,
      "step": 135
    },
    {
      "entropy": 7.043524742126465,
      "epoch": 0.015724153422811252,
      "grad_norm": 0.87109375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.6043,
      "mean_token_accuracy": 0.12342865765094757,
      "num_tokens": 309617.0,
      "step": 140
    },
    {
      "entropy": 7.012852001190185,
      "epoch": 0.0162857303307688,
      "grad_norm": 1.09375,
      "learning_rate": 7.2e-05,
      "loss": 6.6383,
      "mean_token_accuracy": 0.12594973370432855,
      "num_tokens": 321310.0,
      "step": 145
    },
    {
      "entropy": 6.964358329772949,
      "epoch": 0.016847307238726343,
      "grad_norm": 1.09375,
      "learning_rate": 7.45e-05,
      "loss": 6.5806,
      "mean_token_accuracy": 0.1310770720243454,
      "num_tokens": 332243.0,
      "step": 150
    },
    {
      "entropy": 7.011226844787598,
      "epoch": 0.017408884146683887,
      "grad_norm": 0.94140625,
      "learning_rate": 7.7e-05,
      "loss": 6.6,
      "mean_token_accuracy": 0.12499173060059547,
      "num_tokens": 342847.0,
      "step": 155
    },
    {
      "entropy": 6.951114463806152,
      "epoch": 0.017970461054641434,
      "grad_norm": 1.015625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.5329,
      "mean_token_accuracy": 0.1339307375252247,
      "num_tokens": 353570.0,
      "step": 160
    },
    {
      "entropy": 7.024112272262573,
      "epoch": 0.018532037962598977,
      "grad_norm": 1.0625,
      "learning_rate": 8.2e-05,
      "loss": 6.5861,
      "mean_token_accuracy": 0.1328870877623558,
      "num_tokens": 364330.0,
      "step": 165
    },
    {
      "entropy": 6.923271131515503,
      "epoch": 0.01909361487055652,
      "grad_norm": 1.0625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.6291,
      "mean_token_accuracy": 0.12735793143510818,
      "num_tokens": 374930.0,
      "step": 170
    },
    {
      "entropy": 7.001209163665772,
      "epoch": 0.019655191778514068,
      "grad_norm": 1.0703125,
      "learning_rate": 8.7e-05,
      "loss": 6.5985,
      "mean_token_accuracy": 0.12637024670839309,
      "num_tokens": 384690.0,
      "step": 175
    },
    {
      "entropy": 7.093705129623413,
      "epoch": 0.020216768686471612,
      "grad_norm": 1.1953125,
      "learning_rate": 8.95e-05,
      "loss": 6.6222,
      "mean_token_accuracy": 0.13092779442667962,
      "num_tokens": 395753.0,
      "step": 180
    },
    {
      "entropy": 6.9514984607696535,
      "epoch": 0.020778345594429155,
      "grad_norm": 1.0625,
      "learning_rate": 9.2e-05,
      "loss": 6.5706,
      "mean_token_accuracy": 0.13586961552500726,
      "num_tokens": 407049.0,
      "step": 185
    },
    {
      "entropy": 6.99260983467102,
      "epoch": 0.021339922502386702,
      "grad_norm": 1.1484375,
      "learning_rate": 9.45e-05,
      "loss": 6.614,
      "mean_token_accuracy": 0.13324331417679786,
      "num_tokens": 417942.0,
      "step": 190
    },
    {
      "entropy": 6.963890314102173,
      "epoch": 0.021901499410344246,
      "grad_norm": 1.0625,
      "learning_rate": 9.7e-05,
      "loss": 6.6224,
      "mean_token_accuracy": 0.12833026573061942,
      "num_tokens": 428151.0,
      "step": 195
    },
    {
      "entropy": 7.052244853973389,
      "epoch": 0.022463076318301793,
      "grad_norm": 0.9609375,
      "learning_rate": 9.95e-05,
      "loss": 6.7162,
      "mean_token_accuracy": 0.12399227246642112,
      "num_tokens": 439837.0,
      "step": 200
    },
    {
      "entropy": 6.945697355270386,
      "epoch": 0.023024653226259337,
      "grad_norm": 0.96875,
      "learning_rate": 0.000102,
      "loss": 6.5155,
      "mean_token_accuracy": 0.1301749385893345,
      "num_tokens": 450357.0,
      "step": 205
    },
    {
      "entropy": 7.052839756011963,
      "epoch": 0.02358623013421688,
      "grad_norm": 1.171875,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.5709,
      "mean_token_accuracy": 0.13226377889513968,
      "num_tokens": 461174.0,
      "step": 210
    },
    {
      "entropy": 7.019093132019043,
      "epoch": 0.024147807042174427,
      "grad_norm": 1.078125,
      "learning_rate": 0.000107,
      "loss": 6.6459,
      "mean_token_accuracy": 0.12973417043685914,
      "num_tokens": 472785.0,
      "step": 215
    },
    {
      "entropy": 6.997825956344604,
      "epoch": 0.02470938395013197,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0001095,
      "loss": 6.6406,
      "mean_token_accuracy": 0.13002291321754456,
      "num_tokens": 483248.0,
      "step": 220
    },
    {
      "entropy": 6.945517063140869,
      "epoch": 0.025270960858089515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000112,
      "loss": 6.475,
      "mean_token_accuracy": 0.13533480763435363,
      "num_tokens": 494690.0,
      "step": 225
    },
    {
      "entropy": 7.020949125289917,
      "epoch": 0.025832537766047062,
      "grad_norm": 1.0,
      "learning_rate": 0.0001145,
      "loss": 6.7007,
      "mean_token_accuracy": 0.124896689504385,
      "num_tokens": 506192.0,
      "step": 230
    },
    {
      "entropy": 7.028987121582031,
      "epoch": 0.026394114674004605,
      "grad_norm": 1.046875,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.5915,
      "mean_token_accuracy": 0.12858845070004463,
      "num_tokens": 517152.0,
      "step": 235
    },
    {
      "entropy": 6.942857360839843,
      "epoch": 0.02695569158196215,
      "grad_norm": 1.09375,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.5168,
      "mean_token_accuracy": 0.133596034348011,
      "num_tokens": 527836.0,
      "step": 240
    },
    {
      "entropy": 6.9801740646362305,
      "epoch": 0.027517268489919696,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000122,
      "loss": 6.6733,
      "mean_token_accuracy": 0.1276863433420658,
      "num_tokens": 539491.0,
      "step": 245
    },
    {
      "entropy": 7.092193269729615,
      "epoch": 0.02807884539787724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001245,
      "loss": 6.654,
      "mean_token_accuracy": 0.12628911882638932,
      "num_tokens": 550816.0,
      "step": 250
    },
    {
      "entropy": 7.06174373626709,
      "epoch": 0.028640422305834783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000127,
      "loss": 6.7032,
      "mean_token_accuracy": 0.12353009209036828,
      "num_tokens": 561566.0,
      "step": 255
    },
    {
      "entropy": 6.982790279388428,
      "epoch": 0.02920199921379233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001295,
      "loss": 6.6345,
      "mean_token_accuracy": 0.12296314612030983,
      "num_tokens": 573032.0,
      "step": 260
    },
    {
      "entropy": 7.036098527908325,
      "epoch": 0.029763576121749874,
      "grad_norm": 0.953125,
      "learning_rate": 0.000132,
      "loss": 6.5545,
      "mean_token_accuracy": 0.13585324659943582,
      "num_tokens": 584140.0,
      "step": 265
    },
    {
      "entropy": 6.972643518447876,
      "epoch": 0.030325153029707418,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.625,
      "mean_token_accuracy": 0.13363742157816888,
      "num_tokens": 595097.0,
      "step": 270
    },
    {
      "entropy": 7.074052810668945,
      "epoch": 0.030886729937664965,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.6642,
      "mean_token_accuracy": 0.12557960152626038,
      "num_tokens": 606797.0,
      "step": 275
    },
    {
      "entropy": 6.958458995819091,
      "epoch": 0.031448306845622505,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001395,
      "loss": 6.6346,
      "mean_token_accuracy": 0.1282724417746067,
      "num_tokens": 618187.0,
      "step": 280
    },
    {
      "entropy": 7.020027160644531,
      "epoch": 0.032009883753580055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.5731,
      "mean_token_accuracy": 0.13057834804058074,
      "num_tokens": 629829.0,
      "step": 285
    },
    {
      "entropy": 6.9736052513122555,
      "epoch": 0.0325714606615376,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001445,
      "loss": 6.5896,
      "mean_token_accuracy": 0.1329849384725094,
      "num_tokens": 640914.0,
      "step": 290
    },
    {
      "entropy": 6.921159362792968,
      "epoch": 0.03313303756949514,
      "grad_norm": 1.140625,
      "learning_rate": 0.000147,
      "loss": 6.6129,
      "mean_token_accuracy": 0.12800714895129203,
      "num_tokens": 653057.0,
      "step": 295
    },
    {
      "entropy": 6.993884515762329,
      "epoch": 0.033694614477452686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001495,
      "loss": 6.5372,
      "mean_token_accuracy": 0.13587351217865945,
      "num_tokens": 663437.0,
      "step": 300
    },
    {
      "entropy": 6.979153728485107,
      "epoch": 0.03425619138541023,
      "grad_norm": 1.09375,
      "learning_rate": 0.000152,
      "loss": 6.6207,
      "mean_token_accuracy": 0.12917499542236327,
      "num_tokens": 675416.0,
      "step": 305
    },
    {
      "entropy": 6.811232805252075,
      "epoch": 0.03481776829336777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.4163,
      "mean_token_accuracy": 0.13696089312434195,
      "num_tokens": 686633.0,
      "step": 310
    },
    {
      "entropy": 6.947882556915284,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.0625,
      "learning_rate": 0.000157,
      "loss": 6.5374,
      "mean_token_accuracy": 0.13494077548384667,
      "num_tokens": 698065.0,
      "step": 315
    },
    {
      "entropy": 6.984415197372437,
      "epoch": 0.03594092210928287,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001595,
      "loss": 6.5488,
      "mean_token_accuracy": 0.13275664299726486,
      "num_tokens": 708652.0,
      "step": 320
    },
    {
      "entropy": 7.001538467407227,
      "epoch": 0.03650249901724041,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000162,
      "loss": 6.716,
      "mean_token_accuracy": 0.12674812003970146,
      "num_tokens": 720199.0,
      "step": 325
    },
    {
      "entropy": 7.04426908493042,
      "epoch": 0.037064075925197955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.6759,
      "mean_token_accuracy": 0.12783747613430024,
      "num_tokens": 732964.0,
      "step": 330
    },
    {
      "entropy": 6.893573808670044,
      "epoch": 0.0376256528331555,
      "grad_norm": 1.046875,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.4504,
      "mean_token_accuracy": 0.13295350223779678,
      "num_tokens": 743528.0,
      "step": 335
    },
    {
      "entropy": 6.9759503364562985,
      "epoch": 0.03818722974111304,
      "grad_norm": 1.125,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.5345,
      "mean_token_accuracy": 0.13175515085458755,
      "num_tokens": 754365.0,
      "step": 340
    },
    {
      "entropy": 6.8153159618377686,
      "epoch": 0.03874880664907059,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.4614,
      "mean_token_accuracy": 0.131683911383152,
      "num_tokens": 766094.0,
      "step": 345
    },
    {
      "entropy": 6.927649307250976,
      "epoch": 0.039310383557028136,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.5365,
      "mean_token_accuracy": 0.13060494735836983,
      "num_tokens": 777275.0,
      "step": 350
    },
    {
      "entropy": 6.932587432861328,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.078125,
      "learning_rate": 0.000177,
      "loss": 6.6145,
      "mean_token_accuracy": 0.13159823268651963,
      "num_tokens": 789436.0,
      "step": 355
    },
    {
      "entropy": 6.913347625732422,
      "epoch": 0.040433537372943223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001795,
      "loss": 6.6299,
      "mean_token_accuracy": 0.13135476037859917,
      "num_tokens": 800164.0,
      "step": 360
    },
    {
      "entropy": 6.842929649353027,
      "epoch": 0.04099511428090077,
      "grad_norm": 1.125,
      "learning_rate": 0.000182,
      "loss": 6.5077,
      "mean_token_accuracy": 0.13295300006866456,
      "num_tokens": 811797.0,
      "step": 365
    },
    {
      "entropy": 6.945403099060059,
      "epoch": 0.04155669118885831,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001845,
      "loss": 6.5181,
      "mean_token_accuracy": 0.13189431577920913,
      "num_tokens": 824158.0,
      "step": 370
    },
    {
      "entropy": 6.845687580108643,
      "epoch": 0.04211826809681586,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000187,
      "loss": 6.517,
      "mean_token_accuracy": 0.13648107424378395,
      "num_tokens": 835803.0,
      "step": 375
    },
    {
      "entropy": 6.995227909088134,
      "epoch": 0.042679845004773405,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001895,
      "loss": 6.6006,
      "mean_token_accuracy": 0.12846365720033645,
      "num_tokens": 848106.0,
      "step": 380
    },
    {
      "entropy": 6.88704137802124,
      "epoch": 0.04324142191273095,
      "grad_norm": 1.171875,
      "learning_rate": 0.000192,
      "loss": 6.5878,
      "mean_token_accuracy": 0.12994857951998712,
      "num_tokens": 859101.0,
      "step": 385
    },
    {
      "entropy": 6.8107438564300535,
      "epoch": 0.04380299882068849,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001945,
      "loss": 6.4562,
      "mean_token_accuracy": 0.1323493629693985,
      "num_tokens": 870309.0,
      "step": 390
    },
    {
      "entropy": 6.995388984680176,
      "epoch": 0.044364575728646036,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.5557,
      "mean_token_accuracy": 0.1317988380789757,
      "num_tokens": 881515.0,
      "step": 395
    },
    {
      "entropy": 6.84275221824646,
      "epoch": 0.044926152636603586,
      "grad_norm": 1.328125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.5152,
      "mean_token_accuracy": 0.13594210967421533,
      "num_tokens": 892129.0,
      "step": 400
    },
    {
      "entropy": 6.88900990486145,
      "epoch": 0.04548772954456113,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000202,
      "loss": 6.5348,
      "mean_token_accuracy": 0.1257748119533062,
      "num_tokens": 903186.0,
      "step": 405
    },
    {
      "entropy": 6.874643039703369,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.3125,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.5074,
      "mean_token_accuracy": 0.13315046355128288,
      "num_tokens": 913471.0,
      "step": 410
    },
    {
      "entropy": 6.908110284805298,
      "epoch": 0.04661088336047622,
      "grad_norm": 1.109375,
      "learning_rate": 0.000207,
      "loss": 6.5171,
      "mean_token_accuracy": 0.12603104412555693,
      "num_tokens": 924283.0,
      "step": 415
    },
    {
      "entropy": 6.892792367935181,
      "epoch": 0.04717246026843376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002095,
      "loss": 6.5449,
      "mean_token_accuracy": 0.13619182109832764,
      "num_tokens": 934975.0,
      "step": 420
    },
    {
      "entropy": 6.853797531127929,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.21875,
      "learning_rate": 0.000212,
      "loss": 6.4157,
      "mean_token_accuracy": 0.13784429281949998,
      "num_tokens": 945587.0,
      "step": 425
    },
    {
      "entropy": 6.832434177398682,
      "epoch": 0.048295614084348855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002145,
      "loss": 6.4868,
      "mean_token_accuracy": 0.12879071310162543,
      "num_tokens": 955935.0,
      "step": 430
    },
    {
      "entropy": 6.8071524143219,
      "epoch": 0.0488571909923064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.4581,
      "mean_token_accuracy": 0.13443006500601767,
      "num_tokens": 967960.0,
      "step": 435
    },
    {
      "entropy": 6.865250539779663,
      "epoch": 0.04941876790026394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002195,
      "loss": 6.509,
      "mean_token_accuracy": 0.1331017218530178,
      "num_tokens": 979680.0,
      "step": 440
    },
    {
      "entropy": 6.868223047256469,
      "epoch": 0.049980344808221486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000222,
      "loss": 6.5873,
      "mean_token_accuracy": 0.13184501379728317,
      "num_tokens": 990600.0,
      "step": 445
    },
    {
      "entropy": 6.990716028213501,
      "epoch": 0.05054192171617903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002245,
      "loss": 6.6286,
      "mean_token_accuracy": 0.1356397032737732,
      "num_tokens": 1002331.0,
      "step": 450
    },
    {
      "entropy": 6.83350019454956,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.265625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.4445,
      "mean_token_accuracy": 0.13236786276102067,
      "num_tokens": 1013473.0,
      "step": 455
    },
    {
      "entropy": 6.865316438674927,
      "epoch": 0.051665075532094124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.5207,
      "mean_token_accuracy": 0.13039045557379722,
      "num_tokens": 1024434.0,
      "step": 460
    },
    {
      "entropy": 6.890588569641113,
      "epoch": 0.05222665244005167,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.5385,
      "mean_token_accuracy": 0.12809760123491287,
      "num_tokens": 1035452.0,
      "step": 465
    },
    {
      "entropy": 6.826921796798706,
      "epoch": 0.05278822934800921,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.4174,
      "mean_token_accuracy": 0.13453132212162017,
      "num_tokens": 1046213.0,
      "step": 470
    },
    {
      "entropy": 6.771785497665405,
      "epoch": 0.053349806255966754,
      "grad_norm": 1.0625,
      "learning_rate": 0.000237,
      "loss": 6.4467,
      "mean_token_accuracy": 0.13567724153399469,
      "num_tokens": 1057469.0,
      "step": 475
    },
    {
      "entropy": 7.0101700782775875,
      "epoch": 0.0539113831639243,
      "grad_norm": 1.125,
      "learning_rate": 0.0002395,
      "loss": 6.5996,
      "mean_token_accuracy": 0.1290132947266102,
      "num_tokens": 1069543.0,
      "step": 480
    },
    {
      "entropy": 6.71439380645752,
      "epoch": 0.05447296007188184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000242,
      "loss": 6.4728,
      "mean_token_accuracy": 0.13269706442952156,
      "num_tokens": 1080476.0,
      "step": 485
    },
    {
      "entropy": 6.954592084884643,
      "epoch": 0.05503453697983939,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002445,
      "loss": 6.5828,
      "mean_token_accuracy": 0.1320900909602642,
      "num_tokens": 1091700.0,
      "step": 490
    },
    {
      "entropy": 6.786695575714111,
      "epoch": 0.055596113887796936,
      "grad_norm": 1.03125,
      "learning_rate": 0.000247,
      "loss": 6.4695,
      "mean_token_accuracy": 0.13233320266008378,
      "num_tokens": 1103056.0,
      "step": 495
    },
    {
      "entropy": 6.886286687850952,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002495,
      "loss": 6.5359,
      "mean_token_accuracy": 0.13645913898944856,
      "num_tokens": 1114836.0,
      "step": 500
    },
    {
      "entropy": 6.760018682479858,
      "epoch": 0.05671926770371202,
      "grad_norm": 1.140625,
      "learning_rate": 0.000252,
      "loss": 6.447,
      "mean_token_accuracy": 0.13287583217024804,
      "num_tokens": 1125993.0,
      "step": 505
    },
    {
      "entropy": 6.874892234802246,
      "epoch": 0.05728084461166957,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002545,
      "loss": 6.5176,
      "mean_token_accuracy": 0.13570363372564315,
      "num_tokens": 1136719.0,
      "step": 510
    },
    {
      "entropy": 6.795416641235351,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.234375,
      "learning_rate": 0.000257,
      "loss": 6.4023,
      "mean_token_accuracy": 0.13578868433833122,
      "num_tokens": 1146947.0,
      "step": 515
    },
    {
      "entropy": 6.826411437988281,
      "epoch": 0.05840399842758466,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002595,
      "loss": 6.5302,
      "mean_token_accuracy": 0.13053593188524246,
      "num_tokens": 1157974.0,
      "step": 520
    },
    {
      "entropy": 6.750731325149536,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000262,
      "loss": 6.4664,
      "mean_token_accuracy": 0.13042562529444696,
      "num_tokens": 1168573.0,
      "step": 525
    },
    {
      "entropy": 6.815242910385132,
      "epoch": 0.05952715224349975,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.3647,
      "mean_token_accuracy": 0.13693853691220284,
      "num_tokens": 1178446.0,
      "step": 530
    },
    {
      "entropy": 6.811131572723388,
      "epoch": 0.06008872915145729,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.3944,
      "mean_token_accuracy": 0.13583947718143463,
      "num_tokens": 1190558.0,
      "step": 535
    },
    {
      "entropy": 6.737180471420288,
      "epoch": 0.060650306059414835,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.5615,
      "mean_token_accuracy": 0.13333362191915513,
      "num_tokens": 1201929.0,
      "step": 540
    },
    {
      "entropy": 6.711785936355591,
      "epoch": 0.06121188296737238,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.407,
      "mean_token_accuracy": 0.13147006183862686,
      "num_tokens": 1213155.0,
      "step": 545
    },
    {
      "entropy": 6.913523435592651,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002745,
      "loss": 6.4767,
      "mean_token_accuracy": 0.13126003369688988,
      "num_tokens": 1223961.0,
      "step": 550
    },
    {
      "entropy": 6.638333320617676,
      "epoch": 0.06233503678328747,
      "grad_norm": 1.21875,
      "learning_rate": 0.000277,
      "loss": 6.3667,
      "mean_token_accuracy": 0.1402271084487438,
      "num_tokens": 1233888.0,
      "step": 555
    },
    {
      "entropy": 6.733293151855468,
      "epoch": 0.06289661369124501,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002795,
      "loss": 6.4953,
      "mean_token_accuracy": 0.12470907121896743,
      "num_tokens": 1244405.0,
      "step": 560
    },
    {
      "entropy": 6.823939466476441,
      "epoch": 0.06345819059920256,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.4405,
      "mean_token_accuracy": 0.13227839171886444,
      "num_tokens": 1257099.0,
      "step": 565
    },
    {
      "entropy": 6.827079486846924,
      "epoch": 0.06401976750716011,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002845,
      "loss": 6.5325,
      "mean_token_accuracy": 0.13312843218445777,
      "num_tokens": 1267965.0,
      "step": 570
    },
    {
      "entropy": 6.744953536987305,
      "epoch": 0.06458134441511765,
      "grad_norm": 1.1875,
      "learning_rate": 0.000287,
      "loss": 6.4552,
      "mean_token_accuracy": 0.13036884516477584,
      "num_tokens": 1280075.0,
      "step": 575
    },
    {
      "entropy": 6.849108505249023,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002895,
      "loss": 6.4498,
      "mean_token_accuracy": 0.13528711795806886,
      "num_tokens": 1290574.0,
      "step": 580
    },
    {
      "entropy": 6.740403699874878,
      "epoch": 0.06570449823103273,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000292,
      "loss": 6.431,
      "mean_token_accuracy": 0.141810342669487,
      "num_tokens": 1302680.0,
      "step": 585
    },
    {
      "entropy": 6.773650121688843,
      "epoch": 0.06626607513899029,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002945,
      "loss": 6.4158,
      "mean_token_accuracy": 0.13916678130626678,
      "num_tokens": 1314577.0,
      "step": 590
    },
    {
      "entropy": 6.780707979202271,
      "epoch": 0.06682765204694784,
      "grad_norm": 1.109375,
      "learning_rate": 0.000297,
      "loss": 6.5412,
      "mean_token_accuracy": 0.1369754306972027,
      "num_tokens": 1326560.0,
      "step": 595
    },
    {
      "entropy": 6.899291849136352,
      "epoch": 0.06738922895490537,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002995,
      "loss": 6.6101,
      "mean_token_accuracy": 0.128566275537014,
      "num_tokens": 1339105.0,
      "step": 600
    },
    {
      "entropy": 6.5432719707489015,
      "epoch": 0.06795080586286292,
      "grad_norm": 1.046875,
      "learning_rate": 0.000302,
      "loss": 6.2816,
      "mean_token_accuracy": 0.1405505046248436,
      "num_tokens": 1349914.0,
      "step": 605
    },
    {
      "entropy": 6.829094409942627,
      "epoch": 0.06851238277082046,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003045,
      "loss": 6.5035,
      "mean_token_accuracy": 0.1318673647940159,
      "num_tokens": 1361700.0,
      "step": 610
    },
    {
      "entropy": 6.7323582649230955,
      "epoch": 0.06907395967877801,
      "grad_norm": 1.046875,
      "learning_rate": 0.000307,
      "loss": 6.524,
      "mean_token_accuracy": 0.1338668704032898,
      "num_tokens": 1373103.0,
      "step": 615
    },
    {
      "entropy": 6.701350355148316,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003095,
      "loss": 6.3164,
      "mean_token_accuracy": 0.14012459367513658,
      "num_tokens": 1383742.0,
      "step": 620
    },
    {
      "entropy": 6.640402126312256,
      "epoch": 0.0701971134946931,
      "grad_norm": 0.96875,
      "learning_rate": 0.000312,
      "loss": 6.4778,
      "mean_token_accuracy": 0.1333733394742012,
      "num_tokens": 1396656.0,
      "step": 625
    },
    {
      "entropy": 6.783367872238159,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003145,
      "loss": 6.3468,
      "mean_token_accuracy": 0.15129355043172837,
      "num_tokens": 1406709.0,
      "step": 630
    },
    {
      "entropy": 6.7202239513397215,
      "epoch": 0.07132026731060818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000317,
      "loss": 6.3802,
      "mean_token_accuracy": 0.14140321388840676,
      "num_tokens": 1417793.0,
      "step": 635
    },
    {
      "entropy": 6.755014324188233,
      "epoch": 0.07188184421856574,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003195,
      "loss": 6.4691,
      "mean_token_accuracy": 0.13631825447082518,
      "num_tokens": 1429093.0,
      "step": 640
    },
    {
      "entropy": 6.677226877212524,
      "epoch": 0.07244342112652327,
      "grad_norm": 1.03125,
      "learning_rate": 0.000322,
      "loss": 6.3998,
      "mean_token_accuracy": 0.14246537908911705,
      "num_tokens": 1440907.0,
      "step": 645
    },
    {
      "entropy": 6.728086280822754,
      "epoch": 0.07300499803448082,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.4661,
      "mean_token_accuracy": 0.1364982731640339,
      "num_tokens": 1452089.0,
      "step": 650
    },
    {
      "entropy": 6.749984645843506,
      "epoch": 0.07356657494243837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.4943,
      "mean_token_accuracy": 0.1323855087161064,
      "num_tokens": 1464378.0,
      "step": 655
    },
    {
      "entropy": 6.694591379165649,
      "epoch": 0.07412815185039591,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.4131,
      "mean_token_accuracy": 0.13516984954476358,
      "num_tokens": 1476678.0,
      "step": 660
    },
    {
      "entropy": 6.660327816009522,
      "epoch": 0.07468972875835346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.2587,
      "mean_token_accuracy": 0.1437138944864273,
      "num_tokens": 1486808.0,
      "step": 665
    },
    {
      "entropy": 6.783201026916504,
      "epoch": 0.075251305666311,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.5481,
      "mean_token_accuracy": 0.12994321286678315,
      "num_tokens": 1499281.0,
      "step": 670
    },
    {
      "entropy": 6.711549758911133,
      "epoch": 0.07581288257426855,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000337,
      "loss": 6.4505,
      "mean_token_accuracy": 0.13625041544437408,
      "num_tokens": 1510202.0,
      "step": 675
    },
    {
      "entropy": 6.67443437576294,
      "epoch": 0.07637445948222608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003395,
      "loss": 6.3292,
      "mean_token_accuracy": 0.14246201664209365,
      "num_tokens": 1521492.0,
      "step": 680
    },
    {
      "entropy": 6.584014463424682,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.125,
      "learning_rate": 0.000342,
      "loss": 6.3419,
      "mean_token_accuracy": 0.14672905653715135,
      "num_tokens": 1532856.0,
      "step": 685
    },
    {
      "entropy": 6.675293254852295,
      "epoch": 0.07749761329814119,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.3414,
      "mean_token_accuracy": 0.13955900371074675,
      "num_tokens": 1543408.0,
      "step": 690
    },
    {
      "entropy": 6.606938409805298,
      "epoch": 0.07805919020609872,
      "grad_norm": 1.03125,
      "learning_rate": 0.000347,
      "loss": 6.3889,
      "mean_token_accuracy": 0.13311915546655656,
      "num_tokens": 1555474.0,
      "step": 695
    },
    {
      "entropy": 6.718470048904419,
      "epoch": 0.07862076711405627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003495,
      "loss": 6.4749,
      "mean_token_accuracy": 0.13796819001436234,
      "num_tokens": 1567484.0,
      "step": 700
    },
    {
      "entropy": 6.613565444946289,
      "epoch": 0.07918234402201381,
      "grad_norm": 1.0625,
      "learning_rate": 0.000352,
      "loss": 6.3032,
      "mean_token_accuracy": 0.1469588965177536,
      "num_tokens": 1578204.0,
      "step": 705
    },
    {
      "entropy": 6.695013618469238,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003545,
      "loss": 6.3049,
      "mean_token_accuracy": 0.13991812244057655,
      "num_tokens": 1589012.0,
      "step": 710
    },
    {
      "entropy": 6.560267734527588,
      "epoch": 0.08030549783792891,
      "grad_norm": 1.15625,
      "learning_rate": 0.000357,
      "loss": 6.3206,
      "mean_token_accuracy": 0.139244681596756,
      "num_tokens": 1601099.0,
      "step": 715
    },
    {
      "entropy": 6.70271053314209,
      "epoch": 0.08086707474588645,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003595,
      "loss": 6.4769,
      "mean_token_accuracy": 0.13117794394493104,
      "num_tokens": 1612814.0,
      "step": 720
    },
    {
      "entropy": 6.733196353912353,
      "epoch": 0.081428651653844,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000362,
      "loss": 6.3706,
      "mean_token_accuracy": 0.13757371306419372,
      "num_tokens": 1624953.0,
      "step": 725
    },
    {
      "entropy": 6.5949835777282715,
      "epoch": 0.08199022856180153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003645,
      "loss": 6.4122,
      "mean_token_accuracy": 0.1355846755206585,
      "num_tokens": 1635673.0,
      "step": 730
    },
    {
      "entropy": 6.634992599487305,
      "epoch": 0.08255180546975908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000367,
      "loss": 6.3149,
      "mean_token_accuracy": 0.14333953484892845,
      "num_tokens": 1647282.0,
      "step": 735
    },
    {
      "entropy": 6.590029048919678,
      "epoch": 0.08311338237771662,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003695,
      "loss": 6.3328,
      "mean_token_accuracy": 0.1433751843869686,
      "num_tokens": 1658426.0,
      "step": 740
    },
    {
      "entropy": 6.614629554748535,
      "epoch": 0.08367495928567417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000372,
      "loss": 6.3629,
      "mean_token_accuracy": 0.1389339104294777,
      "num_tokens": 1669958.0,
      "step": 745
    },
    {
      "entropy": 6.658067321777343,
      "epoch": 0.08423653619363172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003745,
      "loss": 6.4213,
      "mean_token_accuracy": 0.13475271239876746,
      "num_tokens": 1682409.0,
      "step": 750
    },
    {
      "entropy": 6.597114753723145,
      "epoch": 0.08479811310158926,
      "grad_norm": 1.078125,
      "learning_rate": 0.000377,
      "loss": 6.3431,
      "mean_token_accuracy": 0.1512242317199707,
      "num_tokens": 1692409.0,
      "step": 755
    },
    {
      "entropy": 6.711719036102295,
      "epoch": 0.08535969000954681,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003795,
      "loss": 6.4225,
      "mean_token_accuracy": 0.13909337669610977,
      "num_tokens": 1705018.0,
      "step": 760
    },
    {
      "entropy": 6.600584983825684,
      "epoch": 0.08592126691750435,
      "grad_norm": 1.09375,
      "learning_rate": 0.000382,
      "loss": 6.3506,
      "mean_token_accuracy": 0.1366318315267563,
      "num_tokens": 1715625.0,
      "step": 765
    },
    {
      "entropy": 6.598914957046508,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003845,
      "loss": 6.4358,
      "mean_token_accuracy": 0.13147977441549302,
      "num_tokens": 1727045.0,
      "step": 770
    },
    {
      "entropy": 6.7401669979095455,
      "epoch": 0.08704442073341945,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5042,
      "mean_token_accuracy": 0.13844706639647483,
      "num_tokens": 1737374.0,
      "step": 775
    },
    {
      "entropy": 6.725053167343139,
      "epoch": 0.08760599764137698,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.484,
      "mean_token_accuracy": 0.1349758803844452,
      "num_tokens": 1747788.0,
      "step": 780
    },
    {
      "entropy": 6.476839303970337,
      "epoch": 0.08816757454933453,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.1905,
      "mean_token_accuracy": 0.14935870766639708,
      "num_tokens": 1757769.0,
      "step": 785
    },
    {
      "entropy": 6.5435089588165285,
      "epoch": 0.08872915145729207,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.3098,
      "mean_token_accuracy": 0.13398686051368713,
      "num_tokens": 1769246.0,
      "step": 790
    },
    {
      "entropy": 6.666951322555542,
      "epoch": 0.08929072836524962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4301,
      "mean_token_accuracy": 0.13988927975296975,
      "num_tokens": 1781306.0,
      "step": 795
    },
    {
      "entropy": 6.621778106689453,
      "epoch": 0.08985230527320717,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003995,
      "loss": 6.3204,
      "mean_token_accuracy": 0.14044865369796752,
      "num_tokens": 1792888.0,
      "step": 800
    },
    {
      "entropy": 6.572351598739624,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.15625,
      "learning_rate": 0.000402,
      "loss": 6.3445,
      "mean_token_accuracy": 0.146021768450737,
      "num_tokens": 1802938.0,
      "step": 805
    },
    {
      "entropy": 6.656881046295166,
      "epoch": 0.09097545908912226,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004045,
      "loss": 6.3536,
      "mean_token_accuracy": 0.13862004801630973,
      "num_tokens": 1814738.0,
      "step": 810
    },
    {
      "entropy": 6.619713163375854,
      "epoch": 0.0915370359970798,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4315,
      "mean_token_accuracy": 0.14092215448617934,
      "num_tokens": 1825100.0,
      "step": 815
    },
    {
      "entropy": 6.440285301208496,
      "epoch": 0.09209861290503735,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004095,
      "loss": 6.2709,
      "mean_token_accuracy": 0.14654000774025916,
      "num_tokens": 1835879.0,
      "step": 820
    },
    {
      "entropy": 6.788227272033692,
      "epoch": 0.09266018981299488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412,
      "loss": 6.4236,
      "mean_token_accuracy": 0.13221182823181152,
      "num_tokens": 1848352.0,
      "step": 825
    },
    {
      "entropy": 6.550888586044311,
      "epoch": 0.09322176672095243,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145,
      "loss": 6.3654,
      "mean_token_accuracy": 0.13856692165136336,
      "num_tokens": 1859687.0,
      "step": 830
    },
    {
      "entropy": 6.565574073791504,
      "epoch": 0.09378334362890998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000417,
      "loss": 6.3691,
      "mean_token_accuracy": 0.14064017236232756,
      "num_tokens": 1871438.0,
      "step": 835
    },
    {
      "entropy": 6.58135290145874,
      "epoch": 0.09434492053686752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004195,
      "loss": 6.25,
      "mean_token_accuracy": 0.1493062250316143,
      "num_tokens": 1881519.0,
      "step": 840
    },
    {
      "entropy": 6.656439733505249,
      "epoch": 0.09490649744482507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000422,
      "loss": 6.4275,
      "mean_token_accuracy": 0.13461267650127412,
      "num_tokens": 1892428.0,
      "step": 845
    },
    {
      "entropy": 6.692960023880005,
      "epoch": 0.09546807435278261,
      "grad_norm": 1.125,
      "learning_rate": 0.0004245,
      "loss": 6.3208,
      "mean_token_accuracy": 0.1437561698257923,
      "num_tokens": 1903760.0,
      "step": 850
    },
    {
      "entropy": 6.522662353515625,
      "epoch": 0.09602965126074016,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000427,
      "loss": 6.3783,
      "mean_token_accuracy": 0.13410450741648675,
      "num_tokens": 1915430.0,
      "step": 855
    },
    {
      "entropy": 6.594115972518921,
      "epoch": 0.09659122816869771,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004295,
      "loss": 6.3324,
      "mean_token_accuracy": 0.1378035582602024,
      "num_tokens": 1925709.0,
      "step": 860
    },
    {
      "entropy": 6.645199775695801,
      "epoch": 0.09715280507665525,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000432,
      "loss": 6.3761,
      "mean_token_accuracy": 0.1379880949854851,
      "num_tokens": 1938148.0,
      "step": 865
    },
    {
      "entropy": 6.36707878112793,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345,
      "loss": 6.2556,
      "mean_token_accuracy": 0.14044615030288696,
      "num_tokens": 1949028.0,
      "step": 870
    },
    {
      "entropy": 6.615007495880127,
      "epoch": 0.09827595889257033,
      "grad_norm": 0.984375,
      "learning_rate": 0.000437,
      "loss": 6.3155,
      "mean_token_accuracy": 0.14169368892908096,
      "num_tokens": 1960353.0,
      "step": 875
    },
    {
      "entropy": 6.591008424758911,
      "epoch": 0.09883753580052788,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004395,
      "loss": 6.4238,
      "mean_token_accuracy": 0.13767419308423995,
      "num_tokens": 1972335.0,
      "step": 880
    },
    {
      "entropy": 6.504133701324463,
      "epoch": 0.09939911270848542,
      "grad_norm": 1.0625,
      "learning_rate": 0.000442,
      "loss": 6.2723,
      "mean_token_accuracy": 0.14177839010953902,
      "num_tokens": 1982851.0,
      "step": 885
    },
    {
      "entropy": 6.495877361297607,
      "epoch": 0.09996068961644297,
      "grad_norm": 1.0,
      "learning_rate": 0.0004445,
      "loss": 6.2433,
      "mean_token_accuracy": 0.14337393641471863,
      "num_tokens": 1993931.0,
      "step": 890
    },
    {
      "entropy": 6.582594537734986,
      "epoch": 0.10052226652440052,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000447,
      "loss": 6.381,
      "mean_token_accuracy": 0.1384746566414833,
      "num_tokens": 2004838.0,
      "step": 895
    },
    {
      "entropy": 6.437647104263306,
      "epoch": 0.10108384343235806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.1842,
      "mean_token_accuracy": 0.14243700206279755,
      "num_tokens": 2015620.0,
      "step": 900
    },
    {
      "entropy": 6.569198560714722,
      "epoch": 0.10164542034031561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3106,
      "mean_token_accuracy": 0.14497067034244537,
      "num_tokens": 2027489.0,
      "step": 905
    },
    {
      "entropy": 6.397192287445068,
      "epoch": 0.10220699724827315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.194,
      "mean_token_accuracy": 0.14697765707969665,
      "num_tokens": 2038475.0,
      "step": 910
    },
    {
      "entropy": 6.452962160110474,
      "epoch": 0.1027685741562307,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.1966,
      "mean_token_accuracy": 0.14352490305900573,
      "num_tokens": 2048938.0,
      "step": 915
    },
    {
      "entropy": 6.477091646194458,
      "epoch": 0.10333015106418825,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2972,
      "mean_token_accuracy": 0.13728735148906707,
      "num_tokens": 2060372.0,
      "step": 920
    },
    {
      "entropy": 6.470232009887695,
      "epoch": 0.10389172797214578,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462,
      "loss": 6.2836,
      "mean_token_accuracy": 0.13895049095153808,
      "num_tokens": 2071558.0,
      "step": 925
    },
    {
      "entropy": 6.541862058639526,
      "epoch": 0.10445330488010333,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004645,
      "loss": 6.3173,
      "mean_token_accuracy": 0.13610014989972113,
      "num_tokens": 2082350.0,
      "step": 930
    },
    {
      "entropy": 6.466197681427002,
      "epoch": 0.10501488178806087,
      "grad_norm": 1.015625,
      "learning_rate": 0.000467,
      "loss": 6.37,
      "mean_token_accuracy": 0.14044060856103896,
      "num_tokens": 2093448.0,
      "step": 935
    },
    {
      "entropy": 6.683929204940796,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004695,
      "loss": 6.3979,
      "mean_token_accuracy": 0.13594734519720078,
      "num_tokens": 2104171.0,
      "step": 940
    },
    {
      "entropy": 6.433077430725097,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.15625,
      "learning_rate": 0.000472,
      "loss": 6.2868,
      "mean_token_accuracy": 0.1391531579196453,
      "num_tokens": 2114496.0,
      "step": 945
    },
    {
      "entropy": 6.5762317180633545,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004745,
      "loss": 6.2747,
      "mean_token_accuracy": 0.14130542203783988,
      "num_tokens": 2125930.0,
      "step": 950
    },
    {
      "entropy": 6.492092609405518,
      "epoch": 0.10726118941989106,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477,
      "loss": 6.2387,
      "mean_token_accuracy": 0.14300135299563407,
      "num_tokens": 2136452.0,
      "step": 955
    },
    {
      "entropy": 6.472325229644776,
      "epoch": 0.1078227663278486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004795,
      "loss": 6.3419,
      "mean_token_accuracy": 0.13892039209604262,
      "num_tokens": 2148233.0,
      "step": 960
    },
    {
      "entropy": 6.37879753112793,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000482,
      "loss": 6.1324,
      "mean_token_accuracy": 0.14699541106820108,
      "num_tokens": 2159475.0,
      "step": 965
    },
    {
      "entropy": 6.397939586639405,
      "epoch": 0.10894592014376368,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004845,
      "loss": 6.2593,
      "mean_token_accuracy": 0.14171308875083924,
      "num_tokens": 2170042.0,
      "step": 970
    },
    {
      "entropy": 6.4696125984191895,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000487,
      "loss": 6.2453,
      "mean_token_accuracy": 0.13840788900852202,
      "num_tokens": 2182234.0,
      "step": 975
    },
    {
      "entropy": 6.4058380126953125,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004895,
      "loss": 6.2491,
      "mean_token_accuracy": 0.14312953129410744,
      "num_tokens": 2192477.0,
      "step": 980
    },
    {
      "entropy": 6.388981103897095,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.25,
      "learning_rate": 0.000492,
      "loss": 6.2388,
      "mean_token_accuracy": 0.14177233427762986,
      "num_tokens": 2203082.0,
      "step": 985
    },
    {
      "entropy": 6.4049492359161375,
      "epoch": 0.11119222777559387,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004945,
      "loss": 6.2167,
      "mean_token_accuracy": 0.14464509785175322,
      "num_tokens": 2213684.0,
      "step": 990
    },
    {
      "entropy": 6.415097045898437,
      "epoch": 0.11175380468355141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497,
      "loss": 6.2157,
      "mean_token_accuracy": 0.1419524908065796,
      "num_tokens": 2225524.0,
      "step": 995
    },
    {
      "entropy": 6.487058210372925,
      "epoch": 0.11231538159150896,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995,
      "loss": 6.1869,
      "mean_token_accuracy": 0.1425313711166382,
      "num_tokens": 2236004.0,
      "step": 1000
    },
    {
      "entropy": 6.465081977844238,
      "epoch": 0.1128769584994665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499999997707492,
      "loss": 6.3333,
      "mean_token_accuracy": 0.13484762981534004,
      "num_tokens": 2247730.0,
      "step": 1005
    },
    {
      "entropy": 6.48140435218811,
      "epoch": 0.11343853540742405,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.2913,
      "mean_token_accuracy": 0.14429717510938644,
      "num_tokens": 2260132.0,
      "step": 1010
    },
    {
      "entropy": 6.403030157089233,
      "epoch": 0.1140001123153816,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.2629,
      "mean_token_accuracy": 0.1453886643052101,
      "num_tokens": 2270060.0,
      "step": 1015
    },
    {
      "entropy": 6.4181787967681885,
      "epoch": 0.11456168922333913,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.2805,
      "mean_token_accuracy": 0.14259432330727578,
      "num_tokens": 2281568.0,
      "step": 1020
    },
    {
      "entropy": 6.476824760437012,
      "epoch": 0.11512326613129668,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.2901,
      "mean_token_accuracy": 0.13487386032938958,
      "num_tokens": 2293325.0,
      "step": 1025
    },
    {
      "entropy": 6.481242752075195,
      "epoch": 0.11568484303925422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.293,
      "mean_token_accuracy": 0.14598456174135208,
      "num_tokens": 2304780.0,
      "step": 1030
    },
    {
      "entropy": 6.406403398513794,
      "epoch": 0.11624641994721177,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.2363,
      "mean_token_accuracy": 0.13638271912932395,
      "num_tokens": 2315999.0,
      "step": 1035
    },
    {
      "entropy": 6.472362041473389,
      "epoch": 0.11680799685516932,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999997820685,
      "loss": 6.3868,
      "mean_token_accuracy": 0.14324138388037683,
      "num_tokens": 2326695.0,
      "step": 1040
    },
    {
      "entropy": 6.4760270595550535,
      "epoch": 0.11736957376312686,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.2064,
      "mean_token_accuracy": 0.15152312070131302,
      "num_tokens": 2336925.0,
      "step": 1045
    },
    {
      "entropy": 6.393593263626099,
      "epoch": 0.11793115067108441,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.1246,
      "mean_token_accuracy": 0.1490764707326889,
      "num_tokens": 2347431.0,
      "step": 1050
    },
    {
      "entropy": 6.360946941375732,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.2297,
      "mean_token_accuracy": 0.14785884395241738,
      "num_tokens": 2359894.0,
      "step": 1055
    },
    {
      "entropy": 6.376661109924316,
      "epoch": 0.1190543044869995,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.2571,
      "mean_token_accuracy": 0.14182159453630447,
      "num_tokens": 2371181.0,
      "step": 1060
    },
    {
      "entropy": 6.43308916091919,
      "epoch": 0.11961588139495703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.1609,
      "mean_token_accuracy": 0.14217472076416016,
      "num_tokens": 2381531.0,
      "step": 1065
    },
    {
      "entropy": 6.453187465667725,
      "epoch": 0.12017745830291458,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.2808,
      "mean_token_accuracy": 0.1432340309023857,
      "num_tokens": 2391573.0,
      "step": 1070
    },
    {
      "entropy": 6.3629372119903564,
      "epoch": 0.12073903521087213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.2391,
      "mean_token_accuracy": 0.14391963332891464,
      "num_tokens": 2402822.0,
      "step": 1075
    },
    {
      "entropy": 6.405177211761474,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.1633,
      "mean_token_accuracy": 0.14490702822804452,
      "num_tokens": 2414322.0,
      "step": 1080
    },
    {
      "entropy": 6.475324583053589,
      "epoch": 0.12186218902678722,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.347,
      "mean_token_accuracy": 0.1440906621515751,
      "num_tokens": 2426560.0,
      "step": 1085
    },
    {
      "entropy": 6.456641006469726,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.2586,
      "mean_token_accuracy": 0.1443517878651619,
      "num_tokens": 2437723.0,
      "step": 1090
    },
    {
      "entropy": 6.335519123077392,
      "epoch": 0.12298534284270231,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.2082,
      "mean_token_accuracy": 0.14743260070681571,
      "num_tokens": 2448497.0,
      "step": 1095
    },
    {
      "entropy": 6.297663021087646,
      "epoch": 0.12354691975065986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.0868,
      "mean_token_accuracy": 0.15371182709932327,
      "num_tokens": 2459089.0,
      "step": 1100
    },
    {
      "entropy": 6.3653172016143795,
      "epoch": 0.1241084966586174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.1582,
      "mean_token_accuracy": 0.15073663890361785,
      "num_tokens": 2469882.0,
      "step": 1105
    },
    {
      "entropy": 6.339643955230713,
      "epoch": 0.12467007356657495,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.1994,
      "mean_token_accuracy": 0.14943020194768905,
      "num_tokens": 2482120.0,
      "step": 1110
    },
    {
      "entropy": 6.432038021087647,
      "epoch": 0.1252316504745325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.215,
      "mean_token_accuracy": 0.1437155120074749,
      "num_tokens": 2492626.0,
      "step": 1115
    },
    {
      "entropy": 6.460601425170898,
      "epoch": 0.12579322738249002,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499997970990249,
      "loss": 6.2638,
      "mean_token_accuracy": 0.14326856955885886,
      "num_tokens": 2506347.0,
      "step": 1120
    },
    {
      "entropy": 6.244403314590454,
      "epoch": 0.12635480429044757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.1527,
      "mean_token_accuracy": 0.15141456723213195,
      "num_tokens": 2516584.0,
      "step": 1125
    },
    {
      "entropy": 6.383342456817627,
      "epoch": 0.12691638119840512,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.1966,
      "mean_token_accuracy": 0.1493828549981117,
      "num_tokens": 2526731.0,
      "step": 1130
    },
    {
      "entropy": 6.324274206161499,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.0998,
      "mean_token_accuracy": 0.15336599498987197,
      "num_tokens": 2537180.0,
      "step": 1135
    },
    {
      "entropy": 6.393867540359497,
      "epoch": 0.12803953501432022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.1904,
      "mean_token_accuracy": 0.14489513039588928,
      "num_tokens": 2547617.0,
      "step": 1140
    },
    {
      "entropy": 6.284448671340942,
      "epoch": 0.12860111192227774,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.0558,
      "mean_token_accuracy": 0.15631555318832396,
      "num_tokens": 2557743.0,
      "step": 1145
    },
    {
      "entropy": 6.3023943424224855,
      "epoch": 0.1291626888302353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.1694,
      "mean_token_accuracy": 0.14730207920074462,
      "num_tokens": 2568351.0,
      "step": 1150
    },
    {
      "entropy": 6.266475439071655,
      "epoch": 0.12972426573819285,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.2243,
      "mean_token_accuracy": 0.1486775517463684,
      "num_tokens": 2579703.0,
      "step": 1155
    },
    {
      "entropy": 6.48288254737854,
      "epoch": 0.1302858426461504,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.202,
      "mean_token_accuracy": 0.1458014190196991,
      "num_tokens": 2591197.0,
      "step": 1160
    },
    {
      "entropy": 6.293277406692505,
      "epoch": 0.13084741955410795,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499996146305169,
      "loss": 6.1625,
      "mean_token_accuracy": 0.14280997961759567,
      "num_tokens": 2601836.0,
      "step": 1165
    },
    {
      "entropy": 6.247816705703736,
      "epoch": 0.13140899646206547,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.1171,
      "mean_token_accuracy": 0.14851301163434982,
      "num_tokens": 2613513.0,
      "step": 1170
    },
    {
      "entropy": 6.337481117248535,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.1561,
      "mean_token_accuracy": 0.14839788228273393,
      "num_tokens": 2625246.0,
      "step": 1175
    },
    {
      "entropy": 6.3013852596282955,
      "epoch": 0.13253215027798057,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.1219,
      "mean_token_accuracy": 0.1513855129480362,
      "num_tokens": 2636375.0,
      "step": 1180
    },
    {
      "entropy": 6.3086591243743895,
      "epoch": 0.13309372718593812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.211,
      "mean_token_accuracy": 0.14393428042531015,
      "num_tokens": 2648586.0,
      "step": 1185
    },
    {
      "entropy": 6.401763534545898,
      "epoch": 0.13365530409389567,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.2164,
      "mean_token_accuracy": 0.15318685173988342,
      "num_tokens": 2660847.0,
      "step": 1190
    },
    {
      "entropy": 6.29134726524353,
      "epoch": 0.1342168810018532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049999460746976,
      "loss": 6.1473,
      "mean_token_accuracy": 0.14875885099172592,
      "num_tokens": 2671160.0,
      "step": 1195
    },
    {
      "entropy": 6.482362604141235,
      "epoch": 0.13477845790981074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.2605,
      "mean_token_accuracy": 0.144464373588562,
      "num_tokens": 2681622.0,
      "step": 1200
    },
    {
      "entropy": 6.207872486114502,
      "epoch": 0.1353400348177683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.0459,
      "mean_token_accuracy": 0.15137153267860412,
      "num_tokens": 2691503.0,
      "step": 1205
    },
    {
      "entropy": 6.269325542449951,
      "epoch": 0.13590161172572585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.1656,
      "mean_token_accuracy": 0.14382996857166291,
      "num_tokens": 2702207.0,
      "step": 1210
    },
    {
      "entropy": 6.334103918075561,
      "epoch": 0.1364631886336834,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.155,
      "mean_token_accuracy": 0.1459783896803856,
      "num_tokens": 2713178.0,
      "step": 1215
    },
    {
      "entropy": 6.360417366027832,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.1601,
      "mean_token_accuracy": 0.15585049539804458,
      "num_tokens": 2723717.0,
      "step": 1220
    },
    {
      "entropy": 6.296167993545533,
      "epoch": 0.13758634244959847,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.2208,
      "mean_token_accuracy": 0.142020982503891,
      "num_tokens": 2735669.0,
      "step": 1225
    },
    {
      "entropy": 6.2710144996643065,
      "epoch": 0.13814791935755602,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.1297,
      "mean_token_accuracy": 0.14939224869012832,
      "num_tokens": 2746087.0,
      "step": 1230
    },
    {
      "entropy": 6.290295934677124,
      "epoch": 0.13870949626551357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.143,
      "mean_token_accuracy": 0.1474170580506325,
      "num_tokens": 2757143.0,
      "step": 1235
    },
    {
      "entropy": 6.402835416793823,
      "epoch": 0.1392710731734711,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.2416,
      "mean_token_accuracy": 0.14436521381139755,
      "num_tokens": 2769340.0,
      "step": 1240
    },
    {
      "entropy": 6.353951168060303,
      "epoch": 0.13983265008142864,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.1601,
      "mean_token_accuracy": 0.14525908827781678,
      "num_tokens": 2779514.0,
      "step": 1245
    },
    {
      "entropy": 6.312722253799438,
      "epoch": 0.1403942269893862,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.0998,
      "mean_token_accuracy": 0.1497655987739563,
      "num_tokens": 2790801.0,
      "step": 1250
    },
    {
      "entropy": 6.249200296401978,
      "epoch": 0.14095580389734375,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.0947,
      "mean_token_accuracy": 0.1503822013735771,
      "num_tokens": 2801082.0,
      "step": 1255
    },
    {
      "entropy": 6.284300231933594,
      "epoch": 0.1415173808053013,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.1092,
      "mean_token_accuracy": 0.14815400838851928,
      "num_tokens": 2812719.0,
      "step": 1260
    },
    {
      "entropy": 6.227870082855224,
      "epoch": 0.14207895771325882,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499990013909322,
      "loss": 6.1087,
      "mean_token_accuracy": 0.14787399768829346,
      "num_tokens": 2823766.0,
      "step": 1265
    },
    {
      "entropy": 6.272064590454102,
      "epoch": 0.14264053462121637,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499989632069227,
      "loss": 6.1973,
      "mean_token_accuracy": 0.14443453252315522,
      "num_tokens": 2834801.0,
      "step": 1270
    },
    {
      "entropy": 6.274245309829712,
      "epoch": 0.14320211152917392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.0189,
      "mean_token_accuracy": 0.1519506961107254,
      "num_tokens": 2845769.0,
      "step": 1275
    },
    {
      "entropy": 6.176938724517822,
      "epoch": 0.14376368843713147,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.0579,
      "mean_token_accuracy": 0.1528223842382431,
      "num_tokens": 2856001.0,
      "step": 1280
    },
    {
      "entropy": 6.2550255298614506,
      "epoch": 0.14432526534508902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.0874,
      "mean_token_accuracy": 0.15212968960404397,
      "num_tokens": 2867703.0,
      "step": 1285
    },
    {
      "entropy": 6.179595184326172,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999880330714,
      "loss": 6.0396,
      "mean_token_accuracy": 0.15258833020925522,
      "num_tokens": 2878671.0,
      "step": 1290
    },
    {
      "entropy": 6.2422667980194095,
      "epoch": 0.1454484191610041,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.1536,
      "mean_token_accuracy": 0.1454537458717823,
      "num_tokens": 2890137.0,
      "step": 1295
    },
    {
      "entropy": 6.2193653106689455,
      "epoch": 0.14600999606896164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.0568,
      "mean_token_accuracy": 0.15350484400987624,
      "num_tokens": 2900344.0,
      "step": 1300
    },
    {
      "entropy": 6.197004318237305,
      "epoch": 0.1465715729769192,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.0832,
      "mean_token_accuracy": 0.1501095861196518,
      "num_tokens": 2912270.0,
      "step": 1305
    },
    {
      "entropy": 6.325214147567749,
      "epoch": 0.14713314988487675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.1474,
      "mean_token_accuracy": 0.142381751537323,
      "num_tokens": 2924677.0,
      "step": 1310
    },
    {
      "entropy": 6.198146820068359,
      "epoch": 0.14769472679283427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.0754,
      "mean_token_accuracy": 0.1535286396741867,
      "num_tokens": 2935794.0,
      "step": 1315
    },
    {
      "entropy": 6.309402561187744,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.1807,
      "mean_token_accuracy": 0.1483509808778763,
      "num_tokens": 2945958.0,
      "step": 1320
    },
    {
      "entropy": 6.274414014816284,
      "epoch": 0.14881788060874937,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.1433,
      "mean_token_accuracy": 0.14606737792491914,
      "num_tokens": 2957536.0,
      "step": 1325
    },
    {
      "entropy": 6.277726364135742,
      "epoch": 0.14937945751670692,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.1094,
      "mean_token_accuracy": 0.15047910585999488,
      "num_tokens": 2967787.0,
      "step": 1330
    },
    {
      "entropy": 6.2430198192596436,
      "epoch": 0.14994103442466447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499984016250827,
      "loss": 6.1327,
      "mean_token_accuracy": 0.14211324453353882,
      "num_tokens": 2980522.0,
      "step": 1335
    },
    {
      "entropy": 6.135755300521851,
      "epoch": 0.150502611332622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499983534119342,
      "loss": 6.0791,
      "mean_token_accuracy": 0.15327682048082353,
      "num_tokens": 2990497.0,
      "step": 1340
    },
    {
      "entropy": 6.3184624195098875,
      "epoch": 0.15106418824057954,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.1647,
      "mean_token_accuracy": 0.15013931393623353,
      "num_tokens": 3002931.0,
      "step": 1345
    },
    {
      "entropy": 6.258245277404785,
      "epoch": 0.1516257651485371,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.0849,
      "mean_token_accuracy": 0.1547642767429352,
      "num_tokens": 3014137.0,
      "step": 1350
    },
    {
      "entropy": 6.369079113006592,
      "epoch": 0.15218734205649465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.2573,
      "mean_token_accuracy": 0.13950591012835503,
      "num_tokens": 3026145.0,
      "step": 1355
    },
    {
      "entropy": 6.16160569190979,
      "epoch": 0.15274891896445217,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.1087,
      "mean_token_accuracy": 0.14567982107400895,
      "num_tokens": 3037068.0,
      "step": 1360
    },
    {
      "entropy": 6.293409633636474,
      "epoch": 0.15331049587240972,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.0279,
      "mean_token_accuracy": 0.14883914291858674,
      "num_tokens": 3047836.0,
      "step": 1365
    },
    {
      "entropy": 6.2063792705535885,
      "epoch": 0.15387207278036727,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.1668,
      "mean_token_accuracy": 0.15080713033676146,
      "num_tokens": 3060309.0,
      "step": 1370
    },
    {
      "entropy": 6.290639400482178,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.1341,
      "mean_token_accuracy": 0.1524991884827614,
      "num_tokens": 3071789.0,
      "step": 1375
    },
    {
      "entropy": 6.208878803253174,
      "epoch": 0.15499522659628237,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499979419180532,
      "loss": 6.0495,
      "mean_token_accuracy": 0.15470616593956948,
      "num_tokens": 3081756.0,
      "step": 1380
    },
    {
      "entropy": 6.266907453536987,
      "epoch": 0.1555568035042399,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.1499,
      "mean_token_accuracy": 0.14703571647405625,
      "num_tokens": 3093456.0,
      "step": 1385
    },
    {
      "entropy": 6.318468141555786,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.0872,
      "mean_token_accuracy": 0.15315306335687637,
      "num_tokens": 3104565.0,
      "step": 1390
    },
    {
      "entropy": 6.206810712814331,
      "epoch": 0.156679957320155,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999777578813904,
      "loss": 6.1015,
      "mean_token_accuracy": 0.14678023308515548,
      "num_tokens": 3115577.0,
      "step": 1395
    },
    {
      "entropy": 6.182242918014526,
      "epoch": 0.15724153422811254,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499977189788228,
      "loss": 6.0038,
      "mean_token_accuracy": 0.1520043984055519,
      "num_tokens": 3126019.0,
      "step": 1400
    },
    {
      "entropy": 6.278607749938965,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.1645,
      "mean_token_accuracy": 0.13868181630969048,
      "num_tokens": 3136191.0,
      "step": 1405
    },
    {
      "entropy": 6.29112024307251,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.0267,
      "mean_token_accuracy": 0.1518556460738182,
      "num_tokens": 3147737.0,
      "step": 1410
    },
    {
      "entropy": 6.1147044658660885,
      "epoch": 0.15892626495198517,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.1018,
      "mean_token_accuracy": 0.14828776121139525,
      "num_tokens": 3158494.0,
      "step": 1415
    },
    {
      "entropy": 6.346787929534912,
      "epoch": 0.15948784185994272,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.1186,
      "mean_token_accuracy": 0.15080327838659285,
      "num_tokens": 3170550.0,
      "step": 1420
    },
    {
      "entropy": 6.109242296218872,
      "epoch": 0.16004941876790027,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499974241872371,
      "loss": 6.1044,
      "mean_token_accuracy": 0.15341940075159072,
      "num_tokens": 3182860.0,
      "step": 1425
    },
    {
      "entropy": 6.2393961429595945,
      "epoch": 0.16061099567585782,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.0998,
      "mean_token_accuracy": 0.1468542292714119,
      "num_tokens": 3193807.0,
      "step": 1430
    },
    {
      "entropy": 6.112961626052856,
      "epoch": 0.16117257258381534,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999730125630255,
      "loss": 5.9753,
      "mean_token_accuracy": 0.14808735698461534,
      "num_tokens": 3203959.0,
      "step": 1435
    },
    {
      "entropy": 6.200899696350097,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999723871635008,
      "loss": 5.9863,
      "mean_token_accuracy": 0.16002811044454573,
      "num_tokens": 3215416.0,
      "step": 1440
    },
    {
      "entropy": 6.27506537437439,
      "epoch": 0.16229572639973044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.1366,
      "mean_token_accuracy": 0.14617353230714797,
      "num_tokens": 3226938.0,
      "step": 1445
    },
    {
      "entropy": 6.093537378311157,
      "epoch": 0.162857303307688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999711148748476,
      "loss": 5.9578,
      "mean_token_accuracy": 0.15727282613515853,
      "num_tokens": 3238406.0,
      "step": 1450
    },
    {
      "entropy": 6.092117643356323,
      "epoch": 0.16341888021564555,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999704679857594,
      "loss": 5.9807,
      "mean_token_accuracy": 0.15305008739233017,
      "num_tokens": 3249695.0,
      "step": 1455
    },
    {
      "entropy": 6.300736570358277,
      "epoch": 0.16398045712360307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.1224,
      "mean_token_accuracy": 0.15480272024869918,
      "num_tokens": 3259656.0,
      "step": 1460
    },
    {
      "entropy": 6.161792993545532,
      "epoch": 0.16454203403156062,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999691527181632,
      "loss": 5.9581,
      "mean_token_accuracy": 0.1575998604297638,
      "num_tokens": 3269712.0,
      "step": 1465
    },
    {
      "entropy": 6.017685747146606,
      "epoch": 0.16510361093951817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999684843396968,
      "loss": 5.9473,
      "mean_token_accuracy": 0.1620100051164627,
      "num_tokens": 3280343.0,
      "step": 1470
    },
    {
      "entropy": 6.132945108413696,
      "epoch": 0.16566518784747572,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999678087981467,
      "loss": 5.9832,
      "mean_token_accuracy": 0.15548000037670134,
      "num_tokens": 3291649.0,
      "step": 1475
    },
    {
      "entropy": 6.260512495040894,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.1602,
      "mean_token_accuracy": 0.15033089965581894,
      "num_tokens": 3304190.0,
      "step": 1480
    },
    {
      "entropy": 6.118093347549438,
      "epoch": 0.1667883416633908,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999664362258814,
      "loss": 5.8834,
      "mean_token_accuracy": 0.16507079303264618,
      "num_tokens": 3313966.0,
      "step": 1485
    },
    {
      "entropy": 6.124730253219605,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.1394,
      "mean_token_accuracy": 0.15149711593985557,
      "num_tokens": 3326017.0,
      "step": 1490
    },
    {
      "entropy": 6.235573196411133,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499965035001542,
      "loss": 6.0274,
      "mean_token_accuracy": 0.1512436166405678,
      "num_tokens": 3338264.0,
      "step": 1495
    },
    {
      "entropy": 6.172114086151123,
      "epoch": 0.16847307238726345,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.0762,
      "mean_token_accuracy": 0.15923445373773576,
      "num_tokens": 3348901.0,
      "step": 1500
    },
    {
      "entropy": 6.156464338302612,
      "epoch": 0.16903464929522097,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.01,
      "mean_token_accuracy": 0.1561951994895935,
      "num_tokens": 3358441.0,
      "step": 1505
    },
    {
      "entropy": 6.230216312408447,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.0741,
      "mean_token_accuracy": 0.1592329427599907,
      "num_tokens": 3369334.0,
      "step": 1510
    },
    {
      "entropy": 6.171168613433838,
      "epoch": 0.17015780311113607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.0778,
      "mean_token_accuracy": 0.15496481209993362,
      "num_tokens": 3380698.0,
      "step": 1515
    },
    {
      "entropy": 6.309156847000122,
      "epoch": 0.17071938001909362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.0899,
      "mean_token_accuracy": 0.14700602293014525,
      "num_tokens": 3391061.0,
      "step": 1520
    },
    {
      "entropy": 6.197608804702758,
      "epoch": 0.17128095692705117,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.1364,
      "mean_token_accuracy": 0.15024443566799164,
      "num_tokens": 3403983.0,
      "step": 1525
    },
    {
      "entropy": 6.244645118713379,
      "epoch": 0.1718425338350087,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049995990508388,
      "loss": 6.0779,
      "mean_token_accuracy": 0.15672694891691208,
      "num_tokens": 3413838.0,
      "step": 1530
    },
    {
      "entropy": 6.046824502944946,
      "epoch": 0.17240411074296624,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999591435870674,
      "loss": 5.9089,
      "mean_token_accuracy": 0.15403609424829484,
      "num_tokens": 3423978.0,
      "step": 1535
    },
    {
      "entropy": 6.206249952316284,
      "epoch": 0.1729656876509238,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.075,
      "mean_token_accuracy": 0.1552916258573532,
      "num_tokens": 3433416.0,
      "step": 1540
    },
    {
      "entropy": 6.05125732421875,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999575991051074,
      "loss": 5.9337,
      "mean_token_accuracy": 0.15946417003870011,
      "num_tokens": 3444983.0,
      "step": 1545
    },
    {
      "entropy": 6.104172706604004,
      "epoch": 0.1740888414668389,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999568161200093,
      "loss": 5.9132,
      "mean_token_accuracy": 0.16011491119861604,
      "num_tokens": 3455808.0,
      "step": 1550
    },
    {
      "entropy": 6.068035030364991,
      "epoch": 0.17465041837479642,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.0633,
      "mean_token_accuracy": 0.14862860813736917,
      "num_tokens": 3467503.0,
      "step": 1555
    },
    {
      "entropy": 6.333144569396973,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.2627,
      "mean_token_accuracy": 0.14934148043394088,
      "num_tokens": 3479941.0,
      "step": 1560
    },
    {
      "entropy": 6.0875753402709964,
      "epoch": 0.17577357219071152,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999544241885421,
      "loss": 5.9409,
      "mean_token_accuracy": 0.1608643963932991,
      "num_tokens": 3491069.0,
      "step": 1565
    },
    {
      "entropy": 6.185438203811645,
      "epoch": 0.17633514909866907,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.1219,
      "mean_token_accuracy": 0.14819481000304222,
      "num_tokens": 3502065.0,
      "step": 1570
    },
    {
      "entropy": 6.242910718917846,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499952793754341,
      "loss": 6.0773,
      "mean_token_accuracy": 0.15577124506235124,
      "num_tokens": 3514228.0,
      "step": 1575
    },
    {
      "entropy": 6.135730504989624,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.0138,
      "mean_token_accuracy": 0.1518117979168892,
      "num_tokens": 3525801.0,
      "step": 1580
    },
    {
      "entropy": 6.129576301574707,
      "epoch": 0.1780198798225417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999511346698035,
      "loss": 5.9359,
      "mean_token_accuracy": 0.1553277626633644,
      "num_tokens": 3536431.0,
      "step": 1585
    },
    {
      "entropy": 6.099354934692383,
      "epoch": 0.17858145673049924,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.051,
      "mean_token_accuracy": 0.15079501271247864,
      "num_tokens": 3549098.0,
      "step": 1590
    },
    {
      "entropy": 6.210941171646118,
      "epoch": 0.1791430336384568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.0178,
      "mean_token_accuracy": 0.14901258572936057,
      "num_tokens": 3560532.0,
      "step": 1595
    },
    {
      "entropy": 6.065124273300171,
      "epoch": 0.17970461054641435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.0173,
      "mean_token_accuracy": 0.1576265998184681,
      "num_tokens": 3572175.0,
      "step": 1600
    },
    {
      "entropy": 6.19850378036499,
      "epoch": 0.18026618745437187,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999477305505675,
      "loss": 5.9437,
      "mean_token_accuracy": 0.17433226928114892,
      "num_tokens": 3582862.0,
      "step": 1605
    },
    {
      "entropy": 6.204102182388306,
      "epoch": 0.18082776436232942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.1143,
      "mean_token_accuracy": 0.15742965340614318,
      "num_tokens": 3594685.0,
      "step": 1610
    },
    {
      "entropy": 6.077350044250489,
      "epoch": 0.18138934127028697,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999459855163026,
      "loss": 5.9041,
      "mean_token_accuracy": 0.16472332179546356,
      "num_tokens": 3604847.0,
      "step": 1615
    },
    {
      "entropy": 6.121081256866455,
      "epoch": 0.18195091817824452,
      "grad_norm": 1.0,
      "learning_rate": 0.000499945102255605,
      "loss": 6.0043,
      "mean_token_accuracy": 0.1578712470829487,
      "num_tokens": 3615549.0,
      "step": 1620
    },
    {
      "entropy": 6.102070951461792,
      "epoch": 0.18251249508620204,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999442118325682,
      "loss": 5.9327,
      "mean_token_accuracy": 0.16479116231203078,
      "num_tokens": 3626300.0,
      "step": 1625
    },
    {
      "entropy": 6.038302087783814,
      "epoch": 0.1830740719941596,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999433142472205,
      "loss": 5.9514,
      "mean_token_accuracy": 0.15335373133420943,
      "num_tokens": 3637832.0,
      "step": 1630
    },
    {
      "entropy": 6.115744400024414,
      "epoch": 0.18363564890211714,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999424094995902,
      "loss": 5.9789,
      "mean_token_accuracy": 0.15215082317590714,
      "num_tokens": 3649179.0,
      "step": 1635
    },
    {
      "entropy": 6.125009965896607,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.0199,
      "mean_token_accuracy": 0.1541282758116722,
      "num_tokens": 3660011.0,
      "step": 1640
    },
    {
      "entropy": 6.208988380432129,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.0967,
      "mean_token_accuracy": 0.1505162686109543,
      "num_tokens": 3672791.0,
      "step": 1645
    },
    {
      "entropy": 6.054459905624389,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999396522832949,
      "loss": 5.9203,
      "mean_token_accuracy": 0.15148744434118272,
      "num_tokens": 3684709.0,
      "step": 1650
    },
    {
      "entropy": 6.075024271011353,
      "epoch": 0.18588195653394732,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999387188868256,
      "loss": 5.8949,
      "mean_token_accuracy": 0.15390077084302903,
      "num_tokens": 3694675.0,
      "step": 1655
    },
    {
      "entropy": 6.017467975616455,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999377783282203,
      "loss": 5.8928,
      "mean_token_accuracy": 0.1555798292160034,
      "num_tokens": 3706384.0,
      "step": 1660
    },
    {
      "entropy": 6.1620267868042,
      "epoch": 0.18700511034986242,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999368306075089,
      "loss": 5.9712,
      "mean_token_accuracy": 0.15607626289129256,
      "num_tokens": 3717419.0,
      "step": 1665
    },
    {
      "entropy": 6.114587497711182,
      "epoch": 0.18756668725781997,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.0022,
      "mean_token_accuracy": 0.15720313042402267,
      "num_tokens": 3727306.0,
      "step": 1670
    },
    {
      "entropy": 6.247707271575928,
      "epoch": 0.1881282641657775,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.1246,
      "mean_token_accuracy": 0.14902206510305405,
      "num_tokens": 3739476.0,
      "step": 1675
    },
    {
      "entropy": 6.045167350769043,
      "epoch": 0.18868984107373504,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999339444730409,
      "loss": 5.8815,
      "mean_token_accuracy": 0.16054750233888626,
      "num_tokens": 3751035.0,
      "step": 1680
    },
    {
      "entropy": 6.094293832778931,
      "epoch": 0.1892514179816926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999329681042092,
      "loss": 5.9816,
      "mean_token_accuracy": 0.15803841948509217,
      "num_tokens": 3762881.0,
      "step": 1685
    },
    {
      "entropy": 6.10405535697937,
      "epoch": 0.18981299488965014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999319845734243,
      "loss": 5.9942,
      "mean_token_accuracy": 0.15585220605134964,
      "num_tokens": 3774180.0,
      "step": 1690
    },
    {
      "entropy": 6.085261487960816,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999309938807181,
      "loss": 5.9852,
      "mean_token_accuracy": 0.1583825647830963,
      "num_tokens": 3784864.0,
      "step": 1695
    },
    {
      "entropy": 6.159953594207764,
      "epoch": 0.19093614870556522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999299960261216,
      "loss": 5.9661,
      "mean_token_accuracy": 0.15621330142021178,
      "num_tokens": 3797693.0,
      "step": 1700
    },
    {
      "entropy": 6.121345901489258,
      "epoch": 0.19149772561352277,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999289910096669,
      "loss": 5.9815,
      "mean_token_accuracy": 0.1569610618054867,
      "num_tokens": 3809379.0,
      "step": 1705
    },
    {
      "entropy": 5.980222129821778,
      "epoch": 0.19205930252148032,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999279788313858,
      "loss": 5.8466,
      "mean_token_accuracy": 0.16340151876211167,
      "num_tokens": 3819473.0,
      "step": 1710
    },
    {
      "entropy": 5.951950740814209,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999269594913107,
      "loss": 5.8416,
      "mean_token_accuracy": 0.16119720190763473,
      "num_tokens": 3831037.0,
      "step": 1715
    },
    {
      "entropy": 6.028258419036865,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999259329894739,
      "loss": 5.9828,
      "mean_token_accuracy": 0.1525442384183407,
      "num_tokens": 3841767.0,
      "step": 1720
    },
    {
      "entropy": 5.972176122665405,
      "epoch": 0.19374403324535294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999248993259081,
      "loss": 5.8144,
      "mean_token_accuracy": 0.16172564327716826,
      "num_tokens": 3851629.0,
      "step": 1725
    },
    {
      "entropy": 6.02704439163208,
      "epoch": 0.1943056101533105,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999238585006463,
      "loss": 5.888,
      "mean_token_accuracy": 0.15862395614385605,
      "num_tokens": 3863247.0,
      "step": 1730
    },
    {
      "entropy": 6.070574378967285,
      "epoch": 0.19486718706126804,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999228105137218,
      "loss": 5.8708,
      "mean_token_accuracy": 0.15669515430927278,
      "num_tokens": 3873949.0,
      "step": 1735
    },
    {
      "entropy": 6.089202356338501,
      "epoch": 0.1954287639692256,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999217553651676,
      "loss": 5.9734,
      "mean_token_accuracy": 0.1567447692155838,
      "num_tokens": 3884992.0,
      "step": 1740
    },
    {
      "entropy": 6.181120681762695,
      "epoch": 0.19599034087718312,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.0919,
      "mean_token_accuracy": 0.16048870831727982,
      "num_tokens": 3897029.0,
      "step": 1745
    },
    {
      "entropy": 6.045256233215332,
      "epoch": 0.19655191778514067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999196235833054,
      "loss": 5.9408,
      "mean_token_accuracy": 0.1560034304857254,
      "num_tokens": 3909510.0,
      "step": 1750
    },
    {
      "entropy": 6.065906572341919,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999185469500651,
      "loss": 5.8437,
      "mean_token_accuracy": 0.1581835687160492,
      "num_tokens": 3920593.0,
      "step": 1755
    },
    {
      "entropy": 6.056929016113282,
      "epoch": 0.19767507160105577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499917463155331,
      "loss": 5.9727,
      "mean_token_accuracy": 0.15137458369135856,
      "num_tokens": 3932215.0,
      "step": 1760
    },
    {
      "entropy": 6.078279876708985,
      "epoch": 0.19823664850901332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999163721991378,
      "loss": 5.9258,
      "mean_token_accuracy": 0.1602189153432846,
      "num_tokens": 3943876.0,
      "step": 1765
    },
    {
      "entropy": 5.983368635177612,
      "epoch": 0.19879822541697084,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049991527408152,
      "loss": 5.9166,
      "mean_token_accuracy": 0.1628502070903778,
      "num_tokens": 3954533.0,
      "step": 1770
    },
    {
      "entropy": 6.112651348114014,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999141688025127,
      "loss": 5.9763,
      "mean_token_accuracy": 0.16206707656383515,
      "num_tokens": 3965722.0,
      "step": 1775
    },
    {
      "entropy": 6.054543972015381,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499913056362151,
      "loss": 5.9075,
      "mean_token_accuracy": 0.1556747227907181,
      "num_tokens": 3976737.0,
      "step": 1780
    },
    {
      "entropy": 6.070246934890747,
      "epoch": 0.2004829561408435,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999119367604703,
      "loss": 5.9879,
      "mean_token_accuracy": 0.16562585681676864,
      "num_tokens": 3986102.0,
      "step": 1785
    },
    {
      "entropy": 6.096411323547363,
      "epoch": 0.20104453304880104,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999108099975064,
      "loss": 5.9518,
      "mean_token_accuracy": 0.15884745121002197,
      "num_tokens": 3996534.0,
      "step": 1790
    },
    {
      "entropy": 6.006507396697998,
      "epoch": 0.20160610995675857,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499909676073295,
      "loss": 5.8641,
      "mean_token_accuracy": 0.15670980811119078,
      "num_tokens": 4006947.0,
      "step": 1795
    },
    {
      "entropy": 6.010478734970093,
      "epoch": 0.20216768686471612,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999085349878723,
      "loss": 5.9085,
      "mean_token_accuracy": 0.16509864330291749,
      "num_tokens": 4017723.0,
      "step": 1800
    },
    {
      "entropy": 6.017948055267334,
      "epoch": 0.20272926377267367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999073867412746,
      "loss": 5.9857,
      "mean_token_accuracy": 0.1533601000905037,
      "num_tokens": 4028347.0,
      "step": 1805
    },
    {
      "entropy": 6.06327452659607,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999062313335385,
      "loss": 5.846,
      "mean_token_accuracy": 0.16621426939964296,
      "num_tokens": 4039648.0,
      "step": 1810
    },
    {
      "entropy": 6.054847764968872,
      "epoch": 0.20385241758858877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999050687647008,
      "loss": 5.9093,
      "mean_token_accuracy": 0.15702853202819825,
      "num_tokens": 4050786.0,
      "step": 1815
    },
    {
      "entropy": 6.053362274169922,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999038990347984,
      "loss": 5.9447,
      "mean_token_accuracy": 0.16274867802858353,
      "num_tokens": 4061184.0,
      "step": 1820
    },
    {
      "entropy": 6.157382345199585,
      "epoch": 0.20497557140450384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.0754,
      "mean_token_accuracy": 0.15608905404806137,
      "num_tokens": 4073607.0,
      "step": 1825
    },
    {
      "entropy": 5.970077323913574,
      "epoch": 0.2055371483124614,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499901538091949,
      "loss": 5.8927,
      "mean_token_accuracy": 0.1595301076769829,
      "num_tokens": 4083924.0,
      "step": 1830
    },
    {
      "entropy": 6.023391008377075,
      "epoch": 0.20609872522041894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999003468790773,
      "loss": 5.8918,
      "mean_token_accuracy": 0.16054748892784118,
      "num_tokens": 4094083.0,
      "step": 1835
    },
    {
      "entropy": 6.0351855754852295,
      "epoch": 0.2066603021283765,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998991485052911,
      "loss": 5.9218,
      "mean_token_accuracy": 0.16628677994012833,
      "num_tokens": 4105124.0,
      "step": 1840
    },
    {
      "entropy": 6.1698376655578615,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499897942970629,
      "loss": 6.0908,
      "mean_token_accuracy": 0.1494095504283905,
      "num_tokens": 4116792.0,
      "step": 1845
    },
    {
      "entropy": 6.220897674560547,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499896730275129,
      "loss": 6.0706,
      "mean_token_accuracy": 0.15271387100219727,
      "num_tokens": 4127441.0,
      "step": 1850
    },
    {
      "entropy": 6.071835470199585,
      "epoch": 0.20834503285224912,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049989551041883,
      "loss": 6.0561,
      "mean_token_accuracy": 0.15183976292610168,
      "num_tokens": 4138696.0,
      "step": 1855
    },
    {
      "entropy": 6.061728000640869,
      "epoch": 0.20890660976020667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998942834017707,
      "loss": 5.915,
      "mean_token_accuracy": 0.15786469280719756,
      "num_tokens": 4151254.0,
      "step": 1860
    },
    {
      "entropy": 5.997844457626343,
      "epoch": 0.2094681866681642,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998930492239901,
      "loss": 5.8837,
      "mean_token_accuracy": 0.1549679934978485,
      "num_tokens": 4162046.0,
      "step": 1865
    },
    {
      "entropy": 5.995732355117798,
      "epoch": 0.21002976357612174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998918078855277,
      "loss": 5.8325,
      "mean_token_accuracy": 0.1628045290708542,
      "num_tokens": 4172409.0,
      "step": 1870
    },
    {
      "entropy": 5.986175918579102,
      "epoch": 0.2105913404840793,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998905593864227,
      "loss": 5.9844,
      "mean_token_accuracy": 0.16071528941392899,
      "num_tokens": 4182949.0,
      "step": 1875
    },
    {
      "entropy": 6.048963737487793,
      "epoch": 0.21115291739203684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998893037267153,
      "loss": 5.9605,
      "mean_token_accuracy": 0.15584419518709183,
      "num_tokens": 4194164.0,
      "step": 1880
    },
    {
      "entropy": 6.133493423461914,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998880409064452,
      "loss": 5.9129,
      "mean_token_accuracy": 0.16219342201948167,
      "num_tokens": 4206056.0,
      "step": 1885
    },
    {
      "entropy": 5.956058120727539,
      "epoch": 0.21227607120795192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998867709256527,
      "loss": 5.9854,
      "mean_token_accuracy": 0.15834779739379884,
      "num_tokens": 4216660.0,
      "step": 1890
    },
    {
      "entropy": 6.011321544647217,
      "epoch": 0.21283764811590947,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499885493784378,
      "loss": 5.8419,
      "mean_token_accuracy": 0.16390256732702255,
      "num_tokens": 4228495.0,
      "step": 1895
    },
    {
      "entropy": 6.061332416534424,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998842094826621,
      "loss": 5.8681,
      "mean_token_accuracy": 0.1624634876847267,
      "num_tokens": 4238560.0,
      "step": 1900
    },
    {
      "entropy": 6.160039234161377,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.0703,
      "mean_token_accuracy": 0.15231714472174646,
      "num_tokens": 4250553.0,
      "step": 1905
    },
    {
      "entropy": 6.0768242359161375,
      "epoch": 0.21452237883978212,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998816193980701,
      "loss": 5.9904,
      "mean_token_accuracy": 0.15530536472797393,
      "num_tokens": 4262050.0,
      "step": 1910
    },
    {
      "entropy": 6.142356586456299,
      "epoch": 0.21508395574773964,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998803136152764,
      "loss": 5.9473,
      "mean_token_accuracy": 0.15309242606163026,
      "num_tokens": 4273019.0,
      "step": 1915
    },
    {
      "entropy": 6.004923295974732,
      "epoch": 0.2156455326556972,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998790006722063,
      "loss": 5.8254,
      "mean_token_accuracy": 0.16646857857704161,
      "num_tokens": 4283585.0,
      "step": 1920
    },
    {
      "entropy": 5.993122625350952,
      "epoch": 0.21620710956365474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998776805689016,
      "loss": 5.9043,
      "mean_token_accuracy": 0.161615289747715,
      "num_tokens": 4294905.0,
      "step": 1925
    },
    {
      "entropy": 6.161378240585327,
      "epoch": 0.2167686864716123,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998763533054045,
      "loss": 6.0087,
      "mean_token_accuracy": 0.15805915147066116,
      "num_tokens": 4305926.0,
      "step": 1930
    },
    {
      "entropy": 5.925935316085815,
      "epoch": 0.21733026337956984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998750188817568,
      "loss": 5.9809,
      "mean_token_accuracy": 0.1576686106622219,
      "num_tokens": 4318297.0,
      "step": 1935
    },
    {
      "entropy": 5.998038911819458,
      "epoch": 0.21789184028752737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.8028,
      "mean_token_accuracy": 0.16835505217313768,
      "num_tokens": 4328795.0,
      "step": 1940
    },
    {
      "entropy": 5.970638608932495,
      "epoch": 0.21845341719548492,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499872328554181,
      "loss": 5.8275,
      "mean_token_accuracy": 0.16290033608675003,
      "num_tokens": 4340279.0,
      "step": 1945
    },
    {
      "entropy": 6.099689674377442,
      "epoch": 0.21901499410344247,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998709726503383,
      "loss": 6.0003,
      "mean_token_accuracy": 0.1518409989774227,
      "num_tokens": 4352157.0,
      "step": 1950
    },
    {
      "entropy": 5.986471176147461,
      "epoch": 0.21957657101140002,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998696095865168,
      "loss": 5.824,
      "mean_token_accuracy": 0.16872964352369307,
      "num_tokens": 4361909.0,
      "step": 1955
    },
    {
      "entropy": 5.843346214294433,
      "epoch": 0.22013814791935757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.7623,
      "mean_token_accuracy": 0.16021531522274018,
      "num_tokens": 4372177.0,
      "step": 1960
    },
    {
      "entropy": 5.961346864700317,
      "epoch": 0.2206997248273151,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.8063,
      "mean_token_accuracy": 0.16359231173992156,
      "num_tokens": 4382872.0,
      "step": 1965
    },
    {
      "entropy": 6.00990309715271,
      "epoch": 0.22126130173527264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998654774356131,
      "loss": 5.9066,
      "mean_token_accuracy": 0.15922408998012544,
      "num_tokens": 4393623.0,
      "step": 1970
    },
    {
      "entropy": 5.948709440231323,
      "epoch": 0.2218228786432302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998640857323118,
      "loss": 5.8402,
      "mean_token_accuracy": 0.15961378663778306,
      "num_tokens": 4404215.0,
      "step": 1975
    },
    {
      "entropy": 6.1061646938323975,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998626868692509,
      "loss": 6.0948,
      "mean_token_accuracy": 0.15816606134176253,
      "num_tokens": 4415978.0,
      "step": 1980
    },
    {
      "entropy": 6.046572494506836,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998612808464748,
      "loss": 5.8441,
      "mean_token_accuracy": 0.16600802540779114,
      "num_tokens": 4427060.0,
      "step": 1985
    },
    {
      "entropy": 5.935293436050415,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998598676640283,
      "loss": 5.9204,
      "mean_token_accuracy": 0.15637087672948838,
      "num_tokens": 4438145.0,
      "step": 1990
    },
    {
      "entropy": 6.165454578399658,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998584473219563,
      "loss": 5.9893,
      "mean_token_accuracy": 0.1606801465153694,
      "num_tokens": 4449324.0,
      "step": 1995
    },
    {
      "entropy": 5.938971281051636,
      "epoch": 0.22463076318301792,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.7297,
      "mean_token_accuracy": 0.1644525095820427,
      "num_tokens": 4459728.0,
      "step": 2000
    },
    {
      "entropy": 5.961595249176026,
      "epoch": 0.22519234009097547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998555851591175,
      "loss": 5.9458,
      "mean_token_accuracy": 0.16193436831235886,
      "num_tokens": 4471164.0,
      "step": 2005
    },
    {
      "entropy": 6.059179592132568,
      "epoch": 0.225753916998933,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.0017,
      "mean_token_accuracy": 0.1561392977833748,
      "num_tokens": 4483349.0,
      "step": 2010
    },
    {
      "entropy": 6.103359651565552,
      "epoch": 0.22631549390689054,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998526943583226,
      "loss": 5.9563,
      "mean_token_accuracy": 0.15778704583644867,
      "num_tokens": 4495311.0,
      "step": 2015
    },
    {
      "entropy": 5.927799892425537,
      "epoch": 0.2268770708148481,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998512382188066,
      "loss": 5.8836,
      "mean_token_accuracy": 0.1617691360414028,
      "num_tokens": 4507017.0,
      "step": 2020
    },
    {
      "entropy": 5.996227025985718,
      "epoch": 0.22743864772280564,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998497749199401,
      "loss": 5.9212,
      "mean_token_accuracy": 0.15882934331893922,
      "num_tokens": 4517746.0,
      "step": 2025
    },
    {
      "entropy": 6.099315357208252,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998483044617694,
      "loss": 6.0097,
      "mean_token_accuracy": 0.15352682620286942,
      "num_tokens": 4529149.0,
      "step": 2030
    },
    {
      "entropy": 6.046715021133423,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998468268443416,
      "loss": 5.8928,
      "mean_token_accuracy": 0.1632772609591484,
      "num_tokens": 4540630.0,
      "step": 2035
    },
    {
      "entropy": 5.875202322006226,
      "epoch": 0.22912337844667827,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.8252,
      "mean_token_accuracy": 0.16278187185525894,
      "num_tokens": 4552877.0,
      "step": 2040
    },
    {
      "entropy": 6.0447197437286375,
      "epoch": 0.22968495535463582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998438501319025,
      "loss": 5.9238,
      "mean_token_accuracy": 0.16725802570581436,
      "num_tokens": 4563634.0,
      "step": 2045
    },
    {
      "entropy": 6.03820834159851,
      "epoch": 0.23024653226259337,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998423510369862,
      "loss": 5.8429,
      "mean_token_accuracy": 0.16436342447996138,
      "num_tokens": 4574007.0,
      "step": 2050
    },
    {
      "entropy": 5.93084487915039,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998408447830022,
      "loss": 5.8515,
      "mean_token_accuracy": 0.16042328476905823,
      "num_tokens": 4584146.0,
      "step": 2055
    },
    {
      "entropy": 6.019020128250122,
      "epoch": 0.23136968607850844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998393313699985,
      "loss": 5.8658,
      "mean_token_accuracy": 0.15834631770849228,
      "num_tokens": 4594839.0,
      "step": 2060
    },
    {
      "entropy": 6.118397998809814,
      "epoch": 0.231931262986466,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998378107980233,
      "loss": 6.0378,
      "mean_token_accuracy": 0.15390192568302155,
      "num_tokens": 4608229.0,
      "step": 2065
    },
    {
      "entropy": 6.0200855255126955,
      "epoch": 0.23249283989442354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998362830671249,
      "loss": 5.8386,
      "mean_token_accuracy": 0.1594712480902672,
      "num_tokens": 4619206.0,
      "step": 2070
    },
    {
      "entropy": 5.911860418319702,
      "epoch": 0.2330544168023811,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998347481773522,
      "loss": 5.8782,
      "mean_token_accuracy": 0.16451769918203354,
      "num_tokens": 4628966.0,
      "step": 2075
    },
    {
      "entropy": 6.036902618408203,
      "epoch": 0.23361599371033864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998332061287538,
      "loss": 5.753,
      "mean_token_accuracy": 0.16608522683382035,
      "num_tokens": 4639935.0,
      "step": 2080
    },
    {
      "entropy": 6.02183985710144,
      "epoch": 0.23417757061829617,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499831656921379,
      "loss": 5.9959,
      "mean_token_accuracy": 0.15738053619861603,
      "num_tokens": 4651998.0,
      "step": 2085
    },
    {
      "entropy": 6.092220067977905,
      "epoch": 0.23473914752625372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499830100555277,
      "loss": 5.9932,
      "mean_token_accuracy": 0.15611726194620132,
      "num_tokens": 4664002.0,
      "step": 2090
    },
    {
      "entropy": 5.96814341545105,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998285370304974,
      "loss": 5.8598,
      "mean_token_accuracy": 0.16325256526470183,
      "num_tokens": 4675343.0,
      "step": 2095
    },
    {
      "entropy": 6.047848653793335,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049982696634709,
      "loss": 5.8987,
      "mean_token_accuracy": 0.16740916967391967,
      "num_tokens": 4686410.0,
      "step": 2100
    },
    {
      "entropy": 5.984486770629883,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998253885051048,
      "loss": 5.9357,
      "mean_token_accuracy": 0.16218028217554092,
      "num_tokens": 4697768.0,
      "step": 2105
    },
    {
      "entropy": 6.095366096496582,
      "epoch": 0.2369854551580839,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499823803504592,
      "loss": 5.9388,
      "mean_token_accuracy": 0.1598934069275856,
      "num_tokens": 4709933.0,
      "step": 2110
    },
    {
      "entropy": 6.017884826660156,
      "epoch": 0.23754703206604144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998222113456022,
      "loss": 5.9774,
      "mean_token_accuracy": 0.1559494897723198,
      "num_tokens": 4721385.0,
      "step": 2115
    },
    {
      "entropy": 5.9762331485748295,
      "epoch": 0.238108608973999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998206120281858,
      "loss": 5.7984,
      "mean_token_accuracy": 0.1687019482254982,
      "num_tokens": 4732562.0,
      "step": 2120
    },
    {
      "entropy": 5.9636999607086185,
      "epoch": 0.23867018588195654,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499819005552394,
      "loss": 5.8117,
      "mean_token_accuracy": 0.16452171057462692,
      "num_tokens": 4742794.0,
      "step": 2125
    },
    {
      "entropy": 5.919935417175293,
      "epoch": 0.23923176278991407,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.7881,
      "mean_token_accuracy": 0.166632841527462,
      "num_tokens": 4754903.0,
      "step": 2130
    },
    {
      "entropy": 5.9008917808532715,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499815771125889,
      "loss": 5.8175,
      "mean_token_accuracy": 0.16199010461568833,
      "num_tokens": 4766165.0,
      "step": 2135
    },
    {
      "entropy": 5.972467041015625,
      "epoch": 0.24035491660582917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998141431752785,
      "loss": 5.8284,
      "mean_token_accuracy": 0.16583052426576614,
      "num_tokens": 4776551.0,
      "step": 2140
    },
    {
      "entropy": 6.0888008117675785,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998125080664987,
      "loss": 6.0294,
      "mean_token_accuracy": 0.15712772458791732,
      "num_tokens": 4788344.0,
      "step": 2145
    },
    {
      "entropy": 6.032654619216919,
      "epoch": 0.24147807042174427,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.8201,
      "mean_token_accuracy": 0.16204651147127153,
      "num_tokens": 4799314.0,
      "step": 2150
    },
    {
      "entropy": 5.93705735206604,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998092163746387,
      "loss": 5.8618,
      "mean_token_accuracy": 0.16488695591688157,
      "num_tokens": 4809588.0,
      "step": 2155
    },
    {
      "entropy": 5.907848787307739,
      "epoch": 0.24260122423765934,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.6812,
      "mean_token_accuracy": 0.17253921180963516,
      "num_tokens": 4820076.0,
      "step": 2160
    },
    {
      "entropy": 5.989373302459716,
      "epoch": 0.2431628011456169,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998058960507286,
      "loss": 5.9562,
      "mean_token_accuracy": 0.1578843265771866,
      "num_tokens": 4831504.0,
      "step": 2165
    },
    {
      "entropy": 5.992100238800049,
      "epoch": 0.24372437805357444,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.8054,
      "mean_token_accuracy": 0.16406459957361222,
      "num_tokens": 4842826.0,
      "step": 2170
    },
    {
      "entropy": 5.86922173500061,
      "epoch": 0.244285954961532,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998025470951908,
      "loss": 5.8005,
      "mean_token_accuracy": 0.16625914573669434,
      "num_tokens": 4853279.0,
      "step": 2175
    },
    {
      "entropy": 6.003297185897827,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998008618806949,
      "loss": 5.8589,
      "mean_token_accuracy": 0.1688696026802063,
      "num_tokens": 4863332.0,
      "step": 2180
    },
    {
      "entropy": 5.924489688873291,
      "epoch": 0.24540910877744707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997991695084523,
      "loss": 5.9807,
      "mean_token_accuracy": 0.15316371172666549,
      "num_tokens": 4874217.0,
      "step": 2185
    },
    {
      "entropy": 6.1011065483093265,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997974699785169,
      "loss": 5.8362,
      "mean_token_accuracy": 0.16836950108408927,
      "num_tokens": 4885322.0,
      "step": 2190
    },
    {
      "entropy": 5.828060865402222,
      "epoch": 0.24653226259336217,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.781,
      "mean_token_accuracy": 0.16213512569665908,
      "num_tokens": 4895546.0,
      "step": 2195
    },
    {
      "entropy": 6.1020842552185055,
      "epoch": 0.24709383950131972,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997940494457846,
      "loss": 5.9813,
      "mean_token_accuracy": 0.157517471909523,
      "num_tokens": 4906957.0,
      "step": 2200
    },
    {
      "entropy": 6.115909624099731,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997923284430967,
      "loss": 5.9104,
      "mean_token_accuracy": 0.1537712886929512,
      "num_tokens": 4917965.0,
      "step": 2205
    },
    {
      "entropy": 5.8366883277893065,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.7987,
      "mean_token_accuracy": 0.16758462488651277,
      "num_tokens": 4928585.0,
      "step": 2210
    },
    {
      "entropy": 5.961793851852417,
      "epoch": 0.24877857022519234,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499788864965351,
      "loss": 5.8096,
      "mean_token_accuracy": 0.16297494918107985,
      "num_tokens": 4939818.0,
      "step": 2215
    },
    {
      "entropy": 6.018909597396851,
      "epoch": 0.2493401471331499,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997871224904036,
      "loss": 5.8987,
      "mean_token_accuracy": 0.1577194705605507,
      "num_tokens": 4951752.0,
      "step": 2220
    },
    {
      "entropy": 5.85117998123169,
      "epoch": 0.24990172404110744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.7255,
      "mean_token_accuracy": 0.17753110378980635,
      "num_tokens": 4961846.0,
      "step": 2225
    },
    {
      "entropy": 5.963651466369629,
      "epoch": 0.250463300949065,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997836160686368,
      "loss": 5.8962,
      "mean_token_accuracy": 0.15754956007003784,
      "num_tokens": 4973693.0,
      "step": 2230
    },
    {
      "entropy": 5.936974811553955,
      "epoch": 0.25102487785702254,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16598349809646606,
      "num_tokens": 4986156.0,
      "step": 2235
    },
    {
      "entropy": 5.9454022407531735,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997800810180803,
      "loss": 5.8731,
      "mean_token_accuracy": 0.16364911198616028,
      "num_tokens": 4997615.0,
      "step": 2240
    },
    {
      "entropy": 5.981971311569214,
      "epoch": 0.2521480316729376,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.8097,
      "mean_token_accuracy": 0.16514143496751785,
      "num_tokens": 5007655.0,
      "step": 2245
    },
    {
      "entropy": 5.966670131683349,
      "epoch": 0.25270960858089514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997765173391841,
      "loss": 5.8812,
      "mean_token_accuracy": 0.16588087528944015,
      "num_tokens": 5018933.0,
      "step": 2250
    },
    {
      "entropy": 6.058870506286621,
      "epoch": 0.2532711854888527,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997747247642505,
      "loss": 5.8748,
      "mean_token_accuracy": 0.16249170750379563,
      "num_tokens": 5030495.0,
      "step": 2255
    },
    {
      "entropy": 5.938293361663819,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997729250324023,
      "loss": 5.916,
      "mean_token_accuracy": 0.15933229476213456,
      "num_tokens": 5041553.0,
      "step": 2260
    },
    {
      "entropy": 5.946228456497193,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997711181436971,
      "loss": 5.8639,
      "mean_token_accuracy": 0.16556437611579894,
      "num_tokens": 5052312.0,
      "step": 2265
    },
    {
      "entropy": 6.082585287094116,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997693040981922,
      "loss": 6.0487,
      "mean_token_accuracy": 0.1565891683101654,
      "num_tokens": 5063445.0,
      "step": 2270
    },
    {
      "entropy": 6.03095555305481,
      "epoch": 0.2555174931206829,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.8301,
      "mean_token_accuracy": 0.16110838502645491,
      "num_tokens": 5073052.0,
      "step": 2275
    },
    {
      "entropy": 5.920227956771851,
      "epoch": 0.25607907002864044,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997656545370151,
      "loss": 5.8694,
      "mean_token_accuracy": 0.16425251513719558,
      "num_tokens": 5084174.0,
      "step": 2280
    },
    {
      "entropy": 6.02871994972229,
      "epoch": 0.256640646936598,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499763819021459,
      "loss": 5.8353,
      "mean_token_accuracy": 0.16484149694442748,
      "num_tokens": 5094106.0,
      "step": 2285
    },
    {
      "entropy": 6.0660443782806395,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997619763493357,
      "loss": 5.9288,
      "mean_token_accuracy": 0.1632024973630905,
      "num_tokens": 5106963.0,
      "step": 2290
    },
    {
      "entropy": 5.8431133270263675,
      "epoch": 0.25776380075251304,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499760126520704,
      "loss": 5.7377,
      "mean_token_accuracy": 0.16144067347049712,
      "num_tokens": 5119476.0,
      "step": 2295
    },
    {
      "entropy": 6.031280374526977,
      "epoch": 0.2583253776604706,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997582695356225,
      "loss": 5.9879,
      "mean_token_accuracy": 0.15699385106563568,
      "num_tokens": 5132169.0,
      "step": 2300
    },
    {
      "entropy": 5.939564180374146,
      "epoch": 0.25888695456842814,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997564053941507,
      "loss": 5.82,
      "mean_token_accuracy": 0.16523557752370835,
      "num_tokens": 5143354.0,
      "step": 2305
    },
    {
      "entropy": 5.880933618545532,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997545340963478,
      "loss": 5.8924,
      "mean_token_accuracy": 0.15893735885620117,
      "num_tokens": 5154624.0,
      "step": 2310
    },
    {
      "entropy": 6.041430854797364,
      "epoch": 0.26001010838434324,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997526556422733,
      "loss": 5.7748,
      "mean_token_accuracy": 0.16341781467199326,
      "num_tokens": 5165349.0,
      "step": 2315
    },
    {
      "entropy": 5.887167644500733,
      "epoch": 0.2605716852923008,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997507700319872,
      "loss": 5.8504,
      "mean_token_accuracy": 0.16064828485250474,
      "num_tokens": 5175874.0,
      "step": 2320
    },
    {
      "entropy": 5.964330816268921,
      "epoch": 0.26113326220025834,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997488772655492,
      "loss": 5.7554,
      "mean_token_accuracy": 0.16929169446229936,
      "num_tokens": 5186591.0,
      "step": 2325
    },
    {
      "entropy": 5.903693675994873,
      "epoch": 0.2616948391082159,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997469773430199,
      "loss": 5.8519,
      "mean_token_accuracy": 0.17121090292930602,
      "num_tokens": 5197307.0,
      "step": 2330
    },
    {
      "entropy": 5.88110842704773,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.818,
      "mean_token_accuracy": 0.1627274125814438,
      "num_tokens": 5209868.0,
      "step": 2335
    },
    {
      "entropy": 5.976365756988526,
      "epoch": 0.26281799292413094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.8265,
      "mean_token_accuracy": 0.16824889183044434,
      "num_tokens": 5220365.0,
      "step": 2340
    },
    {
      "entropy": 5.904721117019653,
      "epoch": 0.2633795698320885,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.74,
      "mean_token_accuracy": 0.16793981045484543,
      "num_tokens": 5231575.0,
      "step": 2345
    },
    {
      "entropy": 5.969769144058228,
      "epoch": 0.26394114674004604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.8382,
      "mean_token_accuracy": 0.16210115253925322,
      "num_tokens": 5241852.0,
      "step": 2350
    },
    {
      "entropy": 5.970647525787354,
      "epoch": 0.2645027236480036,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997373703911273,
      "loss": 5.8794,
      "mean_token_accuracy": 0.16796267330646514,
      "num_tokens": 5254536.0,
      "step": 2355
    },
    {
      "entropy": 5.984567260742187,
      "epoch": 0.26506430055596114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.7922,
      "mean_token_accuracy": 0.16661520600318908,
      "num_tokens": 5264946.0,
      "step": 2360
    },
    {
      "entropy": 5.692500972747803,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499733477519865,
      "loss": 5.6292,
      "mean_token_accuracy": 0.16376495808362962,
      "num_tokens": 5275527.0,
      "step": 2365
    },
    {
      "entropy": 5.928407621383667,
      "epoch": 0.26618745437187624,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997315203508013,
      "loss": 5.8023,
      "mean_token_accuracy": 0.16296649128198623,
      "num_tokens": 5285471.0,
      "step": 2370
    },
    {
      "entropy": 5.835812282562256,
      "epoch": 0.2667490312798338,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.7219,
      "mean_token_accuracy": 0.16512494385242463,
      "num_tokens": 5295743.0,
      "step": 2375
    },
    {
      "entropy": 5.956495523452759,
      "epoch": 0.26731060818779134,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049972758454612,
      "loss": 5.8692,
      "mean_token_accuracy": 0.15754660218954086,
      "num_tokens": 5307963.0,
      "step": 2380
    },
    {
      "entropy": 5.9242268085479735,
      "epoch": 0.26787218509574884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.7534,
      "mean_token_accuracy": 0.1652416855096817,
      "num_tokens": 5318394.0,
      "step": 2385
    },
    {
      "entropy": 5.935852479934693,
      "epoch": 0.2684337620037064,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997236201197853,
      "loss": 5.9947,
      "mean_token_accuracy": 0.14979661107063294,
      "num_tokens": 5331728.0,
      "step": 2390
    },
    {
      "entropy": 5.94870285987854,
      "epoch": 0.26899533891166394,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997216271736554,
      "loss": 5.869,
      "mean_token_accuracy": 0.17228117436170579,
      "num_tokens": 5342825.0,
      "step": 2395
    },
    {
      "entropy": 5.978704738616943,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.7792,
      "mean_token_accuracy": 0.1661553055047989,
      "num_tokens": 5354232.0,
      "step": 2400
    },
    {
      "entropy": 5.882722234725952,
      "epoch": 0.27011849272757904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997176198157881,
      "loss": 5.8705,
      "mean_token_accuracy": 0.16065609604120254,
      "num_tokens": 5364807.0,
      "step": 2405
    },
    {
      "entropy": 5.937169790267944,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.8442,
      "mean_token_accuracy": 0.1606588914990425,
      "num_tokens": 5375593.0,
      "step": 2410
    },
    {
      "entropy": 5.968391370773316,
      "epoch": 0.27124164654349414,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.7928,
      "mean_token_accuracy": 0.17065857201814652,
      "num_tokens": 5386740.0,
      "step": 2415
    },
    {
      "entropy": 5.825274181365967,
      "epoch": 0.2718032234514517,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.7859,
      "mean_token_accuracy": 0.16433975994586944,
      "num_tokens": 5397732.0,
      "step": 2420
    },
    {
      "entropy": 5.979048871994019,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499709519239414,
      "loss": 5.8433,
      "mean_token_accuracy": 0.159425051510334,
      "num_tokens": 5409392.0,
      "step": 2425
    },
    {
      "entropy": 5.79268445968628,
      "epoch": 0.2729263772673668,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.7906,
      "mean_token_accuracy": 0.17156518399715423,
      "num_tokens": 5420903.0,
      "step": 2430
    },
    {
      "entropy": 5.852681636810303,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499705426021939,
      "loss": 5.7318,
      "mean_token_accuracy": 0.16576131135225297,
      "num_tokens": 5432631.0,
      "step": 2435
    },
    {
      "entropy": 5.858410882949829,
      "epoch": 0.27404953108328184,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499703368681107,
      "loss": 5.6874,
      "mean_token_accuracy": 0.1763172999024391,
      "num_tokens": 5443463.0,
      "step": 2440
    },
    {
      "entropy": 5.990561437606812,
      "epoch": 0.2746111079912394,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997013041856326,
      "loss": 6.0024,
      "mean_token_accuracy": 0.15876681730151176,
      "num_tokens": 5455903.0,
      "step": 2445
    },
    {
      "entropy": 5.941286277770996,
      "epoch": 0.27517268489919694,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.826,
      "mean_token_accuracy": 0.1625517025589943,
      "num_tokens": 5467201.0,
      "step": 2450
    },
    {
      "entropy": 5.9081957817077635,
      "epoch": 0.2757342618071545,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.7079,
      "mean_token_accuracy": 0.16979653686285018,
      "num_tokens": 5478109.0,
      "step": 2455
    },
    {
      "entropy": 5.908603048324585,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.8226,
      "mean_token_accuracy": 0.16229735761880876,
      "num_tokens": 5489621.0,
      "step": 2460
    },
    {
      "entropy": 5.919325733184815,
      "epoch": 0.2768574156230696,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499692974658629,
      "loss": 5.7826,
      "mean_token_accuracy": 0.16799012571573257,
      "num_tokens": 5500606.0,
      "step": 2465
    },
    {
      "entropy": 5.937128257751465,
      "epoch": 0.27741899253102714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.7359,
      "mean_token_accuracy": 0.16903213709592818,
      "num_tokens": 5510768.0,
      "step": 2470
    },
    {
      "entropy": 5.778143644332886,
      "epoch": 0.2779805694389847,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.7554,
      "mean_token_accuracy": 0.16829170882701874,
      "num_tokens": 5522210.0,
      "step": 2475
    },
    {
      "entropy": 6.009134101867676,
      "epoch": 0.2785421463469422,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996866523928747,
      "loss": 5.867,
      "mean_token_accuracy": 0.16028780341148377,
      "num_tokens": 5533055.0,
      "step": 2480
    },
    {
      "entropy": 5.925482892990113,
      "epoch": 0.27910372325489974,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996845306626467,
      "loss": 5.8813,
      "mean_token_accuracy": 0.1584608882665634,
      "num_tokens": 5544345.0,
      "step": 2485
    },
    {
      "entropy": 5.937867641448975,
      "epoch": 0.2796653001628573,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.7935,
      "mean_token_accuracy": 0.16448234766721725,
      "num_tokens": 5556734.0,
      "step": 2490
    },
    {
      "entropy": 5.975261259078979,
      "epoch": 0.28022687707081484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.8289,
      "mean_token_accuracy": 0.1649043843150139,
      "num_tokens": 5566874.0,
      "step": 2495
    },
    {
      "entropy": 5.915608930587768,
      "epoch": 0.2807884539787724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.8426,
      "mean_token_accuracy": 0.16331143230199813,
      "num_tokens": 5578812.0,
      "step": 2500
    },
    {
      "entropy": 5.944158029556275,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996759722019872,
      "loss": 5.8711,
      "mean_token_accuracy": 0.16549691557884216,
      "num_tokens": 5590021.0,
      "step": 2505
    },
    {
      "entropy": 5.941056251525879,
      "epoch": 0.2819116077946875,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996738147022262,
      "loss": 5.8929,
      "mean_token_accuracy": 0.1572035700082779,
      "num_tokens": 5600921.0,
      "step": 2510
    },
    {
      "entropy": 5.888544178009033,
      "epoch": 0.28247318470264504,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.8032,
      "mean_token_accuracy": 0.16694664508104323,
      "num_tokens": 5610615.0,
      "step": 2515
    },
    {
      "entropy": 5.788763236999512,
      "epoch": 0.2830347616106026,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499669478241669,
      "loss": 5.7268,
      "mean_token_accuracy": 0.16717411577701569,
      "num_tokens": 5621596.0,
      "step": 2520
    },
    {
      "entropy": 6.033792781829834,
      "epoch": 0.28359633851856014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.7846,
      "mean_token_accuracy": 0.16378311067819595,
      "num_tokens": 5633047.0,
      "step": 2525
    },
    {
      "entropy": 5.867732572555542,
      "epoch": 0.28415791542651764,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.7647,
      "mean_token_accuracy": 0.1646569311618805,
      "num_tokens": 5644970.0,
      "step": 2530
    },
    {
      "entropy": 5.851826286315918,
      "epoch": 0.2847194923344752,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996629198992823,
      "loss": 5.8849,
      "mean_token_accuracy": 0.16250034421682358,
      "num_tokens": 5656631.0,
      "step": 2535
    },
    {
      "entropy": 5.936293029785157,
      "epoch": 0.28528106924243274,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.8458,
      "mean_token_accuracy": 0.16177234947681426,
      "num_tokens": 5668808.0,
      "step": 2540
    },
    {
      "entropy": 5.834509754180909,
      "epoch": 0.2858426461503903,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.7581,
      "mean_token_accuracy": 0.16683479249477387,
      "num_tokens": 5679493.0,
      "step": 2545
    },
    {
      "entropy": 5.961978435516357,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.7875,
      "mean_token_accuracy": 0.17051588296890258,
      "num_tokens": 5689879.0,
      "step": 2550
    },
    {
      "entropy": 5.955478048324585,
      "epoch": 0.2869657999663054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499654075296133,
      "loss": 5.8868,
      "mean_token_accuracy": 0.16586922556161882,
      "num_tokens": 5699934.0,
      "step": 2555
    },
    {
      "entropy": 5.846363353729248,
      "epoch": 0.28752737687426294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499651846262487,
      "loss": 5.7991,
      "mean_token_accuracy": 0.17178635001182557,
      "num_tokens": 5710171.0,
      "step": 2560
    },
    {
      "entropy": 5.9693992137908936,
      "epoch": 0.2880889537822205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.788,
      "mean_token_accuracy": 0.15814389437437057,
      "num_tokens": 5721190.0,
      "step": 2565
    },
    {
      "entropy": 5.962607526779175,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996473667362604,
      "loss": 5.9165,
      "mean_token_accuracy": 0.15922336876392365,
      "num_tokens": 5733706.0,
      "step": 2570
    },
    {
      "entropy": 5.861262464523316,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996451162438224,
      "loss": 5.8646,
      "mean_token_accuracy": 0.16674669682979584,
      "num_tokens": 5745712.0,
      "step": 2575
    },
    {
      "entropy": 5.7940126895904545,
      "epoch": 0.2897736845060931,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.682,
      "mean_token_accuracy": 0.17691746205091477,
      "num_tokens": 5757915.0,
      "step": 2580
    },
    {
      "entropy": 5.956466007232666,
      "epoch": 0.29033526141405064,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.8215,
      "mean_token_accuracy": 0.17083913683891297,
      "num_tokens": 5768788.0,
      "step": 2585
    },
    {
      "entropy": 5.9059301853179935,
      "epoch": 0.2908968383220082,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.7558,
      "mean_token_accuracy": 0.1652416318655014,
      "num_tokens": 5779806.0,
      "step": 2590
    },
    {
      "entropy": 5.77002182006836,
      "epoch": 0.29145841522996574,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499636042746914,
      "loss": 5.7209,
      "mean_token_accuracy": 0.15952013283967972,
      "num_tokens": 5790177.0,
      "step": 2595
    },
    {
      "entropy": 5.923643255233765,
      "epoch": 0.2920199921379233,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499633756491259,
      "loss": 5.7569,
      "mean_token_accuracy": 0.1707965239882469,
      "num_tokens": 5800251.0,
      "step": 2600
    },
    {
      "entropy": 5.936167621612549,
      "epoch": 0.29258156904588084,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996314630831781,
      "loss": 5.8339,
      "mean_token_accuracy": 0.16565884947776793,
      "num_tokens": 5812210.0,
      "step": 2605
    },
    {
      "entropy": 5.918135690689087,
      "epoch": 0.2931431459538384,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.823,
      "mean_token_accuracy": 0.1682283952832222,
      "num_tokens": 5823280.0,
      "step": 2610
    },
    {
      "entropy": 5.983985757827758,
      "epoch": 0.29370472286179594,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996268548100307,
      "loss": 5.8846,
      "mean_token_accuracy": 0.15846949815750122,
      "num_tokens": 5835076.0,
      "step": 2615
    },
    {
      "entropy": 5.774041032791137,
      "epoch": 0.2942662997697535,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499624539945111,
      "loss": 5.6842,
      "mean_token_accuracy": 0.17633943408727645,
      "num_tokens": 5845376.0,
      "step": 2620
    },
    {
      "entropy": 5.780922889709473,
      "epoch": 0.294827876677711,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.7233,
      "mean_token_accuracy": 0.16174683272838591,
      "num_tokens": 5857124.0,
      "step": 2625
    },
    {
      "entropy": 5.963794898986817,
      "epoch": 0.29538945358566854,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.8263,
      "mean_token_accuracy": 0.15918904915452003,
      "num_tokens": 5868681.0,
      "step": 2630
    },
    {
      "entropy": 5.872399187088012,
      "epoch": 0.2959510304936261,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.8141,
      "mean_token_accuracy": 0.16962093859910965,
      "num_tokens": 5880532.0,
      "step": 2635
    },
    {
      "entropy": 5.928860807418824,
      "epoch": 0.29651260740158364,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499615208964848,
      "loss": 5.7669,
      "mean_token_accuracy": 0.1708823785185814,
      "num_tokens": 5891582.0,
      "step": 2640
    },
    {
      "entropy": 5.85929856300354,
      "epoch": 0.2970741843095412,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.7778,
      "mean_token_accuracy": 0.163731949031353,
      "num_tokens": 5903543.0,
      "step": 2645
    },
    {
      "entropy": 5.830670642852783,
      "epoch": 0.29763576121749874,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.7276,
      "mean_token_accuracy": 0.1748351737856865,
      "num_tokens": 5913140.0,
      "step": 2650
    },
    {
      "entropy": 5.886608219146728,
      "epoch": 0.2981973381254563,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.7024,
      "mean_token_accuracy": 0.17049144208431244,
      "num_tokens": 5924010.0,
      "step": 2655
    },
    {
      "entropy": 5.901228237152099,
      "epoch": 0.29875891503341384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996057635552237,
      "loss": 5.9004,
      "mean_token_accuracy": 0.16316385865211486,
      "num_tokens": 5936246.0,
      "step": 2660
    },
    {
      "entropy": 5.900568342208862,
      "epoch": 0.2993204919413714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.7884,
      "mean_token_accuracy": 0.15582166761159896,
      "num_tokens": 5947594.0,
      "step": 2665
    },
    {
      "entropy": 5.927651834487915,
      "epoch": 0.29988206884932894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996009979409025,
      "loss": 5.7314,
      "mean_token_accuracy": 0.15971839278936387,
      "num_tokens": 5957660.0,
      "step": 2670
    },
    {
      "entropy": 5.685725259780884,
      "epoch": 0.30044364575728644,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.5661,
      "mean_token_accuracy": 0.17532341331243514,
      "num_tokens": 5968881.0,
      "step": 2675
    },
    {
      "entropy": 5.8464141368865965,
      "epoch": 0.301005222665244,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049959620372105,
      "loss": 5.7354,
      "mean_token_accuracy": 0.16859683096408845,
      "num_tokens": 5980032.0,
      "step": 2680
    },
    {
      "entropy": 5.720075225830078,
      "epoch": 0.30156679957320154,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.7348,
      "mean_token_accuracy": 0.16227123737335206,
      "num_tokens": 5991781.0,
      "step": 2685
    },
    {
      "entropy": 5.883849287033081,
      "epoch": 0.3021283764811591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.7686,
      "mean_token_accuracy": 0.16267271637916564,
      "num_tokens": 6002303.0,
      "step": 2690
    },
    {
      "entropy": 5.965817070007324,
      "epoch": 0.30268995338911664,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995889587572366,
      "loss": 5.8528,
      "mean_token_accuracy": 0.16787079870700836,
      "num_tokens": 6011523.0,
      "step": 2695
    },
    {
      "entropy": 5.912099599838257,
      "epoch": 0.3032515302970742,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499586529467197,
      "loss": 5.662,
      "mean_token_accuracy": 0.17042718827724457,
      "num_tokens": 6022169.0,
      "step": 2700
    },
    {
      "entropy": 5.723714351654053,
      "epoch": 0.30381310720503174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.7268,
      "mean_token_accuracy": 0.16638493165373802,
      "num_tokens": 6033393.0,
      "step": 2705
    },
    {
      "entropy": 5.8972358226776125,
      "epoch": 0.3043746841129893,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.7846,
      "mean_token_accuracy": 0.1596575379371643,
      "num_tokens": 6045365.0,
      "step": 2710
    },
    {
      "entropy": 5.826961803436279,
      "epoch": 0.30493626102094684,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.8057,
      "mean_token_accuracy": 0.16703701317310332,
      "num_tokens": 6057482.0,
      "step": 2715
    },
    {
      "entropy": 5.891117000579834,
      "epoch": 0.30549783792890434,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.7156,
      "mean_token_accuracy": 0.17238480895757674,
      "num_tokens": 6068194.0,
      "step": 2720
    },
    {
      "entropy": 5.850578546524048,
      "epoch": 0.3060594148368619,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.7649,
      "mean_token_accuracy": 0.17026786655187606,
      "num_tokens": 6080540.0,
      "step": 2725
    },
    {
      "entropy": 5.857915544509888,
      "epoch": 0.30662099174481944,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499571803560316,
      "loss": 5.8099,
      "mean_token_accuracy": 0.1647024117410183,
      "num_tokens": 6092464.0,
      "step": 2730
    },
    {
      "entropy": 5.915971326828003,
      "epoch": 0.307182568652777,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.7596,
      "mean_token_accuracy": 0.16778793632984162,
      "num_tokens": 6103242.0,
      "step": 2735
    },
    {
      "entropy": 5.825624752044678,
      "epoch": 0.30774414556073454,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.7974,
      "mean_token_accuracy": 0.16479281783103944,
      "num_tokens": 6116383.0,
      "step": 2740
    },
    {
      "entropy": 5.792344808578491,
      "epoch": 0.3083057224686921,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.6016,
      "mean_token_accuracy": 0.17332126349210739,
      "num_tokens": 6126883.0,
      "step": 2745
    },
    {
      "entropy": 5.887543678283691,
      "epoch": 0.30886729937664964,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.7993,
      "mean_token_accuracy": 0.16024126410484313,
      "num_tokens": 6139613.0,
      "step": 2750
    },
    {
      "entropy": 5.885481071472168,
      "epoch": 0.3094288762846072,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995593353330753,
      "loss": 5.8573,
      "mean_token_accuracy": 0.16386310905218124,
      "num_tokens": 6151950.0,
      "step": 2755
    },
    {
      "entropy": 5.814095878601075,
      "epoch": 0.30999045319256474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.6341,
      "mean_token_accuracy": 0.17681725472211837,
      "num_tokens": 6161563.0,
      "step": 2760
    },
    {
      "entropy": 5.80485954284668,
      "epoch": 0.3105520301005223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.6624,
      "mean_token_accuracy": 0.17834541499614714,
      "num_tokens": 6170966.0,
      "step": 2765
    },
    {
      "entropy": 5.818015003204346,
      "epoch": 0.3111136070084798,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.7308,
      "mean_token_accuracy": 0.16718895733356476,
      "num_tokens": 6182563.0,
      "step": 2770
    },
    {
      "entropy": 5.846281957626343,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.8416,
      "mean_token_accuracy": 0.16075171381235123,
      "num_tokens": 6194495.0,
      "step": 2775
    },
    {
      "entropy": 5.8640399932861325,
      "epoch": 0.3122367608243949,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.7408,
      "mean_token_accuracy": 0.16369348615407944,
      "num_tokens": 6206594.0,
      "step": 2780
    },
    {
      "entropy": 5.895187568664551,
      "epoch": 0.31279833773235244,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.7627,
      "mean_token_accuracy": 0.16485580503940583,
      "num_tokens": 6218371.0,
      "step": 2785
    },
    {
      "entropy": 5.882254314422608,
      "epoch": 0.31335991464031,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.6976,
      "mean_token_accuracy": 0.17356550693511963,
      "num_tokens": 6228723.0,
      "step": 2790
    },
    {
      "entropy": 5.886617183685303,
      "epoch": 0.31392149154826754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995390143707665,
      "loss": 5.8831,
      "mean_token_accuracy": 0.16673346012830734,
      "num_tokens": 6240535.0,
      "step": 2795
    },
    {
      "entropy": 5.822242879867554,
      "epoch": 0.3144830684562251,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.5854,
      "mean_token_accuracy": 0.17410123497247695,
      "num_tokens": 6250643.0,
      "step": 2800
    },
    {
      "entropy": 5.7944286346435545,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.7097,
      "mean_token_accuracy": 0.1716044783592224,
      "num_tokens": 6259569.0,
      "step": 2805
    },
    {
      "entropy": 5.858190250396729,
      "epoch": 0.3156062222721402,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499531276042413,
      "loss": 5.8734,
      "mean_token_accuracy": 0.16926339864730836,
      "num_tokens": 6271108.0,
      "step": 2810
    },
    {
      "entropy": 5.889335870742798,
      "epoch": 0.31616779918009774,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.7088,
      "mean_token_accuracy": 0.15809194445610047,
      "num_tokens": 6282352.0,
      "step": 2815
    },
    {
      "entropy": 5.8390350341796875,
      "epoch": 0.31672937608805524,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.8235,
      "mean_token_accuracy": 0.16358305364847184,
      "num_tokens": 6294946.0,
      "step": 2820
    },
    {
      "entropy": 5.754707479476929,
      "epoch": 0.3172909529960128,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.631,
      "mean_token_accuracy": 0.1745324283838272,
      "num_tokens": 6305552.0,
      "step": 2825
    },
    {
      "entropy": 5.797818899154663,
      "epoch": 0.31785252990397034,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.6993,
      "mean_token_accuracy": 0.1769035428762436,
      "num_tokens": 6315815.0,
      "step": 2830
    },
    {
      "entropy": 5.864303636550903,
      "epoch": 0.3184141068119279,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.7426,
      "mean_token_accuracy": 0.165255768597126,
      "num_tokens": 6327384.0,
      "step": 2835
    },
    {
      "entropy": 5.861385011672974,
      "epoch": 0.31897568371988544,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1629460111260414,
      "num_tokens": 6338083.0,
      "step": 2840
    },
    {
      "entropy": 5.876327419281006,
      "epoch": 0.319537260627843,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.7197,
      "mean_token_accuracy": 0.17031879872083663,
      "num_tokens": 6349058.0,
      "step": 2845
    },
    {
      "entropy": 5.823698043823242,
      "epoch": 0.32009883753580054,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.7282,
      "mean_token_accuracy": 0.17293349653482437,
      "num_tokens": 6359655.0,
      "step": 2850
    },
    {
      "entropy": 5.917513751983643,
      "epoch": 0.3206604144437581,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995076750115258,
      "loss": 5.8714,
      "mean_token_accuracy": 0.1634015515446663,
      "num_tokens": 6372088.0,
      "step": 2855
    },
    {
      "entropy": 5.820725727081299,
      "epoch": 0.32122199135171564,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.757,
      "mean_token_accuracy": 0.16435178220272065,
      "num_tokens": 6382946.0,
      "step": 2860
    },
    {
      "entropy": 5.829253482818603,
      "epoch": 0.32178356825967314,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.7626,
      "mean_token_accuracy": 0.1649421587586403,
      "num_tokens": 6392591.0,
      "step": 2865
    },
    {
      "entropy": 5.767510843276978,
      "epoch": 0.3223451451676307,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.6885,
      "mean_token_accuracy": 0.17057053446769715,
      "num_tokens": 6403159.0,
      "step": 2870
    },
    {
      "entropy": 5.859529638290406,
      "epoch": 0.32290672207558824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.7251,
      "mean_token_accuracy": 0.16644418686628343,
      "num_tokens": 6414232.0,
      "step": 2875
    },
    {
      "entropy": 5.8600890159606935,
      "epoch": 0.3234682989835458,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.6921,
      "mean_token_accuracy": 0.18009964227676392,
      "num_tokens": 6425762.0,
      "step": 2880
    },
    {
      "entropy": 5.778131580352783,
      "epoch": 0.32402987589150334,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.7359,
      "mean_token_accuracy": 0.16423172056674956,
      "num_tokens": 6439077.0,
      "step": 2885
    },
    {
      "entropy": 5.909442615509033,
      "epoch": 0.3245914527994609,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.8157,
      "mean_token_accuracy": 0.16724281013011932,
      "num_tokens": 6450425.0,
      "step": 2890
    },
    {
      "entropy": 5.829175615310669,
      "epoch": 0.32515302970741844,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.7707,
      "mean_token_accuracy": 0.1715887501835823,
      "num_tokens": 6461347.0,
      "step": 2895
    },
    {
      "entropy": 5.865241003036499,
      "epoch": 0.325714606615376,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.759,
      "mean_token_accuracy": 0.16879544407129288,
      "num_tokens": 6472428.0,
      "step": 2900
    },
    {
      "entropy": 5.7962500095367435,
      "epoch": 0.32627618352333354,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.7168,
      "mean_token_accuracy": 0.17446729838848113,
      "num_tokens": 6483393.0,
      "step": 2905
    },
    {
      "entropy": 5.80073094367981,
      "epoch": 0.3268377604312911,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.7014,
      "mean_token_accuracy": 0.17042361497879027,
      "num_tokens": 6493520.0,
      "step": 2910
    },
    {
      "entropy": 5.751598072052002,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.6518,
      "mean_token_accuracy": 0.17481115609407424,
      "num_tokens": 6505282.0,
      "step": 2915
    },
    {
      "entropy": 5.936983394622803,
      "epoch": 0.32796091424720614,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.816,
      "mean_token_accuracy": 0.1736079603433609,
      "num_tokens": 6516808.0,
      "step": 2920
    },
    {
      "entropy": 5.871405839920044,
      "epoch": 0.3285224911551637,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.7754,
      "mean_token_accuracy": 0.1640280917286873,
      "num_tokens": 6529184.0,
      "step": 2925
    },
    {
      "entropy": 5.744284677505493,
      "epoch": 0.32908406806312124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.6271,
      "mean_token_accuracy": 0.17572139203548431,
      "num_tokens": 6539909.0,
      "step": 2930
    },
    {
      "entropy": 5.8266712665557865,
      "epoch": 0.3296456449710788,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.6743,
      "mean_token_accuracy": 0.1735612004995346,
      "num_tokens": 6550870.0,
      "step": 2935
    },
    {
      "entropy": 5.756214761734009,
      "epoch": 0.33020722187903634,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499461515538303,
      "loss": 5.7291,
      "mean_token_accuracy": 0.17583443224430084,
      "num_tokens": 6561784.0,
      "step": 2940
    },
    {
      "entropy": 5.757110977172852,
      "epoch": 0.3307687987869939,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.6696,
      "mean_token_accuracy": 0.17046401351690293,
      "num_tokens": 6573422.0,
      "step": 2945
    },
    {
      "entropy": 5.896243619918823,
      "epoch": 0.33133037569495144,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499455949211845,
      "loss": 5.7409,
      "mean_token_accuracy": 0.17513602674007417,
      "num_tokens": 6584565.0,
      "step": 2950
    },
    {
      "entropy": 5.869036149978638,
      "epoch": 0.331891952602909,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.6615,
      "mean_token_accuracy": 0.16997261941432953,
      "num_tokens": 6594238.0,
      "step": 2955
    },
    {
      "entropy": 5.723050689697265,
      "epoch": 0.3324535295108665,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.741,
      "mean_token_accuracy": 0.16911158263683318,
      "num_tokens": 6606309.0,
      "step": 2960
    },
    {
      "entropy": 5.956376457214356,
      "epoch": 0.33301510641882404,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.8312,
      "mean_token_accuracy": 0.1619592562317848,
      "num_tokens": 6617916.0,
      "step": 2965
    },
    {
      "entropy": 5.807530403137207,
      "epoch": 0.3335766833267816,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.6867,
      "mean_token_accuracy": 0.16937974691390992,
      "num_tokens": 6627951.0,
      "step": 2970
    },
    {
      "entropy": 5.740665912628174,
      "epoch": 0.33413826023473914,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.7516,
      "mean_token_accuracy": 0.17015371173620225,
      "num_tokens": 6639285.0,
      "step": 2975
    },
    {
      "entropy": 5.817653036117553,
      "epoch": 0.3346998371426967,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.6058,
      "mean_token_accuracy": 0.1782895877957344,
      "num_tokens": 6649973.0,
      "step": 2980
    },
    {
      "entropy": 5.73169355392456,
      "epoch": 0.33526141405065424,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.6212,
      "mean_token_accuracy": 0.17402680665254594,
      "num_tokens": 6660625.0,
      "step": 2985
    },
    {
      "entropy": 5.835042095184326,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049943339804258,
      "loss": 5.7113,
      "mean_token_accuracy": 0.16824356615543365,
      "num_tokens": 6673191.0,
      "step": 2990
    },
    {
      "entropy": 5.834809255599976,
      "epoch": 0.33638456786656934,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.716,
      "mean_token_accuracy": 0.17247826904058455,
      "num_tokens": 6683993.0,
      "step": 2995
    },
    {
      "entropy": 5.763989877700806,
      "epoch": 0.3369461447745269,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.7292,
      "mean_token_accuracy": 0.1751634359359741,
      "num_tokens": 6694881.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.631992234315398,
      "eval_loss": 5.732239246368408,
      "eval_mean_token_accuracy": 0.17665448291421798,
      "eval_num_tokens": 6694881.0,
      "eval_runtime": 18.4364,
      "eval_samples_per_second": 1682.164,
      "eval_steps_per_second": 210.291,
      "step": 3000
    },
    {
      "entropy": 5.819136333465576,
      "epoch": 0.33750772168248444,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.733,
      "mean_token_accuracy": 0.16972203850746154,
      "num_tokens": 6706648.0,
      "step": 3005
    },
    {
      "entropy": 5.801159763336182,
      "epoch": 0.33806929859044194,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499421950949953,
      "loss": 5.7347,
      "mean_token_accuracy": 0.17212952226400374,
      "num_tokens": 6716858.0,
      "step": 3010
    },
    {
      "entropy": 5.724242496490478,
      "epoch": 0.3386308754983995,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499419071312363,
      "loss": 5.6889,
      "mean_token_accuracy": 0.17499449700117112,
      "num_tokens": 6726558.0,
      "step": 3015
    },
    {
      "entropy": 5.881882572174073,
      "epoch": 0.33919245240635704,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.6801,
      "mean_token_accuracy": 0.16776044964790343,
      "num_tokens": 6737819.0,
      "step": 3020
    },
    {
      "entropy": 5.794957828521729,
      "epoch": 0.3397540293143146,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.6322,
      "mean_token_accuracy": 0.1724249988794327,
      "num_tokens": 6748593.0,
      "step": 3025
    },
    {
      "entropy": 5.63336067199707,
      "epoch": 0.34031560622227214,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.7422,
      "mean_token_accuracy": 0.162245012819767,
      "num_tokens": 6760631.0,
      "step": 3030
    },
    {
      "entropy": 5.919184780120849,
      "epoch": 0.3408771831302297,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.6398,
      "mean_token_accuracy": 0.16651962250471114,
      "num_tokens": 6772318.0,
      "step": 3035
    },
    {
      "entropy": 5.715067386627197,
      "epoch": 0.34143876003818724,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.6772,
      "mean_token_accuracy": 0.1750547170639038,
      "num_tokens": 6782684.0,
      "step": 3040
    },
    {
      "entropy": 5.785168123245239,
      "epoch": 0.3420003369461448,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.5655,
      "mean_token_accuracy": 0.1777344137430191,
      "num_tokens": 6793895.0,
      "step": 3045
    },
    {
      "entropy": 5.684515047073364,
      "epoch": 0.34256191385410234,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.6293,
      "mean_token_accuracy": 0.17251270562410354,
      "num_tokens": 6804794.0,
      "step": 3050
    },
    {
      "entropy": 5.8218306541442875,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499395776978143,
      "loss": 5.7589,
      "mean_token_accuracy": 0.16612927317619325,
      "num_tokens": 6815976.0,
      "step": 3055
    },
    {
      "entropy": 5.817735004425049,
      "epoch": 0.3436850676700174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.742,
      "mean_token_accuracy": 0.16376831382513046,
      "num_tokens": 6828027.0,
      "step": 3060
    },
    {
      "entropy": 5.815735101699829,
      "epoch": 0.34424664457797494,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.7233,
      "mean_token_accuracy": 0.1691370740532875,
      "num_tokens": 6839195.0,
      "step": 3065
    },
    {
      "entropy": 5.656938266754151,
      "epoch": 0.3448082214859325,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.5894,
      "mean_token_accuracy": 0.17244611978530883,
      "num_tokens": 6851731.0,
      "step": 3070
    },
    {
      "entropy": 5.8333851337432865,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.6397,
      "mean_token_accuracy": 0.17510228455066681,
      "num_tokens": 6862916.0,
      "step": 3075
    },
    {
      "entropy": 5.888251066207886,
      "epoch": 0.3459313753018476,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.8157,
      "mean_token_accuracy": 0.162273408472538,
      "num_tokens": 6876003.0,
      "step": 3080
    },
    {
      "entropy": 5.739641332626343,
      "epoch": 0.34649295220980514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.7806,
      "mean_token_accuracy": 0.1702667608857155,
      "num_tokens": 6886040.0,
      "step": 3085
    },
    {
      "entropy": 5.829275894165039,
      "epoch": 0.3470545291177627,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.6865,
      "mean_token_accuracy": 0.1768171027302742,
      "num_tokens": 6896758.0,
      "step": 3090
    },
    {
      "entropy": 5.833682680130005,
      "epoch": 0.34761610602572024,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.7384,
      "mean_token_accuracy": 0.17008273154497147,
      "num_tokens": 6907287.0,
      "step": 3095
    },
    {
      "entropy": 5.768305683135987,
      "epoch": 0.3481776829336778,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.7566,
      "mean_token_accuracy": 0.16723249703645707,
      "num_tokens": 6918992.0,
      "step": 3100
    },
    {
      "entropy": 5.890691232681275,
      "epoch": 0.3487392598416353,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.77,
      "mean_token_accuracy": 0.1689866676926613,
      "num_tokens": 6931493.0,
      "step": 3105
    },
    {
      "entropy": 5.7417604446411135,
      "epoch": 0.34930083674959284,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.5921,
      "mean_token_accuracy": 0.1775926724076271,
      "num_tokens": 6941807.0,
      "step": 3110
    },
    {
      "entropy": 5.861759185791016,
      "epoch": 0.3498624136575504,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.7927,
      "mean_token_accuracy": 0.16744445115327836,
      "num_tokens": 6953436.0,
      "step": 3115
    },
    {
      "entropy": 5.968575429916382,
      "epoch": 0.35042399056550794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993569484348359,
      "loss": 5.9075,
      "mean_token_accuracy": 0.16151681840419768,
      "num_tokens": 6965247.0,
      "step": 3120
    },
    {
      "entropy": 5.7765837669372555,
      "epoch": 0.3509855674734655,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.6958,
      "mean_token_accuracy": 0.17401523888111115,
      "num_tokens": 6976605.0,
      "step": 3125
    },
    {
      "entropy": 5.710376501083374,
      "epoch": 0.35154714438142304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.6508,
      "mean_token_accuracy": 0.18309797793626786,
      "num_tokens": 6986205.0,
      "step": 3130
    },
    {
      "entropy": 5.774714517593384,
      "epoch": 0.3521087212893806,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.6054,
      "mean_token_accuracy": 0.178724704682827,
      "num_tokens": 6997699.0,
      "step": 3135
    },
    {
      "entropy": 5.690198755264282,
      "epoch": 0.35267029819733814,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.62,
      "mean_token_accuracy": 0.1739739567041397,
      "num_tokens": 7008968.0,
      "step": 3140
    },
    {
      "entropy": 5.8412824153900145,
      "epoch": 0.3532318751052957,
      "grad_norm": 1.0,
      "learning_rate": 0.000499341692905686,
      "loss": 5.6581,
      "mean_token_accuracy": 0.1689726173877716,
      "num_tokens": 7020633.0,
      "step": 3145
    },
    {
      "entropy": 5.73580117225647,
      "epoch": 0.35379345201325324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.579,
      "mean_token_accuracy": 0.18208679258823396,
      "num_tokens": 7031187.0,
      "step": 3150
    },
    {
      "entropy": 5.816767406463623,
      "epoch": 0.35435502892121074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.7948,
      "mean_token_accuracy": 0.16400412023067473,
      "num_tokens": 7042662.0,
      "step": 3155
    },
    {
      "entropy": 5.848151445388794,
      "epoch": 0.3549166058291683,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.7857,
      "mean_token_accuracy": 0.173075395822525,
      "num_tokens": 7054297.0,
      "step": 3160
    },
    {
      "entropy": 5.826444005966186,
      "epoch": 0.35547818273712584,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.6413,
      "mean_token_accuracy": 0.17617398798465728,
      "num_tokens": 7064961.0,
      "step": 3165
    },
    {
      "entropy": 5.798005485534668,
      "epoch": 0.3560397596450834,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.7994,
      "mean_token_accuracy": 0.1694578006863594,
      "num_tokens": 7076857.0,
      "step": 3170
    },
    {
      "entropy": 5.838046503067017,
      "epoch": 0.35660133655304094,
      "grad_norm": 1.0,
      "learning_rate": 0.000499323150548625,
      "loss": 5.6704,
      "mean_token_accuracy": 0.17216843664646148,
      "num_tokens": 7088926.0,
      "step": 3175
    },
    {
      "entropy": 5.656442022323608,
      "epoch": 0.3571629134609985,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.5703,
      "mean_token_accuracy": 0.18148448467254638,
      "num_tokens": 7099478.0,
      "step": 3180
    },
    {
      "entropy": 5.744642448425293,
      "epoch": 0.35772449036895604,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.6312,
      "mean_token_accuracy": 0.1722390189766884,
      "num_tokens": 7111776.0,
      "step": 3185
    },
    {
      "entropy": 5.88515830039978,
      "epoch": 0.3582860672769136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.7242,
      "mean_token_accuracy": 0.16789288222789764,
      "num_tokens": 7123084.0,
      "step": 3190
    },
    {
      "entropy": 5.764853668212891,
      "epoch": 0.35884764418487114,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.704,
      "mean_token_accuracy": 0.171440427005291,
      "num_tokens": 7134973.0,
      "step": 3195
    },
    {
      "entropy": 5.692889022827148,
      "epoch": 0.3594092210928287,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.7244,
      "mean_token_accuracy": 0.16945530772209166,
      "num_tokens": 7147227.0,
      "step": 3200
    },
    {
      "entropy": 5.874735164642334,
      "epoch": 0.3599707980007862,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.7013,
      "mean_token_accuracy": 0.17460730224847792,
      "num_tokens": 7158201.0,
      "step": 3205
    },
    {
      "entropy": 5.709667730331421,
      "epoch": 0.36053237490874374,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499301192814866,
      "loss": 5.5983,
      "mean_token_accuracy": 0.17740452587604522,
      "num_tokens": 7168909.0,
      "step": 3210
    },
    {
      "entropy": 5.801937246322632,
      "epoch": 0.3610939518167013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.753,
      "mean_token_accuracy": 0.16828656196594238,
      "num_tokens": 7179877.0,
      "step": 3215
    },
    {
      "entropy": 5.793533849716186,
      "epoch": 0.36165552872465884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.6603,
      "mean_token_accuracy": 0.1773525133728981,
      "num_tokens": 7190823.0,
      "step": 3220
    },
    {
      "entropy": 5.752917909622193,
      "epoch": 0.3622171056326164,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.5894,
      "mean_token_accuracy": 0.17470155656337738,
      "num_tokens": 7201700.0,
      "step": 3225
    },
    {
      "entropy": 5.780609560012818,
      "epoch": 0.36277868254057394,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499288488415819,
      "loss": 5.6493,
      "mean_token_accuracy": 0.17959498316049577,
      "num_tokens": 7213618.0,
      "step": 3230
    },
    {
      "entropy": 5.866898870468139,
      "epoch": 0.3633402594485315,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.8387,
      "mean_token_accuracy": 0.16976505815982817,
      "num_tokens": 7226207.0,
      "step": 3235
    },
    {
      "entropy": 5.7547344207763675,
      "epoch": 0.36390183635648904,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.6698,
      "mean_token_accuracy": 0.1762295573949814,
      "num_tokens": 7237545.0,
      "step": 3240
    },
    {
      "entropy": 5.717173767089844,
      "epoch": 0.3644634132644466,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.6514,
      "mean_token_accuracy": 0.1721198230981827,
      "num_tokens": 7247799.0,
      "step": 3245
    },
    {
      "entropy": 5.746833848953247,
      "epoch": 0.3650249901724041,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.6435,
      "mean_token_accuracy": 0.16833308786153794,
      "num_tokens": 7260229.0,
      "step": 3250
    },
    {
      "entropy": 5.735746622085571,
      "epoch": 0.36558656708036164,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.7514,
      "mean_token_accuracy": 0.1684183195233345,
      "num_tokens": 7271457.0,
      "step": 3255
    },
    {
      "entropy": 5.780439186096191,
      "epoch": 0.3661481439883192,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.65,
      "mean_token_accuracy": 0.1729257583618164,
      "num_tokens": 7282124.0,
      "step": 3260
    },
    {
      "entropy": 5.824368476867676,
      "epoch": 0.36670972089627674,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.7521,
      "mean_token_accuracy": 0.16994550675153733,
      "num_tokens": 7292981.0,
      "step": 3265
    },
    {
      "entropy": 5.935243320465088,
      "epoch": 0.3672712978042343,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992627368355112,
      "loss": 5.9006,
      "mean_token_accuracy": 0.16785538494586943,
      "num_tokens": 7304591.0,
      "step": 3270
    },
    {
      "entropy": 5.849437046051025,
      "epoch": 0.36783287471219184,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.6396,
      "mean_token_accuracy": 0.17560438215732574,
      "num_tokens": 7315918.0,
      "step": 3275
    },
    {
      "entropy": 5.777096176147461,
      "epoch": 0.3683944516201494,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.7355,
      "mean_token_accuracy": 0.16889662742614747,
      "num_tokens": 7327725.0,
      "step": 3280
    },
    {
      "entropy": 5.88145899772644,
      "epoch": 0.36895602852810694,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.7405,
      "mean_token_accuracy": 0.16651637107133865,
      "num_tokens": 7339787.0,
      "step": 3285
    },
    {
      "entropy": 5.858824682235718,
      "epoch": 0.3695176054360645,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.7294,
      "mean_token_accuracy": 0.17076862901449202,
      "num_tokens": 7351284.0,
      "step": 3290
    },
    {
      "entropy": 5.658685493469238,
      "epoch": 0.37007918234402204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.5524,
      "mean_token_accuracy": 0.17875730544328688,
      "num_tokens": 7362105.0,
      "step": 3295
    },
    {
      "entropy": 5.771422719955444,
      "epoch": 0.37064075925197953,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.6694,
      "mean_token_accuracy": 0.1705271676182747,
      "num_tokens": 7375204.0,
      "step": 3300
    },
    {
      "entropy": 5.90152678489685,
      "epoch": 0.3712023361599371,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.8067,
      "mean_token_accuracy": 0.1683753490447998,
      "num_tokens": 7385853.0,
      "step": 3305
    },
    {
      "entropy": 5.754110288619995,
      "epoch": 0.37176391306789464,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.6737,
      "mean_token_accuracy": 0.16615278273820877,
      "num_tokens": 7397459.0,
      "step": 3310
    },
    {
      "entropy": 5.775053882598877,
      "epoch": 0.3723254899758522,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.6533,
      "mean_token_accuracy": 0.16877304911613464,
      "num_tokens": 7408996.0,
      "step": 3315
    },
    {
      "entropy": 5.84871244430542,
      "epoch": 0.37288706688380974,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.7501,
      "mean_token_accuracy": 0.16261120289564132,
      "num_tokens": 7419938.0,
      "step": 3320
    },
    {
      "entropy": 5.813049745559693,
      "epoch": 0.3734486437917673,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.7612,
      "mean_token_accuracy": 0.16559012234210968,
      "num_tokens": 7430181.0,
      "step": 3325
    },
    {
      "entropy": 5.810550117492676,
      "epoch": 0.37401022069972484,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.7099,
      "mean_token_accuracy": 0.16987437307834624,
      "num_tokens": 7440774.0,
      "step": 3330
    },
    {
      "entropy": 5.818269824981689,
      "epoch": 0.3745717976076824,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499219915847757,
      "loss": 5.6408,
      "mean_token_accuracy": 0.17308392822742463,
      "num_tokens": 7451744.0,
      "step": 3335
    },
    {
      "entropy": 5.647943449020386,
      "epoch": 0.37513337451563994,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.6826,
      "mean_token_accuracy": 0.17259350717067717,
      "num_tokens": 7463731.0,
      "step": 3340
    },
    {
      "entropy": 5.836802339553833,
      "epoch": 0.37569495142359743,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.6712,
      "mean_token_accuracy": 0.1763611614704132,
      "num_tokens": 7474392.0,
      "step": 3345
    },
    {
      "entropy": 5.715862989425659,
      "epoch": 0.376256528331555,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.5792,
      "mean_token_accuracy": 0.18714895993471145,
      "num_tokens": 7484860.0,
      "step": 3350
    },
    {
      "entropy": 5.692795515060425,
      "epoch": 0.37681810523951254,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499206497413985,
      "loss": 5.6433,
      "mean_token_accuracy": 0.17612822651863097,
      "num_tokens": 7495877.0,
      "step": 3355
    },
    {
      "entropy": 5.870680618286133,
      "epoch": 0.3773796821474701,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.7041,
      "mean_token_accuracy": 0.17734302431344987,
      "num_tokens": 7506879.0,
      "step": 3360
    },
    {
      "entropy": 5.706663513183594,
      "epoch": 0.37794125905542764,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.5412,
      "mean_token_accuracy": 0.18161289244890214,
      "num_tokens": 7517405.0,
      "step": 3365
    },
    {
      "entropy": 5.715374231338501,
      "epoch": 0.3785028359633852,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.6328,
      "mean_token_accuracy": 0.1720275640487671,
      "num_tokens": 7528832.0,
      "step": 3370
    },
    {
      "entropy": 5.757056188583374,
      "epoch": 0.37906441287134274,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.6105,
      "mean_token_accuracy": 0.17205780744552612,
      "num_tokens": 7541048.0,
      "step": 3375
    },
    {
      "entropy": 5.754649496078491,
      "epoch": 0.3796259897793003,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.6185,
      "mean_token_accuracy": 0.17652998715639115,
      "num_tokens": 7551849.0,
      "step": 3380
    },
    {
      "entropy": 5.662470054626465,
      "epoch": 0.38018756668725784,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.6357,
      "mean_token_accuracy": 0.17981449663639068,
      "num_tokens": 7563148.0,
      "step": 3385
    },
    {
      "entropy": 5.894662761688233,
      "epoch": 0.3807491435952154,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.8225,
      "mean_token_accuracy": 0.1672139585018158,
      "num_tokens": 7575119.0,
      "step": 3390
    },
    {
      "entropy": 5.8430053234100345,
      "epoch": 0.3813107205031729,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.6704,
      "mean_token_accuracy": 0.17408028542995452,
      "num_tokens": 7585899.0,
      "step": 3395
    },
    {
      "entropy": 5.792122459411621,
      "epoch": 0.38187229741113043,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.5733,
      "mean_token_accuracy": 0.18275393694639205,
      "num_tokens": 7598938.0,
      "step": 3400
    },
    {
      "entropy": 5.628747367858887,
      "epoch": 0.382433874319088,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.6521,
      "mean_token_accuracy": 0.16920132786035538,
      "num_tokens": 7610087.0,
      "step": 3405
    },
    {
      "entropy": 5.887259387969971,
      "epoch": 0.38299545122704554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.7794,
      "mean_token_accuracy": 0.16809322983026503,
      "num_tokens": 7622186.0,
      "step": 3410
    },
    {
      "entropy": 5.874405384063721,
      "epoch": 0.3835570281350031,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.7338,
      "mean_token_accuracy": 0.17035729736089705,
      "num_tokens": 7633501.0,
      "step": 3415
    },
    {
      "entropy": 5.654582500457764,
      "epoch": 0.38411860504296064,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.6487,
      "mean_token_accuracy": 0.1724038079380989,
      "num_tokens": 7643656.0,
      "step": 3420
    },
    {
      "entropy": 5.780475711822509,
      "epoch": 0.3846801819509182,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.6597,
      "mean_token_accuracy": 0.17520257979631423,
      "num_tokens": 7654565.0,
      "step": 3425
    },
    {
      "entropy": 5.750225830078125,
      "epoch": 0.38524175885887574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.5788,
      "mean_token_accuracy": 0.1814594268798828,
      "num_tokens": 7664725.0,
      "step": 3430
    },
    {
      "entropy": 5.658280611038208,
      "epoch": 0.3858033357668333,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.6667,
      "mean_token_accuracy": 0.17267031818628312,
      "num_tokens": 7674772.0,
      "step": 3435
    },
    {
      "entropy": 5.855649948120117,
      "epoch": 0.38636491267479084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499148194874525,
      "loss": 5.7189,
      "mean_token_accuracy": 0.16851244121789932,
      "num_tokens": 7686488.0,
      "step": 3440
    },
    {
      "entropy": 5.850823068618775,
      "epoch": 0.38692648958274833,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.7083,
      "mean_token_accuracy": 0.1727656215429306,
      "num_tokens": 7698842.0,
      "step": 3445
    },
    {
      "entropy": 5.685489082336426,
      "epoch": 0.3874880664907059,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.4988,
      "mean_token_accuracy": 0.17627423107624055,
      "num_tokens": 7709255.0,
      "step": 3450
    },
    {
      "entropy": 5.605527257919311,
      "epoch": 0.38804964339866344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.5555,
      "mean_token_accuracy": 0.17667909264564513,
      "num_tokens": 7720182.0,
      "step": 3455
    },
    {
      "entropy": 5.701817989349365,
      "epoch": 0.388611220306621,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.5803,
      "mean_token_accuracy": 0.18403764963150024,
      "num_tokens": 7730944.0,
      "step": 3460
    },
    {
      "entropy": 5.778051900863647,
      "epoch": 0.38917279721457854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049913065451149,
      "loss": 5.6454,
      "mean_token_accuracy": 0.1756534218788147,
      "num_tokens": 7741686.0,
      "step": 3465
    },
    {
      "entropy": 5.7675353527069095,
      "epoch": 0.3897343741225361,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.6537,
      "mean_token_accuracy": 0.17629104107618332,
      "num_tokens": 7752245.0,
      "step": 3470
    },
    {
      "entropy": 5.637860631942749,
      "epoch": 0.39029595103049364,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.6335,
      "mean_token_accuracy": 0.16780858039855956,
      "num_tokens": 7762622.0,
      "step": 3475
    },
    {
      "entropy": 5.866631984710693,
      "epoch": 0.3908575279384512,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.7482,
      "mean_token_accuracy": 0.17427458316087724,
      "num_tokens": 7773550.0,
      "step": 3480
    },
    {
      "entropy": 5.693408346176147,
      "epoch": 0.39141910484640874,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.6254,
      "mean_token_accuracy": 0.17077690064907075,
      "num_tokens": 7784988.0,
      "step": 3485
    },
    {
      "entropy": 5.7872772216796875,
      "epoch": 0.39198068175436623,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.7126,
      "mean_token_accuracy": 0.17314721494913102,
      "num_tokens": 7797706.0,
      "step": 3490
    },
    {
      "entropy": 5.722473287582398,
      "epoch": 0.3925422586623238,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.5935,
      "mean_token_accuracy": 0.17936303317546845,
      "num_tokens": 7808960.0,
      "step": 3495
    },
    {
      "entropy": 5.681925201416016,
      "epoch": 0.39310383557028133,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.5715,
      "mean_token_accuracy": 0.18486556261777878,
      "num_tokens": 7819271.0,
      "step": 3500
    },
    {
      "entropy": 5.7344615936279295,
      "epoch": 0.3936654124782389,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.6962,
      "mean_token_accuracy": 0.17472131848335265,
      "num_tokens": 7832257.0,
      "step": 3505
    },
    {
      "entropy": 5.852534914016724,
      "epoch": 0.39422698938619644,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049909863227389,
      "loss": 5.7017,
      "mean_token_accuracy": 0.16729053854942322,
      "num_tokens": 7842684.0,
      "step": 3510
    },
    {
      "entropy": 5.770964336395264,
      "epoch": 0.394788566294154,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.719,
      "mean_token_accuracy": 0.16776077896356584,
      "num_tokens": 7855031.0,
      "step": 3515
    },
    {
      "entropy": 5.674540424346924,
      "epoch": 0.39535014320211154,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.5148,
      "mean_token_accuracy": 0.1746792569756508,
      "num_tokens": 7865291.0,
      "step": 3520
    },
    {
      "entropy": 5.6400972366333,
      "epoch": 0.3959117201100691,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.4908,
      "mean_token_accuracy": 0.17839647978544235,
      "num_tokens": 7876401.0,
      "step": 3525
    },
    {
      "entropy": 5.634608030319214,
      "epoch": 0.39647329701802664,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.536,
      "mean_token_accuracy": 0.17374856024980545,
      "num_tokens": 7887787.0,
      "step": 3530
    },
    {
      "entropy": 5.707843494415283,
      "epoch": 0.3970348739259842,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.523,
      "mean_token_accuracy": 0.1763603776693344,
      "num_tokens": 7898550.0,
      "step": 3535
    },
    {
      "entropy": 5.68929762840271,
      "epoch": 0.3975964508339417,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.6747,
      "mean_token_accuracy": 0.17355783879756928,
      "num_tokens": 7909254.0,
      "step": 3540
    },
    {
      "entropy": 5.852826881408691,
      "epoch": 0.39815802774189923,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990733265032,
      "loss": 5.7042,
      "mean_token_accuracy": 0.17926168888807298,
      "num_tokens": 7920164.0,
      "step": 3545
    },
    {
      "entropy": 5.705505084991455,
      "epoch": 0.3987196046498568,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.6434,
      "mean_token_accuracy": 0.17162723541259767,
      "num_tokens": 7931390.0,
      "step": 3550
    },
    {
      "entropy": 5.765172052383423,
      "epoch": 0.39928118155781434,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499066032070036,
      "loss": 5.6634,
      "mean_token_accuracy": 0.1701713427901268,
      "num_tokens": 7943765.0,
      "step": 3555
    },
    {
      "entropy": 5.695808792114258,
      "epoch": 0.3998427584657719,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.5016,
      "mean_token_accuracy": 0.1803964599967003,
      "num_tokens": 7954621.0,
      "step": 3560
    },
    {
      "entropy": 5.671160173416138,
      "epoch": 0.40040433537372944,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499058709099475,
      "loss": 5.7128,
      "mean_token_accuracy": 0.16859437972307206,
      "num_tokens": 7966063.0,
      "step": 3565
    },
    {
      "entropy": 5.835380983352661,
      "epoch": 0.400965912281687,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.6232,
      "mean_token_accuracy": 0.17717592418193817,
      "num_tokens": 7977391.0,
      "step": 3570
    },
    {
      "entropy": 5.743214464187622,
      "epoch": 0.40152748918964454,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.7367,
      "mean_token_accuracy": 0.16896894723176956,
      "num_tokens": 7988786.0,
      "step": 3575
    },
    {
      "entropy": 5.742875957489014,
      "epoch": 0.4020890660976021,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.5241,
      "mean_token_accuracy": 0.17627668231725693,
      "num_tokens": 8000848.0,
      "step": 3580
    },
    {
      "entropy": 5.750096702575684,
      "epoch": 0.40265064300555964,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.6817,
      "mean_token_accuracy": 0.17042557001113892,
      "num_tokens": 8012621.0,
      "step": 3585
    },
    {
      "entropy": 5.833896017074585,
      "epoch": 0.40321221991351713,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499040276828089,
      "loss": 5.7354,
      "mean_token_accuracy": 0.16869859397411346,
      "num_tokens": 8023599.0,
      "step": 3590
    },
    {
      "entropy": 5.644566774368286,
      "epoch": 0.4037737968214747,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.5444,
      "mean_token_accuracy": 0.17929765433073044,
      "num_tokens": 8037151.0,
      "step": 3595
    },
    {
      "entropy": 5.742999458312989,
      "epoch": 0.40433537372943223,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.7393,
      "mean_token_accuracy": 0.17098934650421144,
      "num_tokens": 8048606.0,
      "step": 3600
    },
    {
      "entropy": 5.707454442977905,
      "epoch": 0.4048969506373898,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.531,
      "mean_token_accuracy": 0.18790849000215532,
      "num_tokens": 8058430.0,
      "step": 3605
    },
    {
      "entropy": 5.626982164382935,
      "epoch": 0.40545852754534734,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.5427,
      "mean_token_accuracy": 0.17823073863983155,
      "num_tokens": 8069554.0,
      "step": 3610
    },
    {
      "entropy": 5.806381511688232,
      "epoch": 0.4060201044533049,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.7041,
      "mean_token_accuracy": 0.1629281982779503,
      "num_tokens": 8082131.0,
      "step": 3615
    },
    {
      "entropy": 5.690978956222534,
      "epoch": 0.40658168136126244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.5306,
      "mean_token_accuracy": 0.1793044000864029,
      "num_tokens": 8093454.0,
      "step": 3620
    },
    {
      "entropy": 5.816726446151733,
      "epoch": 0.40714325826922,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.7961,
      "mean_token_accuracy": 0.1596296101808548,
      "num_tokens": 8105502.0,
      "step": 3625
    },
    {
      "entropy": 5.728281879425049,
      "epoch": 0.40770483517717754,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.6486,
      "mean_token_accuracy": 0.17166142165660858,
      "num_tokens": 8116524.0,
      "step": 3630
    },
    {
      "entropy": 5.749354457855224,
      "epoch": 0.40826641208513503,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499006649337227,
      "loss": 5.6378,
      "mean_token_accuracy": 0.18110090345144272,
      "num_tokens": 8126949.0,
      "step": 3635
    },
    {
      "entropy": 5.751841020584107,
      "epoch": 0.4088279889930926,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.5948,
      "mean_token_accuracy": 0.18108304888010024,
      "num_tokens": 8137346.0,
      "step": 3640
    },
    {
      "entropy": 5.655619096755982,
      "epoch": 0.40938956590105013,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.6204,
      "mean_token_accuracy": 0.17147953510284425,
      "num_tokens": 8148499.0,
      "step": 3645
    },
    {
      "entropy": 5.795197439193726,
      "epoch": 0.4099511428090077,
      "grad_norm": 1.0,
      "learning_rate": 0.000498995311785053,
      "loss": 5.6606,
      "mean_token_accuracy": 0.17419218868017197,
      "num_tokens": 8159645.0,
      "step": 3650
    },
    {
      "entropy": 5.71441879272461,
      "epoch": 0.41051271971696524,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.5503,
      "mean_token_accuracy": 0.176376548409462,
      "num_tokens": 8171033.0,
      "step": 3655
    },
    {
      "entropy": 5.646851682662964,
      "epoch": 0.4110742966249228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.6198,
      "mean_token_accuracy": 0.17579082101583482,
      "num_tokens": 8181582.0,
      "step": 3660
    },
    {
      "entropy": 5.713270759582519,
      "epoch": 0.41163587353288034,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498983910044002,
      "loss": 5.5233,
      "mean_token_accuracy": 0.17835837006568908,
      "num_tokens": 8192432.0,
      "step": 3665
    },
    {
      "entropy": 5.781804943084717,
      "epoch": 0.4121974504408379,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.7125,
      "mean_token_accuracy": 0.17426385283470153,
      "num_tokens": 8203995.0,
      "step": 3670
    },
    {
      "entropy": 5.669413900375366,
      "epoch": 0.41275902734879544,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.5416,
      "mean_token_accuracy": 0.1784492939710617,
      "num_tokens": 8214323.0,
      "step": 3675
    },
    {
      "entropy": 5.635898303985596,
      "epoch": 0.413320604256753,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.6093,
      "mean_token_accuracy": 0.1708337113261223,
      "num_tokens": 8226285.0,
      "step": 3680
    },
    {
      "entropy": 5.830092430114746,
      "epoch": 0.4138821811647105,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.7605,
      "mean_token_accuracy": 0.16820843517780304,
      "num_tokens": 8237961.0,
      "step": 3685
    },
    {
      "entropy": 5.674856519699096,
      "epoch": 0.41444375807266803,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.519,
      "mean_token_accuracy": 0.18347225934267045,
      "num_tokens": 8248463.0,
      "step": 3690
    },
    {
      "entropy": 5.76221284866333,
      "epoch": 0.4150053349806256,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.6666,
      "mean_token_accuracy": 0.17800566107034682,
      "num_tokens": 8260342.0,
      "step": 3695
    },
    {
      "entropy": 5.7744965076446535,
      "epoch": 0.41556691188858313,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.7668,
      "mean_token_accuracy": 0.17411787509918214,
      "num_tokens": 8272055.0,
      "step": 3700
    },
    {
      "entropy": 5.876493167877197,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.701,
      "mean_token_accuracy": 0.17120005339384078,
      "num_tokens": 8282129.0,
      "step": 3705
    },
    {
      "entropy": 5.768954277038574,
      "epoch": 0.41669006570449824,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.7052,
      "mean_token_accuracy": 0.1731296017765999,
      "num_tokens": 8294663.0,
      "step": 3710
    },
    {
      "entropy": 5.672701168060303,
      "epoch": 0.4172516426124558,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498945440696253,
      "loss": 5.5249,
      "mean_token_accuracy": 0.17951664328575134,
      "num_tokens": 8305882.0,
      "step": 3715
    },
    {
      "entropy": 5.714199781417847,
      "epoch": 0.41781321952041334,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.6191,
      "mean_token_accuracy": 0.1792944297194481,
      "num_tokens": 8316604.0,
      "step": 3720
    },
    {
      "entropy": 5.792916488647461,
      "epoch": 0.4183747964283709,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.7111,
      "mean_token_accuracy": 0.1719786509871483,
      "num_tokens": 8328126.0,
      "step": 3725
    },
    {
      "entropy": 5.734830093383789,
      "epoch": 0.4189363733363284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.6239,
      "mean_token_accuracy": 0.1773782342672348,
      "num_tokens": 8339050.0,
      "step": 3730
    },
    {
      "entropy": 5.642179203033447,
      "epoch": 0.41949795024428593,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.4623,
      "mean_token_accuracy": 0.1804968997836113,
      "num_tokens": 8350216.0,
      "step": 3735
    },
    {
      "entropy": 5.7204163551330565,
      "epoch": 0.4200595271522435,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.6339,
      "mean_token_accuracy": 0.17768179327249528,
      "num_tokens": 8362311.0,
      "step": 3740
    },
    {
      "entropy": 5.783650588989258,
      "epoch": 0.42062110406020103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.7899,
      "mean_token_accuracy": 0.17361119985580445,
      "num_tokens": 8373751.0,
      "step": 3745
    },
    {
      "entropy": 5.797900867462158,
      "epoch": 0.4211826809681586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.6225,
      "mean_token_accuracy": 0.17424166947603226,
      "num_tokens": 8384932.0,
      "step": 3750
    },
    {
      "entropy": 5.630241584777832,
      "epoch": 0.42174425787611614,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.5664,
      "mean_token_accuracy": 0.17962265759706497,
      "num_tokens": 8394973.0,
      "step": 3755
    },
    {
      "entropy": 5.756151723861694,
      "epoch": 0.4223058347840737,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.6639,
      "mean_token_accuracy": 0.17477635145187378,
      "num_tokens": 8406590.0,
      "step": 3760
    },
    {
      "entropy": 5.695920419692993,
      "epoch": 0.42286741169203124,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.592,
      "mean_token_accuracy": 0.17508405297994614,
      "num_tokens": 8417127.0,
      "step": 3765
    },
    {
      "entropy": 5.719182825088501,
      "epoch": 0.4234289885999888,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.6987,
      "mean_token_accuracy": 0.1726223737001419,
      "num_tokens": 8429442.0,
      "step": 3770
    },
    {
      "entropy": 5.696038007736206,
      "epoch": 0.42399056550794634,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.595,
      "mean_token_accuracy": 0.17900979667901992,
      "num_tokens": 8439375.0,
      "step": 3775
    },
    {
      "entropy": 5.667362976074219,
      "epoch": 0.42455214241590383,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.5991,
      "mean_token_accuracy": 0.18217457532882692,
      "num_tokens": 8449817.0,
      "step": 3780
    },
    {
      "entropy": 5.666533088684082,
      "epoch": 0.4251137193238614,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.5059,
      "mean_token_accuracy": 0.18276909589767457,
      "num_tokens": 8460096.0,
      "step": 3785
    },
    {
      "entropy": 5.695093154907227,
      "epoch": 0.42567529623181893,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.6116,
      "mean_token_accuracy": 0.17873996049165725,
      "num_tokens": 8470241.0,
      "step": 3790
    },
    {
      "entropy": 5.745785713195801,
      "epoch": 0.4262368731397765,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.6353,
      "mean_token_accuracy": 0.16566867083311082,
      "num_tokens": 8481438.0,
      "step": 3795
    },
    {
      "entropy": 5.791307163238526,
      "epoch": 0.42679845004773403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498878406420253,
      "loss": 5.6993,
      "mean_token_accuracy": 0.16865878105163573,
      "num_tokens": 8492588.0,
      "step": 3800
    },
    {
      "entropy": 5.663080310821533,
      "epoch": 0.4273600269556916,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.6246,
      "mean_token_accuracy": 0.17759007811546326,
      "num_tokens": 8502298.0,
      "step": 3805
    },
    {
      "entropy": 5.645330810546875,
      "epoch": 0.42792160386364914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.46,
      "mean_token_accuracy": 0.1789509564638138,
      "num_tokens": 8513507.0,
      "step": 3810
    },
    {
      "entropy": 5.767054462432862,
      "epoch": 0.4284831807716067,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.5565,
      "mean_token_accuracy": 0.18155122995376588,
      "num_tokens": 8523888.0,
      "step": 3815
    },
    {
      "entropy": 5.657794332504272,
      "epoch": 0.42904475767956424,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498862334234653,
      "loss": 5.5043,
      "mean_token_accuracy": 0.1860446438193321,
      "num_tokens": 8534032.0,
      "step": 3820
    },
    {
      "entropy": 5.664388275146484,
      "epoch": 0.4296063345875218,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.6528,
      "mean_token_accuracy": 0.17626046091318132,
      "num_tokens": 8545544.0,
      "step": 3825
    },
    {
      "entropy": 5.736386108398437,
      "epoch": 0.4301679114954793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.6058,
      "mean_token_accuracy": 0.1761350154876709,
      "num_tokens": 8555428.0,
      "step": 3830
    },
    {
      "entropy": 5.627249908447266,
      "epoch": 0.43072948840343683,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.4982,
      "mean_token_accuracy": 0.18298625946044922,
      "num_tokens": 8566456.0,
      "step": 3835
    },
    {
      "entropy": 5.6816253662109375,
      "epoch": 0.4312910653113944,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.5049,
      "mean_token_accuracy": 0.18162183314561844,
      "num_tokens": 8578125.0,
      "step": 3840
    },
    {
      "entropy": 5.71356987953186,
      "epoch": 0.43185264221935193,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.5837,
      "mean_token_accuracy": 0.17177168726921083,
      "num_tokens": 8589530.0,
      "step": 3845
    },
    {
      "entropy": 5.738637924194336,
      "epoch": 0.4324142191273095,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.695,
      "mean_token_accuracy": 0.17725915759801864,
      "num_tokens": 8601655.0,
      "step": 3850
    },
    {
      "entropy": 5.653723478317261,
      "epoch": 0.43297579603526704,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.6109,
      "mean_token_accuracy": 0.17556392550468444,
      "num_tokens": 8613355.0,
      "step": 3855
    },
    {
      "entropy": 5.665377569198609,
      "epoch": 0.4335373729432246,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.5393,
      "mean_token_accuracy": 0.18229811638593674,
      "num_tokens": 8624154.0,
      "step": 3860
    },
    {
      "entropy": 5.6893205642700195,
      "epoch": 0.43409894985118214,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049882575484962,
      "loss": 5.6251,
      "mean_token_accuracy": 0.17756401896476745,
      "num_tokens": 8633888.0,
      "step": 3865
    },
    {
      "entropy": 5.708053398132324,
      "epoch": 0.4346605267591397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.5992,
      "mean_token_accuracy": 0.17779675424098967,
      "num_tokens": 8644965.0,
      "step": 3870
    },
    {
      "entropy": 5.706338500976562,
      "epoch": 0.4352221036670972,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.545,
      "mean_token_accuracy": 0.183621321618557,
      "num_tokens": 8655483.0,
      "step": 3875
    },
    {
      "entropy": 5.690682077407837,
      "epoch": 0.43578368057505473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498813433436022,
      "loss": 5.5178,
      "mean_token_accuracy": 0.17505193054676055,
      "num_tokens": 8666443.0,
      "step": 3880
    },
    {
      "entropy": 5.733745527267456,
      "epoch": 0.4363452574830123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.5847,
      "mean_token_accuracy": 0.18064320236444473,
      "num_tokens": 8677198.0,
      "step": 3885
    },
    {
      "entropy": 5.660390377044678,
      "epoch": 0.43690683439096983,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.6425,
      "mean_token_accuracy": 0.17053310573101044,
      "num_tokens": 8688303.0,
      "step": 3890
    },
    {
      "entropy": 5.738673543930053,
      "epoch": 0.4374684112989274,
      "grad_norm": 0.875,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.6068,
      "mean_token_accuracy": 0.17490874230861664,
      "num_tokens": 8700326.0,
      "step": 3895
    },
    {
      "entropy": 5.606908464431763,
      "epoch": 0.43802998820688493,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.5531,
      "mean_token_accuracy": 0.18172637969255448,
      "num_tokens": 8711101.0,
      "step": 3900
    },
    {
      "entropy": 5.669354200363159,
      "epoch": 0.4385915651148425,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.6969,
      "mean_token_accuracy": 0.1714338928461075,
      "num_tokens": 8721751.0,
      "step": 3905
    },
    {
      "entropy": 5.812888717651367,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.7783,
      "mean_token_accuracy": 0.16904164403676986,
      "num_tokens": 8733550.0,
      "step": 3910
    },
    {
      "entropy": 5.734032821655274,
      "epoch": 0.4397147189307576,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.5672,
      "mean_token_accuracy": 0.18006644397974014,
      "num_tokens": 8744218.0,
      "step": 3915
    },
    {
      "entropy": 5.663970279693603,
      "epoch": 0.44027629583871514,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.6093,
      "mean_token_accuracy": 0.1743227168917656,
      "num_tokens": 8756531.0,
      "step": 3920
    },
    {
      "entropy": 5.830068778991699,
      "epoch": 0.44083787274667263,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.6509,
      "mean_token_accuracy": 0.17603119611740112,
      "num_tokens": 8768114.0,
      "step": 3925
    },
    {
      "entropy": 5.6779029846191404,
      "epoch": 0.4413994496546302,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.6435,
      "mean_token_accuracy": 0.17040744572877883,
      "num_tokens": 8780010.0,
      "step": 3930
    },
    {
      "entropy": 5.763619709014892,
      "epoch": 0.44196102656258773,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.6019,
      "mean_token_accuracy": 0.1691674992442131,
      "num_tokens": 8792183.0,
      "step": 3935
    },
    {
      "entropy": 5.758844232559204,
      "epoch": 0.4425226034705453,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.7029,
      "mean_token_accuracy": 0.16982857137918472,
      "num_tokens": 8804513.0,
      "step": 3940
    },
    {
      "entropy": 5.672817754745483,
      "epoch": 0.44308418037850283,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.5779,
      "mean_token_accuracy": 0.1853375628590584,
      "num_tokens": 8814918.0,
      "step": 3945
    },
    {
      "entropy": 5.673820924758911,
      "epoch": 0.4436457572864604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.5445,
      "mean_token_accuracy": 0.17637069821357726,
      "num_tokens": 8826176.0,
      "step": 3950
    },
    {
      "entropy": 5.61413893699646,
      "epoch": 0.44420733419441794,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.5677,
      "mean_token_accuracy": 0.17783082276582718,
      "num_tokens": 8837242.0,
      "step": 3955
    },
    {
      "entropy": 5.780583095550537,
      "epoch": 0.4447689111023755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.5419,
      "mean_token_accuracy": 0.1803451105952263,
      "num_tokens": 8848972.0,
      "step": 3960
    },
    {
      "entropy": 5.590166616439819,
      "epoch": 0.44533048801033304,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.5183,
      "mean_token_accuracy": 0.17309907227754592,
      "num_tokens": 8861108.0,
      "step": 3965
    },
    {
      "entropy": 5.674707412719727,
      "epoch": 0.4458920649182906,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.629,
      "mean_token_accuracy": 0.17385578751564026,
      "num_tokens": 8872161.0,
      "step": 3970
    },
    {
      "entropy": 5.906851768493652,
      "epoch": 0.4464536418262481,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.6923,
      "mean_token_accuracy": 0.17808567732572556,
      "num_tokens": 8882497.0,
      "step": 3975
    },
    {
      "entropy": 5.525469541549683,
      "epoch": 0.44701521873420563,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.4302,
      "mean_token_accuracy": 0.18390298187732695,
      "num_tokens": 8893155.0,
      "step": 3980
    },
    {
      "entropy": 5.612870502471924,
      "epoch": 0.4475767956421632,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498725387911674,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18402701914310454,
      "num_tokens": 8904478.0,
      "step": 3985
    },
    {
      "entropy": 5.669520568847656,
      "epoch": 0.44813837255012073,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.6115,
      "mean_token_accuracy": 0.17307900935411452,
      "num_tokens": 8915470.0,
      "step": 3990
    },
    {
      "entropy": 5.748878002166748,
      "epoch": 0.4486999494580783,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498716838763451,
      "loss": 5.5531,
      "mean_token_accuracy": 0.17950351685285568,
      "num_tokens": 8925644.0,
      "step": 3995
    },
    {
      "entropy": 5.708513021469116,
      "epoch": 0.44926152636603583,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498712553504425,
      "loss": 5.5843,
      "mean_token_accuracy": 0.17659425139427185,
      "num_tokens": 8937000.0,
      "step": 4000
    },
    {
      "entropy": 5.657143020629883,
      "epoch": 0.4498231032739934,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.5782,
      "mean_token_accuracy": 0.1724244773387909,
      "num_tokens": 8948258.0,
      "step": 4005
    },
    {
      "entropy": 5.712979650497436,
      "epoch": 0.45038468018195094,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.5817,
      "mean_token_accuracy": 0.1748627707362175,
      "num_tokens": 8960107.0,
      "step": 4010
    },
    {
      "entropy": 5.744423294067383,
      "epoch": 0.4509462570899085,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.5799,
      "mean_token_accuracy": 0.1767027959227562,
      "num_tokens": 8969612.0,
      "step": 4015
    },
    {
      "entropy": 5.763703680038452,
      "epoch": 0.451507833997866,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.6786,
      "mean_token_accuracy": 0.17461298555135726,
      "num_tokens": 8980970.0,
      "step": 4020
    },
    {
      "entropy": 5.721253871917725,
      "epoch": 0.45206941090582353,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.6045,
      "mean_token_accuracy": 0.1786344677209854,
      "num_tokens": 8992218.0,
      "step": 4025
    },
    {
      "entropy": 5.594237995147705,
      "epoch": 0.4526309878137811,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.5428,
      "mean_token_accuracy": 0.17788070887327195,
      "num_tokens": 9003737.0,
      "step": 4030
    },
    {
      "entropy": 5.679529619216919,
      "epoch": 0.45319256472173863,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.6029,
      "mean_token_accuracy": 0.17558735609054565,
      "num_tokens": 9015427.0,
      "step": 4035
    },
    {
      "entropy": 5.717615461349487,
      "epoch": 0.4537541416296962,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.6126,
      "mean_token_accuracy": 0.1770196735858917,
      "num_tokens": 9025003.0,
      "step": 4040
    },
    {
      "entropy": 5.649677562713623,
      "epoch": 0.45431571853765373,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.5617,
      "mean_token_accuracy": 0.17711025327444077,
      "num_tokens": 9036074.0,
      "step": 4045
    },
    {
      "entropy": 5.8062300205230715,
      "epoch": 0.4548772954456113,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.5675,
      "mean_token_accuracy": 0.17667278349399568,
      "num_tokens": 9046878.0,
      "step": 4050
    },
    {
      "entropy": 5.71130142211914,
      "epoch": 0.45543887235356884,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.622,
      "mean_token_accuracy": 0.17998187839984894,
      "num_tokens": 9058366.0,
      "step": 4055
    },
    {
      "entropy": 5.726722717285156,
      "epoch": 0.4560004492615264,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498660574833995,
      "loss": 5.6611,
      "mean_token_accuracy": 0.17204404175281524,
      "num_tokens": 9069542.0,
      "step": 4060
    },
    {
      "entropy": 5.74843020439148,
      "epoch": 0.45656202616948394,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.8364,
      "mean_token_accuracy": 0.16937779784202575,
      "num_tokens": 9081219.0,
      "step": 4065
    },
    {
      "entropy": 5.757581186294556,
      "epoch": 0.45712360307744143,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.5295,
      "mean_token_accuracy": 0.18289404660463332,
      "num_tokens": 9092101.0,
      "step": 4070
    },
    {
      "entropy": 5.670637702941894,
      "epoch": 0.457685179985399,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498647419924648,
      "loss": 5.5692,
      "mean_token_accuracy": 0.17209210842847825,
      "num_tokens": 9101900.0,
      "step": 4075
    },
    {
      "entropy": 5.769422960281372,
      "epoch": 0.45824675689335653,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498643020712638,
      "loss": 5.7047,
      "mean_token_accuracy": 0.17082441747188568,
      "num_tokens": 9114369.0,
      "step": 4080
    },
    {
      "entropy": 5.673053789138794,
      "epoch": 0.4588083338013141,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.5125,
      "mean_token_accuracy": 0.17445550709962845,
      "num_tokens": 9126548.0,
      "step": 4085
    },
    {
      "entropy": 5.714936351776123,
      "epoch": 0.45936991070927163,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.7072,
      "mean_token_accuracy": 0.17885012328624725,
      "num_tokens": 9138070.0,
      "step": 4090
    },
    {
      "entropy": 5.760405254364014,
      "epoch": 0.4599314876172292,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.6077,
      "mean_token_accuracy": 0.1735104128718376,
      "num_tokens": 9149489.0,
      "step": 4095
    },
    {
      "entropy": 5.67493782043457,
      "epoch": 0.46049306452518673,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.5924,
      "mean_token_accuracy": 0.18236540853977204,
      "num_tokens": 9160363.0,
      "step": 4100
    },
    {
      "entropy": 5.7304445743560795,
      "epoch": 0.4610546414331443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.6528,
      "mean_token_accuracy": 0.1718761533498764,
      "num_tokens": 9171619.0,
      "step": 4105
    },
    {
      "entropy": 5.67438383102417,
      "epoch": 0.46161621834110184,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498616475907006,
      "loss": 5.5035,
      "mean_token_accuracy": 0.17841653674840927,
      "num_tokens": 9182828.0,
      "step": 4110
    },
    {
      "entropy": 5.775537061691284,
      "epoch": 0.46217779524905933,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.6635,
      "mean_token_accuracy": 0.18059315979480745,
      "num_tokens": 9195249.0,
      "step": 4115
    },
    {
      "entropy": 5.564350366592407,
      "epoch": 0.4627393721570169,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.4649,
      "mean_token_accuracy": 0.1827670380473137,
      "num_tokens": 9205770.0,
      "step": 4120
    },
    {
      "entropy": 5.616291856765747,
      "epoch": 0.46330094906497443,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.4841,
      "mean_token_accuracy": 0.18152681291103362,
      "num_tokens": 9216367.0,
      "step": 4125
    },
    {
      "entropy": 5.703184652328491,
      "epoch": 0.463862525972932,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.5842,
      "mean_token_accuracy": 0.17080659717321395,
      "num_tokens": 9228524.0,
      "step": 4130
    },
    {
      "entropy": 5.67804183959961,
      "epoch": 0.46442410288088953,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.5459,
      "mean_token_accuracy": 0.18683858066797257,
      "num_tokens": 9238907.0,
      "step": 4135
    },
    {
      "entropy": 5.7259095191955565,
      "epoch": 0.4649856797888471,
      "grad_norm": 0.828125,
      "learning_rate": 0.000498589674780126,
      "loss": 5.6864,
      "mean_token_accuracy": 0.17780255675315856,
      "num_tokens": 9250491.0,
      "step": 4140
    },
    {
      "entropy": 5.699339294433594,
      "epoch": 0.46554725669680463,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.4742,
      "mean_token_accuracy": 0.18620740622282028,
      "num_tokens": 9260984.0,
      "step": 4145
    },
    {
      "entropy": 5.7042759418487545,
      "epoch": 0.4661088336047622,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.6436,
      "mean_token_accuracy": 0.1757622852921486,
      "num_tokens": 9272416.0,
      "step": 4150
    },
    {
      "entropy": 5.714568996429444,
      "epoch": 0.46667041051271974,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498576178105807,
      "loss": 5.6256,
      "mean_token_accuracy": 0.17754258066415787,
      "num_tokens": 9284513.0,
      "step": 4155
    },
    {
      "entropy": 5.716816473007202,
      "epoch": 0.4672319874206773,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.5227,
      "mean_token_accuracy": 0.1860297590494156,
      "num_tokens": 9293422.0,
      "step": 4160
    },
    {
      "entropy": 5.625087022781372,
      "epoch": 0.4677935643286348,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.51,
      "mean_token_accuracy": 0.1776867017149925,
      "num_tokens": 9304735.0,
      "step": 4165
    },
    {
      "entropy": 5.643379878997803,
      "epoch": 0.46835514123659233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.6197,
      "mean_token_accuracy": 0.17623546570539475,
      "num_tokens": 9315659.0,
      "step": 4170
    },
    {
      "entropy": 5.682793521881104,
      "epoch": 0.4689167181445499,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.6363,
      "mean_token_accuracy": 0.17618870884180068,
      "num_tokens": 9327106.0,
      "step": 4175
    },
    {
      "entropy": 5.805673456192016,
      "epoch": 0.46947829505250743,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.6816,
      "mean_token_accuracy": 0.17468448281288146,
      "num_tokens": 9338762.0,
      "step": 4180
    },
    {
      "entropy": 5.689402008056641,
      "epoch": 0.470039871960465,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498548992554749,
      "loss": 5.5366,
      "mean_token_accuracy": 0.17719955295324324,
      "num_tokens": 9349495.0,
      "step": 4185
    },
    {
      "entropy": 5.751715517044067,
      "epoch": 0.47060144886842253,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.6351,
      "mean_token_accuracy": 0.1837965279817581,
      "num_tokens": 9360830.0,
      "step": 4190
    },
    {
      "entropy": 5.611725521087647,
      "epoch": 0.4711630257763801,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.5128,
      "mean_token_accuracy": 0.17987315803766252,
      "num_tokens": 9371668.0,
      "step": 4195
    },
    {
      "entropy": 5.692264032363892,
      "epoch": 0.47172460268433763,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.6528,
      "mean_token_accuracy": 0.17369603365659714,
      "num_tokens": 9383437.0,
      "step": 4200
    },
    {
      "entropy": 5.6641716957092285,
      "epoch": 0.4722861795922952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.4732,
      "mean_token_accuracy": 0.18241805881261824,
      "num_tokens": 9394597.0,
      "step": 4205
    },
    {
      "entropy": 5.5171393871307375,
      "epoch": 0.47284775650025274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18648850917816162,
      "num_tokens": 9404322.0,
      "step": 4210
    },
    {
      "entropy": 5.739780330657959,
      "epoch": 0.47340933340821023,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.7738,
      "mean_token_accuracy": 0.16741739362478256,
      "num_tokens": 9416740.0,
      "step": 4215
    },
    {
      "entropy": 5.774796772003174,
      "epoch": 0.4739709103161678,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1814936414361,
      "num_tokens": 9428617.0,
      "step": 4220
    },
    {
      "entropy": 5.589036846160889,
      "epoch": 0.47453248722412533,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.5365,
      "mean_token_accuracy": 0.18152604401111602,
      "num_tokens": 9439686.0,
      "step": 4225
    },
    {
      "entropy": 5.66499891281128,
      "epoch": 0.4750940641320829,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.5205,
      "mean_token_accuracy": 0.18545324802398683,
      "num_tokens": 9450378.0,
      "step": 4230
    },
    {
      "entropy": 5.71506986618042,
      "epoch": 0.47565564104004043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.5493,
      "mean_token_accuracy": 0.17825888246297836,
      "num_tokens": 9461659.0,
      "step": 4235
    },
    {
      "entropy": 5.6576087474823,
      "epoch": 0.476217217947998,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.5898,
      "mean_token_accuracy": 0.18185754865407944,
      "num_tokens": 9472528.0,
      "step": 4240
    },
    {
      "entropy": 5.703695344924927,
      "epoch": 0.47677879485595553,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.5287,
      "mean_token_accuracy": 0.1899208128452301,
      "num_tokens": 9483388.0,
      "step": 4245
    },
    {
      "entropy": 5.507798099517823,
      "epoch": 0.4773403717639131,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498489211510521,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1911943331360817,
      "num_tokens": 9495178.0,
      "step": 4250
    },
    {
      "entropy": 5.614229679107666,
      "epoch": 0.47790194867187064,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.6046,
      "mean_token_accuracy": 0.1775161936879158,
      "num_tokens": 9506226.0,
      "step": 4255
    },
    {
      "entropy": 5.740632104873657,
      "epoch": 0.47846352557982813,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.5104,
      "mean_token_accuracy": 0.17976359128952027,
      "num_tokens": 9517671.0,
      "step": 4260
    },
    {
      "entropy": 5.634478282928467,
      "epoch": 0.4790251024877857,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.5976,
      "mean_token_accuracy": 0.18178414106369017,
      "num_tokens": 9527637.0,
      "step": 4265
    },
    {
      "entropy": 5.764854478836059,
      "epoch": 0.47958667939574323,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498470575384638,
      "loss": 5.6761,
      "mean_token_accuracy": 0.17296643704175949,
      "num_tokens": 9539686.0,
      "step": 4270
    },
    {
      "entropy": 5.710594987869262,
      "epoch": 0.4801482563037008,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.6446,
      "mean_token_accuracy": 0.18062152415513993,
      "num_tokens": 9550775.0,
      "step": 4275
    },
    {
      "entropy": 5.712454986572266,
      "epoch": 0.48070983321165833,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.5456,
      "mean_token_accuracy": 0.17569558173418046,
      "num_tokens": 9561760.0,
      "step": 4280
    },
    {
      "entropy": 5.592285585403443,
      "epoch": 0.4812714101196159,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.5841,
      "mean_token_accuracy": 0.17345283627510072,
      "num_tokens": 9573899.0,
      "step": 4285
    },
    {
      "entropy": 5.622520446777344,
      "epoch": 0.48183298702757343,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498451825412523,
      "loss": 5.5414,
      "mean_token_accuracy": 0.17724068611860275,
      "num_tokens": 9584336.0,
      "step": 4290
    },
    {
      "entropy": 5.816002225875854,
      "epoch": 0.482394563935531,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.733,
      "mean_token_accuracy": 0.17272676825523375,
      "num_tokens": 9596614.0,
      "step": 4295
    },
    {
      "entropy": 5.745497703552246,
      "epoch": 0.48295614084348854,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.5683,
      "mean_token_accuracy": 0.1837946966290474,
      "num_tokens": 9606796.0,
      "step": 4300
    },
    {
      "entropy": 5.686303329467774,
      "epoch": 0.4835177177514461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.5547,
      "mean_token_accuracy": 0.1825352668762207,
      "num_tokens": 9617705.0,
      "step": 4305
    },
    {
      "entropy": 5.594384860992432,
      "epoch": 0.4840792946594036,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.5749,
      "mean_token_accuracy": 0.18000427335500718,
      "num_tokens": 9628098.0,
      "step": 4310
    },
    {
      "entropy": 5.730225658416748,
      "epoch": 0.48464087156736113,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.5117,
      "mean_token_accuracy": 0.18523356467485427,
      "num_tokens": 9637697.0,
      "step": 4315
    },
    {
      "entropy": 5.626879787445068,
      "epoch": 0.4852024484753187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.4198,
      "mean_token_accuracy": 0.18843667954206467,
      "num_tokens": 9648221.0,
      "step": 4320
    },
    {
      "entropy": 5.726041078567505,
      "epoch": 0.48576402538327623,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.6037,
      "mean_token_accuracy": 0.17221028804779054,
      "num_tokens": 9660182.0,
      "step": 4325
    },
    {
      "entropy": 5.653705310821533,
      "epoch": 0.4863256022912338,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.6154,
      "mean_token_accuracy": 0.1761000081896782,
      "num_tokens": 9672121.0,
      "step": 4330
    },
    {
      "entropy": 5.697546911239624,
      "epoch": 0.48688717919919133,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.5036,
      "mean_token_accuracy": 0.18291919976472854,
      "num_tokens": 9682651.0,
      "step": 4335
    },
    {
      "entropy": 5.700353384017944,
      "epoch": 0.4874487561071489,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.6686,
      "mean_token_accuracy": 0.17320136576890946,
      "num_tokens": 9693989.0,
      "step": 4340
    },
    {
      "entropy": 5.607551622390747,
      "epoch": 0.48801033301510643,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.5109,
      "mean_token_accuracy": 0.18393711745738983,
      "num_tokens": 9703802.0,
      "step": 4345
    },
    {
      "entropy": 5.662387704849243,
      "epoch": 0.488571909923064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498394892514596,
      "loss": 5.461,
      "mean_token_accuracy": 0.18972480595111846,
      "num_tokens": 9714287.0,
      "step": 4350
    },
    {
      "entropy": 5.616183423995972,
      "epoch": 0.48913348683102154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.4953,
      "mean_token_accuracy": 0.18152210414409636,
      "num_tokens": 9724819.0,
      "step": 4355
    },
    {
      "entropy": 5.637750244140625,
      "epoch": 0.48969506373897903,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.6436,
      "mean_token_accuracy": 0.1711961507797241,
      "num_tokens": 9736326.0,
      "step": 4360
    },
    {
      "entropy": 5.772123241424561,
      "epoch": 0.4902566406469366,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.665,
      "mean_token_accuracy": 0.17694464474916458,
      "num_tokens": 9747883.0,
      "step": 4365
    },
    {
      "entropy": 5.6365245342254635,
      "epoch": 0.49081821755489413,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.5193,
      "mean_token_accuracy": 0.18201339542865752,
      "num_tokens": 9758191.0,
      "step": 4370
    },
    {
      "entropy": 5.715147972106934,
      "epoch": 0.4913797944628517,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.617,
      "mean_token_accuracy": 0.17647611498832702,
      "num_tokens": 9768852.0,
      "step": 4375
    },
    {
      "entropy": 5.734269666671753,
      "epoch": 0.49194137137080923,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.5551,
      "mean_token_accuracy": 0.17609058320522308,
      "num_tokens": 9779033.0,
      "step": 4380
    },
    {
      "entropy": 5.6143005847930905,
      "epoch": 0.4925029482787668,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.4827,
      "mean_token_accuracy": 0.18476453572511672,
      "num_tokens": 9790777.0,
      "step": 4385
    },
    {
      "entropy": 5.769971466064453,
      "epoch": 0.49306452518672433,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.7418,
      "mean_token_accuracy": 0.16620710790157317,
      "num_tokens": 9802305.0,
      "step": 4390
    },
    {
      "entropy": 5.711023330688477,
      "epoch": 0.4936261020946819,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.5426,
      "mean_token_accuracy": 0.17464770376682281,
      "num_tokens": 9813590.0,
      "step": 4395
    },
    {
      "entropy": 5.67334361076355,
      "epoch": 0.49418767900263944,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.6284,
      "mean_token_accuracy": 0.17277866154909133,
      "num_tokens": 9824080.0,
      "step": 4400
    },
    {
      "entropy": 5.697000885009766,
      "epoch": 0.49474925591059693,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.5046,
      "mean_token_accuracy": 0.176873978972435,
      "num_tokens": 9835441.0,
      "step": 4405
    },
    {
      "entropy": 5.627409172058106,
      "epoch": 0.4953108328185545,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.4943,
      "mean_token_accuracy": 0.1827621951699257,
      "num_tokens": 9846750.0,
      "step": 4410
    },
    {
      "entropy": 5.710919618606567,
      "epoch": 0.49587240972651203,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.651,
      "mean_token_accuracy": 0.17651382833719254,
      "num_tokens": 9857987.0,
      "step": 4415
    },
    {
      "entropy": 5.680669784545898,
      "epoch": 0.4964339866344696,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.6017,
      "mean_token_accuracy": 0.17698585242033005,
      "num_tokens": 9869736.0,
      "step": 4420
    },
    {
      "entropy": 5.753222894668579,
      "epoch": 0.49699556354242713,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498322286045317,
      "loss": 5.5771,
      "mean_token_accuracy": 0.17493117451667786,
      "num_tokens": 9881610.0,
      "step": 4425
    },
    {
      "entropy": 5.647378921508789,
      "epoch": 0.4975571404503847,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.5623,
      "mean_token_accuracy": 0.1751056432723999,
      "num_tokens": 9893203.0,
      "step": 4430
    },
    {
      "entropy": 5.6928294658660885,
      "epoch": 0.49811871735834223,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.4897,
      "mean_token_accuracy": 0.17579762488603592,
      "num_tokens": 9904392.0,
      "step": 4435
    },
    {
      "entropy": 5.519991159439087,
      "epoch": 0.4986802942662998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.4734,
      "mean_token_accuracy": 0.18386689871549605,
      "num_tokens": 9916184.0,
      "step": 4440
    },
    {
      "entropy": 5.664324522018433,
      "epoch": 0.49924187117425733,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.5901,
      "mean_token_accuracy": 0.18246736973524094,
      "num_tokens": 9927516.0,
      "step": 4445
    },
    {
      "entropy": 5.741987419128418,
      "epoch": 0.4998034480822149,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.6074,
      "mean_token_accuracy": 0.17554258108139037,
      "num_tokens": 9939182.0,
      "step": 4450
    },
    {
      "entropy": 5.743662643432617,
      "epoch": 0.5003650249901724,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498292795456261,
      "loss": 5.6361,
      "mean_token_accuracy": 0.174177448451519,
      "num_tokens": 9950382.0,
      "step": 4455
    },
    {
      "entropy": 5.707113122940063,
      "epoch": 0.50092660189813,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.6283,
      "mean_token_accuracy": 0.18307442367076873,
      "num_tokens": 9961053.0,
      "step": 4460
    },
    {
      "entropy": 5.644644641876221,
      "epoch": 0.5014881788060875,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.5112,
      "mean_token_accuracy": 0.18234430849552155,
      "num_tokens": 9971313.0,
      "step": 4465
    },
    {
      "entropy": 5.635139608383179,
      "epoch": 0.5020497557140451,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.5398,
      "mean_token_accuracy": 0.1916440635919571,
      "num_tokens": 9981257.0,
      "step": 4470
    },
    {
      "entropy": 5.640893363952637,
      "epoch": 0.5026113326220026,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.4739,
      "mean_token_accuracy": 0.17892623990774154,
      "num_tokens": 9991860.0,
      "step": 4475
    },
    {
      "entropy": 5.623174571990967,
      "epoch": 0.5031729095299601,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.5624,
      "mean_token_accuracy": 0.1814965933561325,
      "num_tokens": 10002320.0,
      "step": 4480
    },
    {
      "entropy": 5.754784631729126,
      "epoch": 0.5037344864379176,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.6358,
      "mean_token_accuracy": 0.17755780518054962,
      "num_tokens": 10013187.0,
      "step": 4485
    },
    {
      "entropy": 5.57007794380188,
      "epoch": 0.5042960633458752,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1904952883720398,
      "num_tokens": 10024268.0,
      "step": 4490
    },
    {
      "entropy": 5.692174434661865,
      "epoch": 0.5048576402538327,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.6562,
      "mean_token_accuracy": 0.17762195765972139,
      "num_tokens": 10034738.0,
      "step": 4495
    },
    {
      "entropy": 5.782405185699463,
      "epoch": 0.5054192171617903,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.6058,
      "mean_token_accuracy": 0.17828063815832138,
      "num_tokens": 10045666.0,
      "step": 4500
    },
    {
      "entropy": 5.604255056381225,
      "epoch": 0.5059807940697478,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.4812,
      "mean_token_accuracy": 0.1805545598268509,
      "num_tokens": 10057084.0,
      "step": 4505
    },
    {
      "entropy": 5.627578830718994,
      "epoch": 0.5065423709777054,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.4564,
      "mean_token_accuracy": 0.1791461542248726,
      "num_tokens": 10066754.0,
      "step": 4510
    },
    {
      "entropy": 5.571236085891724,
      "epoch": 0.5071039478856629,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.4565,
      "mean_token_accuracy": 0.18998907655477523,
      "num_tokens": 10078139.0,
      "step": 4515
    },
    {
      "entropy": 5.666440105438232,
      "epoch": 0.5076655247936205,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.5324,
      "mean_token_accuracy": 0.1763441488146782,
      "num_tokens": 10089652.0,
      "step": 4520
    },
    {
      "entropy": 5.504432582855225,
      "epoch": 0.508227101701578,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.3525,
      "mean_token_accuracy": 0.19223741292953492,
      "num_tokens": 10100642.0,
      "step": 4525
    },
    {
      "entropy": 5.644719839096069,
      "epoch": 0.5087886786095356,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.5838,
      "mean_token_accuracy": 0.17960658222436904,
      "num_tokens": 10112036.0,
      "step": 4530
    },
    {
      "entropy": 5.624582052230835,
      "epoch": 0.5093502555174931,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.455,
      "mean_token_accuracy": 0.1824001595377922,
      "num_tokens": 10121653.0,
      "step": 4535
    },
    {
      "entropy": 5.631822824478149,
      "epoch": 0.5099118324254507,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.5637,
      "mean_token_accuracy": 0.17775342911481856,
      "num_tokens": 10132271.0,
      "step": 4540
    },
    {
      "entropy": 5.5466217517852785,
      "epoch": 0.5104734093334082,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.4496,
      "mean_token_accuracy": 0.18065108060836793,
      "num_tokens": 10143411.0,
      "step": 4545
    },
    {
      "entropy": 5.7046802043914795,
      "epoch": 0.5110349862413658,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.555,
      "mean_token_accuracy": 0.18349169492721557,
      "num_tokens": 10154338.0,
      "step": 4550
    },
    {
      "entropy": 5.668397617340088,
      "epoch": 0.5115965631493233,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.5849,
      "mean_token_accuracy": 0.18315830826759338,
      "num_tokens": 10164810.0,
      "step": 4555
    },
    {
      "entropy": 5.639742231369018,
      "epoch": 0.5121581400572809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.5051,
      "mean_token_accuracy": 0.18220127522945403,
      "num_tokens": 10176323.0,
      "step": 4560
    },
    {
      "entropy": 5.642209482192993,
      "epoch": 0.5127197169652384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.5208,
      "mean_token_accuracy": 0.18486607670783997,
      "num_tokens": 10187110.0,
      "step": 4565
    },
    {
      "entropy": 5.612886762619018,
      "epoch": 0.513281293873196,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.5083,
      "mean_token_accuracy": 0.17804306149482726,
      "num_tokens": 10199099.0,
      "step": 4570
    },
    {
      "entropy": 5.686509990692139,
      "epoch": 0.5138428707811534,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.486,
      "mean_token_accuracy": 0.18251441270112992,
      "num_tokens": 10209726.0,
      "step": 4575
    },
    {
      "entropy": 5.702648544311524,
      "epoch": 0.514404447689111,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049816716337126,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17905279695987703,
      "num_tokens": 10220858.0,
      "step": 4580
    },
    {
      "entropy": 5.518071508407592,
      "epoch": 0.5149660245970685,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.413,
      "mean_token_accuracy": 0.18944605588912963,
      "num_tokens": 10232250.0,
      "step": 4585
    },
    {
      "entropy": 5.725042724609375,
      "epoch": 0.5155276015050261,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.609,
      "mean_token_accuracy": 0.17123129218816757,
      "num_tokens": 10244584.0,
      "step": 4590
    },
    {
      "entropy": 5.716646003723144,
      "epoch": 0.5160891784129836,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.481,
      "mean_token_accuracy": 0.1807915136218071,
      "num_tokens": 10255926.0,
      "step": 4595
    },
    {
      "entropy": 5.665557146072388,
      "epoch": 0.5166507553209412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498146650039855,
      "loss": 5.5723,
      "mean_token_accuracy": 0.18312949687242508,
      "num_tokens": 10266691.0,
      "step": 4600
    },
    {
      "entropy": 5.692429542541504,
      "epoch": 0.5172123322288987,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.6224,
      "mean_token_accuracy": 0.18030687123537065,
      "num_tokens": 10277744.0,
      "step": 4605
    },
    {
      "entropy": 5.600763511657715,
      "epoch": 0.5177739091368563,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.4703,
      "mean_token_accuracy": 0.18670175075531006,
      "num_tokens": 10288792.0,
      "step": 4610
    },
    {
      "entropy": 5.703523778915406,
      "epoch": 0.5183354860448138,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.5218,
      "mean_token_accuracy": 0.18117523938417435,
      "num_tokens": 10299569.0,
      "step": 4615
    },
    {
      "entropy": 5.632540655136109,
      "epoch": 0.5188970629527714,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.4972,
      "mean_token_accuracy": 0.18545956462621688,
      "num_tokens": 10309742.0,
      "step": 4620
    },
    {
      "entropy": 5.680473136901855,
      "epoch": 0.5194586398607289,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.5506,
      "mean_token_accuracy": 0.17689417153596879,
      "num_tokens": 10320603.0,
      "step": 4625
    },
    {
      "entropy": 5.607299709320069,
      "epoch": 0.5200202167686865,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.5088,
      "mean_token_accuracy": 0.18380599915981294,
      "num_tokens": 10331498.0,
      "step": 4630
    },
    {
      "entropy": 5.677219200134277,
      "epoch": 0.520581793676644,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.5408,
      "mean_token_accuracy": 0.18248181641101838,
      "num_tokens": 10342002.0,
      "step": 4635
    },
    {
      "entropy": 5.6430576801300045,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.4927,
      "mean_token_accuracy": 0.18074288517236708,
      "num_tokens": 10351797.0,
      "step": 4640
    },
    {
      "entropy": 5.580838346481324,
      "epoch": 0.5217049474925591,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.4331,
      "mean_token_accuracy": 0.1847847133874893,
      "num_tokens": 10362538.0,
      "step": 4645
    },
    {
      "entropy": 5.600670051574707,
      "epoch": 0.5222665244005167,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18671983033418654,
      "num_tokens": 10373365.0,
      "step": 4650
    },
    {
      "entropy": 5.558433389663696,
      "epoch": 0.5228281013084742,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.454,
      "mean_token_accuracy": 0.18048152774572374,
      "num_tokens": 10385087.0,
      "step": 4655
    },
    {
      "entropy": 5.618748950958252,
      "epoch": 0.5233896782164318,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.5731,
      "mean_token_accuracy": 0.17731207013130187,
      "num_tokens": 10396014.0,
      "step": 4660
    },
    {
      "entropy": 5.579803514480591,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.4968,
      "mean_token_accuracy": 0.18150824904441834,
      "num_tokens": 10407844.0,
      "step": 4665
    },
    {
      "entropy": 5.580438184738159,
      "epoch": 0.5245128320323468,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.5668,
      "mean_token_accuracy": 0.17426397800445556,
      "num_tokens": 10419892.0,
      "step": 4670
    },
    {
      "entropy": 5.690279579162597,
      "epoch": 0.5250744089403043,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.527,
      "mean_token_accuracy": 0.1738398015499115,
      "num_tokens": 10430686.0,
      "step": 4675
    },
    {
      "entropy": 5.643109941482544,
      "epoch": 0.5256359858482619,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.5599,
      "mean_token_accuracy": 0.17802226841449736,
      "num_tokens": 10443785.0,
      "step": 4680
    },
    {
      "entropy": 5.587534379959107,
      "epoch": 0.5261975627562194,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.4399,
      "mean_token_accuracy": 0.1904889687895775,
      "num_tokens": 10454889.0,
      "step": 4685
    },
    {
      "entropy": 5.597082805633545,
      "epoch": 0.526759139664177,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.6024,
      "mean_token_accuracy": 0.1754663571715355,
      "num_tokens": 10466525.0,
      "step": 4690
    },
    {
      "entropy": 5.722401237487793,
      "epoch": 0.5273207165721345,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.5465,
      "mean_token_accuracy": 0.18529189825057985,
      "num_tokens": 10476329.0,
      "step": 4695
    },
    {
      "entropy": 5.748387002944947,
      "epoch": 0.5278822934800921,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.6097,
      "mean_token_accuracy": 0.17898427844047546,
      "num_tokens": 10487178.0,
      "step": 4700
    },
    {
      "entropy": 5.644168710708618,
      "epoch": 0.5284438703880496,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.5554,
      "mean_token_accuracy": 0.1712798923254013,
      "num_tokens": 10499293.0,
      "step": 4705
    },
    {
      "entropy": 5.768489265441895,
      "epoch": 0.5290054472960072,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00049803179494862,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16541629433631896,
      "num_tokens": 10511762.0,
      "step": 4710
    },
    {
      "entropy": 5.667138910293579,
      "epoch": 0.5295670242039647,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.4547,
      "mean_token_accuracy": 0.19035084992647172,
      "num_tokens": 10521725.0,
      "step": 4715
    },
    {
      "entropy": 5.569765329360962,
      "epoch": 0.5301286011119223,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.4007,
      "mean_token_accuracy": 0.18478177189826966,
      "num_tokens": 10531764.0,
      "step": 4720
    },
    {
      "entropy": 5.67407488822937,
      "epoch": 0.5306901780198798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.4897,
      "mean_token_accuracy": 0.18613315969705582,
      "num_tokens": 10542127.0,
      "step": 4725
    },
    {
      "entropy": 5.706964445114136,
      "epoch": 0.5312517549278374,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.5833,
      "mean_token_accuracy": 0.1766618549823761,
      "num_tokens": 10552580.0,
      "step": 4730
    },
    {
      "entropy": 5.628377914428711,
      "epoch": 0.5318133318357949,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.5111,
      "mean_token_accuracy": 0.1801474079489708,
      "num_tokens": 10562961.0,
      "step": 4735
    },
    {
      "entropy": 5.730417633056641,
      "epoch": 0.5323749087437525,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.5057,
      "mean_token_accuracy": 0.18176156133413315,
      "num_tokens": 10573210.0,
      "step": 4740
    },
    {
      "entropy": 5.521605539321899,
      "epoch": 0.53293648565171,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.4754,
      "mean_token_accuracy": 0.17918767482042314,
      "num_tokens": 10583878.0,
      "step": 4745
    },
    {
      "entropy": 5.64623875617981,
      "epoch": 0.5334980625596676,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.5537,
      "mean_token_accuracy": 0.17740156054496764,
      "num_tokens": 10593420.0,
      "step": 4750
    },
    {
      "entropy": 5.704613780975341,
      "epoch": 0.5340596394676251,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497983818025976,
      "loss": 5.5753,
      "mean_token_accuracy": 0.1787201002240181,
      "num_tokens": 10604213.0,
      "step": 4755
    },
    {
      "entropy": 5.696094036102295,
      "epoch": 0.5346212163755827,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.5457,
      "mean_token_accuracy": 0.1722053959965706,
      "num_tokens": 10615502.0,
      "step": 4760
    },
    {
      "entropy": 5.6494340896606445,
      "epoch": 0.5351827932835401,
      "grad_norm": 0.875,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.6009,
      "mean_token_accuracy": 0.17084263116121293,
      "num_tokens": 10627589.0,
      "step": 4765
    },
    {
      "entropy": 5.5736168384552,
      "epoch": 0.5357443701914977,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.4282,
      "mean_token_accuracy": 0.18405757546424867,
      "num_tokens": 10638365.0,
      "step": 4770
    },
    {
      "entropy": 5.637544584274292,
      "epoch": 0.5363059470994552,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497962310344703,
      "loss": 5.5334,
      "mean_token_accuracy": 0.18157838881015778,
      "num_tokens": 10650664.0,
      "step": 4775
    },
    {
      "entropy": 5.681172561645508,
      "epoch": 0.5368675240074128,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00049795691567594,
      "loss": 5.5893,
      "mean_token_accuracy": 0.18533241599798203,
      "num_tokens": 10662418.0,
      "step": 4780
    },
    {
      "entropy": 5.627472829818726,
      "epoch": 0.5374291009153703,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.4668,
      "mean_token_accuracy": 0.184379480779171,
      "num_tokens": 10672793.0,
      "step": 4785
    },
    {
      "entropy": 5.5839722633361815,
      "epoch": 0.5379906778233279,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.5119,
      "mean_token_accuracy": 0.1882890358567238,
      "num_tokens": 10684082.0,
      "step": 4790
    },
    {
      "entropy": 5.492992544174195,
      "epoch": 0.5385522547312854,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.4543,
      "mean_token_accuracy": 0.1841624855995178,
      "num_tokens": 10694490.0,
      "step": 4795
    },
    {
      "entropy": 5.63221583366394,
      "epoch": 0.539113831639243,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.4588,
      "mean_token_accuracy": 0.18302005976438523,
      "num_tokens": 10704731.0,
      "step": 4800
    },
    {
      "entropy": 5.611063766479492,
      "epoch": 0.5396754085472005,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.4887,
      "mean_token_accuracy": 0.18789714425802231,
      "num_tokens": 10715735.0,
      "step": 4805
    },
    {
      "entropy": 5.545275735855102,
      "epoch": 0.5402369854551581,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.4553,
      "mean_token_accuracy": 0.1809619754552841,
      "num_tokens": 10726951.0,
      "step": 4810
    },
    {
      "entropy": 5.602046537399292,
      "epoch": 0.5407985623631156,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.5138,
      "mean_token_accuracy": 0.17767608910799026,
      "num_tokens": 10737609.0,
      "step": 4815
    },
    {
      "entropy": 5.606081628799439,
      "epoch": 0.5413601392710732,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.4109,
      "mean_token_accuracy": 0.19260681569576263,
      "num_tokens": 10748957.0,
      "step": 4820
    },
    {
      "entropy": 5.6902268409729,
      "epoch": 0.5419217161790307,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.5907,
      "mean_token_accuracy": 0.17853567153215408,
      "num_tokens": 10759398.0,
      "step": 4825
    },
    {
      "entropy": 5.654111862182617,
      "epoch": 0.5424832930869883,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.5672,
      "mean_token_accuracy": 0.17787721753120422,
      "num_tokens": 10770436.0,
      "step": 4830
    },
    {
      "entropy": 5.577683019638061,
      "epoch": 0.5430448699949458,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.3983,
      "mean_token_accuracy": 0.19398372024297714,
      "num_tokens": 10781160.0,
      "step": 4835
    },
    {
      "entropy": 5.692035388946533,
      "epoch": 0.5436064469029034,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.5257,
      "mean_token_accuracy": 0.18834299147129058,
      "num_tokens": 10793170.0,
      "step": 4840
    },
    {
      "entropy": 5.574293184280395,
      "epoch": 0.5441680238108609,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.4673,
      "mean_token_accuracy": 0.18172243386507034,
      "num_tokens": 10803515.0,
      "step": 4845
    },
    {
      "entropy": 5.624358606338501,
      "epoch": 0.5447296007188185,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.5051,
      "mean_token_accuracy": 0.18323239982128142,
      "num_tokens": 10814508.0,
      "step": 4850
    },
    {
      "entropy": 5.687337398529053,
      "epoch": 0.545291177626776,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.6114,
      "mean_token_accuracy": 0.17625804543495177,
      "num_tokens": 10825924.0,
      "step": 4855
    },
    {
      "entropy": 5.649694776535034,
      "epoch": 0.5458527545347336,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.5908,
      "mean_token_accuracy": 0.17919801622629167,
      "num_tokens": 10837854.0,
      "step": 4860
    },
    {
      "entropy": 5.60722770690918,
      "epoch": 0.546414331442691,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.4684,
      "mean_token_accuracy": 0.18659446835517884,
      "num_tokens": 10848279.0,
      "step": 4865
    },
    {
      "entropy": 5.557261037826538,
      "epoch": 0.5469759083506486,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.4864,
      "mean_token_accuracy": 0.17544039487838745,
      "num_tokens": 10859615.0,
      "step": 4870
    },
    {
      "entropy": 5.695983839035034,
      "epoch": 0.5475374852586061,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.4622,
      "mean_token_accuracy": 0.18064787834882737,
      "num_tokens": 10871388.0,
      "step": 4875
    },
    {
      "entropy": 5.5865131378173825,
      "epoch": 0.5480990621665637,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.5138,
      "mean_token_accuracy": 0.18382047712802888,
      "num_tokens": 10882411.0,
      "step": 4880
    },
    {
      "entropy": 5.63999924659729,
      "epoch": 0.5486606390745212,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.6316,
      "mean_token_accuracy": 0.18017077893018724,
      "num_tokens": 10894116.0,
      "step": 4885
    },
    {
      "entropy": 5.560580778121948,
      "epoch": 0.5492222159824788,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.41,
      "mean_token_accuracy": 0.18437618613243104,
      "num_tokens": 10904514.0,
      "step": 4890
    },
    {
      "entropy": 5.554110431671143,
      "epoch": 0.5497837928904363,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.4572,
      "mean_token_accuracy": 0.18201902955770494,
      "num_tokens": 10915602.0,
      "step": 4895
    },
    {
      "entropy": 5.588807582855225,
      "epoch": 0.5503453697983939,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.4407,
      "mean_token_accuracy": 0.19006901681423188,
      "num_tokens": 10926172.0,
      "step": 4900
    },
    {
      "entropy": 5.675413513183594,
      "epoch": 0.5509069467063514,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.5649,
      "mean_token_accuracy": 0.17791180312633514,
      "num_tokens": 10938435.0,
      "step": 4905
    },
    {
      "entropy": 5.5930407524108885,
      "epoch": 0.551468523614309,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.433,
      "mean_token_accuracy": 0.18169896602630614,
      "num_tokens": 10949118.0,
      "step": 4910
    },
    {
      "entropy": 5.564508962631225,
      "epoch": 0.5520301005222665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.4879,
      "mean_token_accuracy": 0.1805576741695404,
      "num_tokens": 10960120.0,
      "step": 4915
    },
    {
      "entropy": 5.629130506515503,
      "epoch": 0.5525916774302241,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497802983376339,
      "loss": 5.4775,
      "mean_token_accuracy": 0.18393729776144027,
      "num_tokens": 10971210.0,
      "step": 4920
    },
    {
      "entropy": 5.595309782028198,
      "epoch": 0.5531532543381816,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.4569,
      "mean_token_accuracy": 0.19195475429296494,
      "num_tokens": 10983446.0,
      "step": 4925
    },
    {
      "entropy": 5.653829622268677,
      "epoch": 0.5537148312461392,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.589,
      "mean_token_accuracy": 0.17618100047111512,
      "num_tokens": 10994776.0,
      "step": 4930
    },
    {
      "entropy": 5.699292516708374,
      "epoch": 0.5542764081540967,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497786160684314,
      "loss": 5.4901,
      "mean_token_accuracy": 0.17770370990037918,
      "num_tokens": 11005280.0,
      "step": 4935
    },
    {
      "entropy": 5.538117599487305,
      "epoch": 0.5548379850620543,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.4795,
      "mean_token_accuracy": 0.17729776054620744,
      "num_tokens": 11017509.0,
      "step": 4940
    },
    {
      "entropy": 5.609230899810791,
      "epoch": 0.5553995619700118,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497774910088011,
      "loss": 5.5081,
      "mean_token_accuracy": 0.18174312859773636,
      "num_tokens": 11029087.0,
      "step": 4945
    },
    {
      "entropy": 5.669476509094238,
      "epoch": 0.5559611388779694,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.5614,
      "mean_token_accuracy": 0.17520818561315538,
      "num_tokens": 11041144.0,
      "step": 4950
    },
    {
      "entropy": 5.765043258666992,
      "epoch": 0.5565227157859269,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049776363111875,
      "loss": 5.6188,
      "mean_token_accuracy": 0.17266984283924103,
      "num_tokens": 11054050.0,
      "step": 4955
    },
    {
      "entropy": 5.576721334457398,
      "epoch": 0.5570842926938844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.4415,
      "mean_token_accuracy": 0.1856127619743347,
      "num_tokens": 11064257.0,
      "step": 4960
    },
    {
      "entropy": 5.647933101654052,
      "epoch": 0.5576458696018419,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.5591,
      "mean_token_accuracy": 0.18124807327985765,
      "num_tokens": 11075568.0,
      "step": 4965
    },
    {
      "entropy": 5.651572418212891,
      "epoch": 0.5582074465097995,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.5026,
      "mean_token_accuracy": 0.18101075142621995,
      "num_tokens": 11087078.0,
      "step": 4970
    },
    {
      "entropy": 5.726760768890381,
      "epoch": 0.558769023417757,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.6542,
      "mean_token_accuracy": 0.1728218376636505,
      "num_tokens": 11100655.0,
      "step": 4975
    },
    {
      "entropy": 5.670252847671509,
      "epoch": 0.5593306003257146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.5503,
      "mean_token_accuracy": 0.17902832329273224,
      "num_tokens": 11110675.0,
      "step": 4980
    },
    {
      "entropy": 5.616565752029419,
      "epoch": 0.5598921772336721,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.5456,
      "mean_token_accuracy": 0.18028604239225388,
      "num_tokens": 11122337.0,
      "step": 4985
    },
    {
      "entropy": 5.699768733978272,
      "epoch": 0.5604537541416297,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497723931310065,
      "loss": 5.4543,
      "mean_token_accuracy": 0.18633654564619065,
      "num_tokens": 11132672.0,
      "step": 4990
    },
    {
      "entropy": 5.606082248687744,
      "epoch": 0.5610153310495872,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.4983,
      "mean_token_accuracy": 0.18540901094675064,
      "num_tokens": 11143740.0,
      "step": 4995
    },
    {
      "entropy": 5.561510276794434,
      "epoch": 0.5615769079575448,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.4353,
      "mean_token_accuracy": 0.18260809630155564,
      "num_tokens": 11153722.0,
      "step": 5000
    },
    {
      "entropy": 5.659125947952271,
      "epoch": 0.5621384848655023,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.568,
      "mean_token_accuracy": 0.18011747300624847,
      "num_tokens": 11165042.0,
      "step": 5005
    },
    {
      "entropy": 5.6353111743927,
      "epoch": 0.5627000617734599,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.4419,
      "mean_token_accuracy": 0.18952011615037917,
      "num_tokens": 11176834.0,
      "step": 5010
    },
    {
      "entropy": 5.623426389694214,
      "epoch": 0.5632616386814174,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.5016,
      "mean_token_accuracy": 0.18498848527669906,
      "num_tokens": 11188442.0,
      "step": 5015
    },
    {
      "entropy": 5.576831293106079,
      "epoch": 0.563823215589375,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497689626327256,
      "loss": 5.4149,
      "mean_token_accuracy": 0.19040623009204866,
      "num_tokens": 11197855.0,
      "step": 5020
    },
    {
      "entropy": 5.597378063201904,
      "epoch": 0.5643847924973325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497683884012223,
      "loss": 5.5236,
      "mean_token_accuracy": 0.17814671993255615,
      "num_tokens": 11209300.0,
      "step": 5025
    },
    {
      "entropy": 5.604292917251587,
      "epoch": 0.5649463694052901,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.5237,
      "mean_token_accuracy": 0.17906069606542588,
      "num_tokens": 11222175.0,
      "step": 5030
    },
    {
      "entropy": 5.669887399673462,
      "epoch": 0.5655079463132476,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.5376,
      "mean_token_accuracy": 0.17941499203443528,
      "num_tokens": 11233277.0,
      "step": 5035
    },
    {
      "entropy": 5.647896575927734,
      "epoch": 0.5660695232212052,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.5599,
      "mean_token_accuracy": 0.18245493173599242,
      "num_tokens": 11245241.0,
      "step": 5040
    },
    {
      "entropy": 5.6755060195922855,
      "epoch": 0.5666311001291627,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.5661,
      "mean_token_accuracy": 0.18422424793243408,
      "num_tokens": 11257359.0,
      "step": 5045
    },
    {
      "entropy": 5.614133977890015,
      "epoch": 0.5671926770371203,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.5163,
      "mean_token_accuracy": 0.1842909574508667,
      "num_tokens": 11270651.0,
      "step": 5050
    },
    {
      "entropy": 5.607726955413819,
      "epoch": 0.5677542539450777,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497649281231274,
      "loss": 5.4707,
      "mean_token_accuracy": 0.1895418033003807,
      "num_tokens": 11280164.0,
      "step": 5055
    },
    {
      "entropy": 5.599555587768554,
      "epoch": 0.5683158308530353,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.4141,
      "mean_token_accuracy": 0.18119289427995683,
      "num_tokens": 11290156.0,
      "step": 5060
    },
    {
      "entropy": 5.612470483779907,
      "epoch": 0.5688774077609928,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.4759,
      "mean_token_accuracy": 0.18851063549518585,
      "num_tokens": 11301929.0,
      "step": 5065
    },
    {
      "entropy": 5.5531538963317875,
      "epoch": 0.5694389846689504,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.3883,
      "mean_token_accuracy": 0.19327545613050462,
      "num_tokens": 11312805.0,
      "step": 5070
    },
    {
      "entropy": 5.456476163864136,
      "epoch": 0.5700005615769079,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.2884,
      "mean_token_accuracy": 0.18960177898406982,
      "num_tokens": 11322316.0,
      "step": 5075
    },
    {
      "entropy": 5.5868700504302975,
      "epoch": 0.5705621384848655,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497620250624724,
      "loss": 5.4795,
      "mean_token_accuracy": 0.18839925825595855,
      "num_tokens": 11332994.0,
      "step": 5080
    },
    {
      "entropy": 5.628772020339966,
      "epoch": 0.571123715392823,
      "grad_norm": 0.84375,
      "learning_rate": 0.000497614423237728,
      "loss": 5.4692,
      "mean_token_accuracy": 0.17971829026937486,
      "num_tokens": 11344355.0,
      "step": 5085
    },
    {
      "entropy": 5.582499217987061,
      "epoch": 0.5716852923007806,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.5217,
      "mean_token_accuracy": 0.17927353084087372,
      "num_tokens": 11356275.0,
      "step": 5090
    },
    {
      "entropy": 5.596384048461914,
      "epoch": 0.5722468692087381,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.4939,
      "mean_token_accuracy": 0.18226930350065232,
      "num_tokens": 11368624.0,
      "step": 5095
    },
    {
      "entropy": 5.697848796844482,
      "epoch": 0.5728084461166957,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497596898548706,
      "loss": 5.6062,
      "mean_token_accuracy": 0.17915936410427094,
      "num_tokens": 11380520.0,
      "step": 5100
    },
    {
      "entropy": 5.58489089012146,
      "epoch": 0.5733700230246532,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18580701500177382,
      "num_tokens": 11392293.0,
      "step": 5105
    },
    {
      "entropy": 5.6542316436767575,
      "epoch": 0.5739315999326108,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.5717,
      "mean_token_accuracy": 0.17740507423877716,
      "num_tokens": 11403866.0,
      "step": 5110
    },
    {
      "entropy": 5.558068656921387,
      "epoch": 0.5744931768405683,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000497579310071541,
      "loss": 5.4008,
      "mean_token_accuracy": 0.18831710666418075,
      "num_tokens": 11415334.0,
      "step": 5115
    },
    {
      "entropy": 5.580965232849121,
      "epoch": 0.5750547537485259,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.5464,
      "mean_token_accuracy": 0.18406864255666733,
      "num_tokens": 11426840.0,
      "step": 5120
    },
    {
      "entropy": 5.704878950119019,
      "epoch": 0.5756163306564834,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.6915,
      "mean_token_accuracy": 0.17358105182647704,
      "num_tokens": 11438497.0,
      "step": 5125
    },
    {
      "entropy": 5.608832836151123,
      "epoch": 0.576177907564441,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.4116,
      "mean_token_accuracy": 0.1894557297229767,
      "num_tokens": 11448705.0,
      "step": 5130
    },
    {
      "entropy": 5.725475406646728,
      "epoch": 0.5767394844723985,
      "grad_norm": 0.875,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.6549,
      "mean_token_accuracy": 0.17674582004547118,
      "num_tokens": 11460610.0,
      "step": 5135
    },
    {
      "entropy": 5.615379667282104,
      "epoch": 0.5773010613803561,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.4249,
      "mean_token_accuracy": 0.18672376573085786,
      "num_tokens": 11471170.0,
      "step": 5140
    },
    {
      "entropy": 5.5798238754272464,
      "epoch": 0.5778626382883136,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.4144,
      "mean_token_accuracy": 0.1908004879951477,
      "num_tokens": 11482090.0,
      "step": 5145
    },
    {
      "entropy": 5.552998208999634,
      "epoch": 0.5784242151962711,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.5134,
      "mean_token_accuracy": 0.18054262697696685,
      "num_tokens": 11493650.0,
      "step": 5150
    },
    {
      "entropy": 5.626877212524414,
      "epoch": 0.5789857921042286,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.5028,
      "mean_token_accuracy": 0.1770804211497307,
      "num_tokens": 11505903.0,
      "step": 5155
    },
    {
      "entropy": 5.694047975540161,
      "epoch": 0.5795473690121862,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.5351,
      "mean_token_accuracy": 0.18056992888450624,
      "num_tokens": 11517194.0,
      "step": 5160
    },
    {
      "entropy": 5.645776748657227,
      "epoch": 0.5801089459201437,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.5483,
      "mean_token_accuracy": 0.187668114900589,
      "num_tokens": 11528683.0,
      "step": 5165
    },
    {
      "entropy": 5.619265031814575,
      "epoch": 0.5806705228281013,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.5718,
      "mean_token_accuracy": 0.18432468473911284,
      "num_tokens": 11541940.0,
      "step": 5170
    },
    {
      "entropy": 5.582946252822876,
      "epoch": 0.5812320997360588,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18226873576641084,
      "num_tokens": 11553045.0,
      "step": 5175
    },
    {
      "entropy": 5.567510271072388,
      "epoch": 0.5817936766440164,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.4232,
      "mean_token_accuracy": 0.19102198332548143,
      "num_tokens": 11564401.0,
      "step": 5180
    },
    {
      "entropy": 5.590689706802368,
      "epoch": 0.5823552535519739,
      "grad_norm": 0.875,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.4638,
      "mean_token_accuracy": 0.1833739846944809,
      "num_tokens": 11576424.0,
      "step": 5185
    },
    {
      "entropy": 5.660714054107666,
      "epoch": 0.5829168304599315,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00049749041104062,
      "loss": 5.4818,
      "mean_token_accuracy": 0.18013193905353547,
      "num_tokens": 11587853.0,
      "step": 5190
    },
    {
      "entropy": 5.668058681488037,
      "epoch": 0.583478407367889,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18619268983602524,
      "num_tokens": 11597938.0,
      "step": 5195
    },
    {
      "entropy": 5.704557037353515,
      "epoch": 0.5840399842758466,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.6392,
      "mean_token_accuracy": 0.18226669430732728,
      "num_tokens": 11610077.0,
      "step": 5200
    },
    {
      "entropy": 5.579728794097901,
      "epoch": 0.5846015611838041,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18503529876470565,
      "num_tokens": 11622079.0,
      "step": 5205
    },
    {
      "entropy": 5.523500871658325,
      "epoch": 0.5851631380917617,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.417,
      "mean_token_accuracy": 0.19386857897043228,
      "num_tokens": 11633047.0,
      "step": 5210
    },
    {
      "entropy": 5.549732303619384,
      "epoch": 0.5857247149997192,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.3688,
      "mean_token_accuracy": 0.18848908692598343,
      "num_tokens": 11643858.0,
      "step": 5215
    },
    {
      "entropy": 5.541607427597046,
      "epoch": 0.5862862919076768,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1883080154657364,
      "num_tokens": 11654550.0,
      "step": 5220
    },
    {
      "entropy": 5.5493570327758786,
      "epoch": 0.5868478688156343,
      "grad_norm": 0.84375,
      "learning_rate": 0.000497448379305403,
      "loss": 5.4274,
      "mean_token_accuracy": 0.18527335971593856,
      "num_tokens": 11665787.0,
      "step": 5225
    },
    {
      "entropy": 5.655500841140747,
      "epoch": 0.5874094457235919,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.5622,
      "mean_token_accuracy": 0.17959532588720323,
      "num_tokens": 11677717.0,
      "step": 5230
    },
    {
      "entropy": 5.619990444183349,
      "epoch": 0.5879710226315494,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497436306489741,
      "loss": 5.4358,
      "mean_token_accuracy": 0.18572364002466202,
      "num_tokens": 11687714.0,
      "step": 5235
    },
    {
      "entropy": 5.550149869918823,
      "epoch": 0.588532599539507,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.3355,
      "mean_token_accuracy": 0.19635276347398758,
      "num_tokens": 11698777.0,
      "step": 5240
    },
    {
      "entropy": 5.568590974807739,
      "epoch": 0.5890941764474645,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.4262,
      "mean_token_accuracy": 0.1932222217321396,
      "num_tokens": 11710922.0,
      "step": 5245
    },
    {
      "entropy": 5.575196647644043,
      "epoch": 0.589655753355422,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.4741,
      "mean_token_accuracy": 0.18178150206804275,
      "num_tokens": 11721450.0,
      "step": 5250
    },
    {
      "entropy": 5.520800971984864,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.3861,
      "mean_token_accuracy": 0.18722872734069823,
      "num_tokens": 11732458.0,
      "step": 5255
    },
    {
      "entropy": 5.522866821289062,
      "epoch": 0.5907789071713371,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.35,
      "mean_token_accuracy": 0.19123259037733079,
      "num_tokens": 11742649.0,
      "step": 5260
    },
    {
      "entropy": 5.577361011505127,
      "epoch": 0.5913404840792946,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497399918069926,
      "loss": 5.3894,
      "mean_token_accuracy": 0.19077312350273132,
      "num_tokens": 11753811.0,
      "step": 5265
    },
    {
      "entropy": 5.621190118789673,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.4925,
      "mean_token_accuracy": 0.178723706305027,
      "num_tokens": 11763965.0,
      "step": 5270
    },
    {
      "entropy": 5.614044523239135,
      "epoch": 0.5924636378952097,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.4999,
      "mean_token_accuracy": 0.18876443058252335,
      "num_tokens": 11774323.0,
      "step": 5275
    },
    {
      "entropy": 5.628261756896973,
      "epoch": 0.5930252148031673,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.4729,
      "mean_token_accuracy": 0.18141622245311737,
      "num_tokens": 11786157.0,
      "step": 5280
    },
    {
      "entropy": 5.611000013351441,
      "epoch": 0.5935867917111248,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.45,
      "mean_token_accuracy": 0.19105252325534822,
      "num_tokens": 11797344.0,
      "step": 5285
    },
    {
      "entropy": 5.448312091827392,
      "epoch": 0.5941483686190824,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.3984,
      "mean_token_accuracy": 0.18638614416122437,
      "num_tokens": 11808149.0,
      "step": 5290
    },
    {
      "entropy": 5.552359485626221,
      "epoch": 0.5947099455270399,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.414,
      "mean_token_accuracy": 0.1930566906929016,
      "num_tokens": 11820203.0,
      "step": 5295
    },
    {
      "entropy": 5.617059373855591,
      "epoch": 0.5952715224349975,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.4928,
      "mean_token_accuracy": 0.18311173915863038,
      "num_tokens": 11830172.0,
      "step": 5300
    },
    {
      "entropy": 5.611087846755981,
      "epoch": 0.595833099342955,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.4504,
      "mean_token_accuracy": 0.1895171508193016,
      "num_tokens": 11840203.0,
      "step": 5305
    },
    {
      "entropy": 5.59495210647583,
      "epoch": 0.5963946762509126,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.4147,
      "mean_token_accuracy": 0.18466875255107879,
      "num_tokens": 11851002.0,
      "step": 5310
    },
    {
      "entropy": 5.571983194351196,
      "epoch": 0.5969562531588701,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.4066,
      "mean_token_accuracy": 0.1829525947570801,
      "num_tokens": 11861559.0,
      "step": 5315
    },
    {
      "entropy": 5.55874285697937,
      "epoch": 0.5975178300668277,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.491,
      "mean_token_accuracy": 0.18227005898952484,
      "num_tokens": 11872535.0,
      "step": 5320
    },
    {
      "entropy": 5.640175628662109,
      "epoch": 0.5980794069747852,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.4932,
      "mean_token_accuracy": 0.18847439289093018,
      "num_tokens": 11884308.0,
      "step": 5325
    },
    {
      "entropy": 5.6690717220306395,
      "epoch": 0.5986409838827428,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.5659,
      "mean_token_accuracy": 0.18065213561058044,
      "num_tokens": 11896136.0,
      "step": 5330
    },
    {
      "entropy": 5.595011615753174,
      "epoch": 0.5992025607907003,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497314020447805,
      "loss": 5.4011,
      "mean_token_accuracy": 0.1973426029086113,
      "num_tokens": 11906497.0,
      "step": 5335
    },
    {
      "entropy": 5.6279065132141115,
      "epoch": 0.5997641376986579,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.5099,
      "mean_token_accuracy": 0.1770203560590744,
      "num_tokens": 11917872.0,
      "step": 5340
    },
    {
      "entropy": 5.663007164001465,
      "epoch": 0.6003257146066153,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497301636089344,
      "loss": 5.5404,
      "mean_token_accuracy": 0.18449182361364364,
      "num_tokens": 11929727.0,
      "step": 5345
    },
    {
      "entropy": 5.631331539154052,
      "epoch": 0.6008872915145729,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.3267,
      "mean_token_accuracy": 0.18630826622247695,
      "num_tokens": 11939996.0,
      "step": 5350
    },
    {
      "entropy": 5.533454132080078,
      "epoch": 0.6014488684225304,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.4119,
      "mean_token_accuracy": 0.19154259711503982,
      "num_tokens": 11950683.0,
      "step": 5355
    },
    {
      "entropy": 5.5050122261047365,
      "epoch": 0.602010445330488,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.4089,
      "mean_token_accuracy": 0.18465373665094376,
      "num_tokens": 11962447.0,
      "step": 5360
    },
    {
      "entropy": 5.640342664718628,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.4919,
      "mean_token_accuracy": 0.182956001162529,
      "num_tokens": 11972740.0,
      "step": 5365
    },
    {
      "entropy": 5.555352592468262,
      "epoch": 0.6031335991464031,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.3702,
      "mean_token_accuracy": 0.1916798397898674,
      "num_tokens": 11983292.0,
      "step": 5370
    },
    {
      "entropy": 5.674102401733398,
      "epoch": 0.6036951760543606,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.5922,
      "mean_token_accuracy": 0.18003977239131927,
      "num_tokens": 11996384.0,
      "step": 5375
    },
    {
      "entropy": 5.696470832824707,
      "epoch": 0.6042567529623182,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.5061,
      "mean_token_accuracy": 0.19199381321668624,
      "num_tokens": 12006938.0,
      "step": 5380
    },
    {
      "entropy": 5.482018661499024,
      "epoch": 0.6048183298702757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.3386,
      "mean_token_accuracy": 0.18582624047994614,
      "num_tokens": 12018729.0,
      "step": 5385
    },
    {
      "entropy": 5.7195052146911625,
      "epoch": 0.6053799067782333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.7041,
      "mean_token_accuracy": 0.17384667247533797,
      "num_tokens": 12029625.0,
      "step": 5390
    },
    {
      "entropy": 5.56141209602356,
      "epoch": 0.6059414836861908,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.3895,
      "mean_token_accuracy": 0.18369418978691102,
      "num_tokens": 12040243.0,
      "step": 5395
    },
    {
      "entropy": 5.7068993091583256,
      "epoch": 0.6065030605941484,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.671,
      "mean_token_accuracy": 0.175175179541111,
      "num_tokens": 12051961.0,
      "step": 5400
    },
    {
      "entropy": 5.6262939453125,
      "epoch": 0.6070646375021059,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.5378,
      "mean_token_accuracy": 0.1859945684671402,
      "num_tokens": 12063969.0,
      "step": 5405
    },
    {
      "entropy": 5.647840547561645,
      "epoch": 0.6076262144100635,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.4178,
      "mean_token_accuracy": 0.18723357021808623,
      "num_tokens": 12075682.0,
      "step": 5410
    },
    {
      "entropy": 5.672305297851563,
      "epoch": 0.608187791318021,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.5541,
      "mean_token_accuracy": 0.18671936988830568,
      "num_tokens": 12086595.0,
      "step": 5415
    },
    {
      "entropy": 5.586926698684692,
      "epoch": 0.6087493682259786,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.4268,
      "mean_token_accuracy": 0.18619488626718522,
      "num_tokens": 12099421.0,
      "step": 5420
    },
    {
      "entropy": 5.633434104919433,
      "epoch": 0.6093109451339361,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.5427,
      "mean_token_accuracy": 0.1840139091014862,
      "num_tokens": 12109680.0,
      "step": 5425
    },
    {
      "entropy": 5.543154954910278,
      "epoch": 0.6098725220418937,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.4224,
      "mean_token_accuracy": 0.19202370345592498,
      "num_tokens": 12121196.0,
      "step": 5430
    },
    {
      "entropy": 5.569849824905395,
      "epoch": 0.6104340989498512,
      "grad_norm": 0.828125,
      "learning_rate": 0.000497188902983013,
      "loss": 5.3755,
      "mean_token_accuracy": 0.17969895005226136,
      "num_tokens": 12132262.0,
      "step": 5435
    },
    {
      "entropy": 5.629948139190674,
      "epoch": 0.6109956758578087,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.5264,
      "mean_token_accuracy": 0.18493406623601913,
      "num_tokens": 12143747.0,
      "step": 5440
    },
    {
      "entropy": 5.7156328678131105,
      "epoch": 0.6115572527657662,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.6394,
      "mean_token_accuracy": 0.1741531029343605,
      "num_tokens": 12155887.0,
      "step": 5445
    },
    {
      "entropy": 5.5822920322418215,
      "epoch": 0.6121188296737238,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.4178,
      "mean_token_accuracy": 0.18080265671014786,
      "num_tokens": 12166718.0,
      "step": 5450
    },
    {
      "entropy": 5.620777988433838,
      "epoch": 0.6126804065816813,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.5632,
      "mean_token_accuracy": 0.1833667978644371,
      "num_tokens": 12178540.0,
      "step": 5455
    },
    {
      "entropy": 5.719579219818115,
      "epoch": 0.6132419834896389,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.6076,
      "mean_token_accuracy": 0.1786387026309967,
      "num_tokens": 12189820.0,
      "step": 5460
    },
    {
      "entropy": 5.616104698181152,
      "epoch": 0.6138035603975964,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.4899,
      "mean_token_accuracy": 0.18555515855550767,
      "num_tokens": 12201679.0,
      "step": 5465
    },
    {
      "entropy": 5.6018133640289305,
      "epoch": 0.614365137305554,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.5228,
      "mean_token_accuracy": 0.1803461119532585,
      "num_tokens": 12213793.0,
      "step": 5470
    },
    {
      "entropy": 5.688950681686402,
      "epoch": 0.6149267142135115,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.6075,
      "mean_token_accuracy": 0.1852038875222206,
      "num_tokens": 12224519.0,
      "step": 5475
    },
    {
      "entropy": 5.766947793960571,
      "epoch": 0.6154882911214691,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.5854,
      "mean_token_accuracy": 0.1751009479165077,
      "num_tokens": 12236401.0,
      "step": 5480
    },
    {
      "entropy": 5.640927457809449,
      "epoch": 0.6160498680294266,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.607,
      "mean_token_accuracy": 0.17650003284215926,
      "num_tokens": 12247670.0,
      "step": 5485
    },
    {
      "entropy": 5.5905811309814455,
      "epoch": 0.6166114449373842,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.4792,
      "mean_token_accuracy": 0.18335412591695785,
      "num_tokens": 12258877.0,
      "step": 5490
    },
    {
      "entropy": 5.609956502914429,
      "epoch": 0.6171730218453417,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.4556,
      "mean_token_accuracy": 0.18799896836280822,
      "num_tokens": 12268507.0,
      "step": 5495
    },
    {
      "entropy": 5.591432285308838,
      "epoch": 0.6177345987532993,
      "grad_norm": 0.828125,
      "learning_rate": 0.000497106059033725,
      "loss": 5.4415,
      "mean_token_accuracy": 0.18227818310260774,
      "num_tokens": 12280420.0,
      "step": 5500
    },
    {
      "entropy": 5.568288421630859,
      "epoch": 0.6182961756612568,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.4455,
      "mean_token_accuracy": 0.1805670827627182,
      "num_tokens": 12290876.0,
      "step": 5505
    },
    {
      "entropy": 5.564266633987427,
      "epoch": 0.6188577525692144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.346,
      "mean_token_accuracy": 0.1907796800136566,
      "num_tokens": 12301831.0,
      "step": 5510
    },
    {
      "entropy": 5.4954017162322994,
      "epoch": 0.6194193294771719,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497086771456353,
      "loss": 5.3746,
      "mean_token_accuracy": 0.18919067233800888,
      "num_tokens": 12312678.0,
      "step": 5515
    },
    {
      "entropy": 5.676490306854248,
      "epoch": 0.6199809063851295,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.5504,
      "mean_token_accuracy": 0.17842403501272203,
      "num_tokens": 12322905.0,
      "step": 5520
    },
    {
      "entropy": 5.610180139541626,
      "epoch": 0.620542483293087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.4105,
      "mean_token_accuracy": 0.18521541357040405,
      "num_tokens": 12333091.0,
      "step": 5525
    },
    {
      "entropy": 5.595136499404907,
      "epoch": 0.6211040602010446,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.4231,
      "mean_token_accuracy": 0.18607301265001297,
      "num_tokens": 12346310.0,
      "step": 5530
    },
    {
      "entropy": 5.510883331298828,
      "epoch": 0.621665637109002,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.3426,
      "mean_token_accuracy": 0.18486004024744035,
      "num_tokens": 12357138.0,
      "step": 5535
    },
    {
      "entropy": 5.643247938156128,
      "epoch": 0.6222272140169596,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.434,
      "mean_token_accuracy": 0.19092607945203782,
      "num_tokens": 12367382.0,
      "step": 5540
    },
    {
      "entropy": 5.54682765007019,
      "epoch": 0.6227887909249171,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.4278,
      "mean_token_accuracy": 0.17990480065345765,
      "num_tokens": 12380357.0,
      "step": 5545
    },
    {
      "entropy": 5.649844694137573,
      "epoch": 0.6233503678328747,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.6104,
      "mean_token_accuracy": 0.17453047633171082,
      "num_tokens": 12392187.0,
      "step": 5550
    },
    {
      "entropy": 5.642471170425415,
      "epoch": 0.6239119447408322,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.4626,
      "mean_token_accuracy": 0.1919781044125557,
      "num_tokens": 12402924.0,
      "step": 5555
    },
    {
      "entropy": 5.690981531143189,
      "epoch": 0.6244735216487898,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497028526894682,
      "loss": 5.6001,
      "mean_token_accuracy": 0.18040583431720733,
      "num_tokens": 12413554.0,
      "step": 5560
    },
    {
      "entropy": 5.594972753524781,
      "epoch": 0.6250350985567473,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.4719,
      "mean_token_accuracy": 0.18820811361074447,
      "num_tokens": 12424525.0,
      "step": 5565
    },
    {
      "entropy": 5.5729231357574465,
      "epoch": 0.6255966754647049,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.4757,
      "mean_token_accuracy": 0.18470598757266998,
      "num_tokens": 12436405.0,
      "step": 5570
    },
    {
      "entropy": 5.630585432052612,
      "epoch": 0.6261582523726624,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.5239,
      "mean_token_accuracy": 0.1797418400645256,
      "num_tokens": 12446909.0,
      "step": 5575
    },
    {
      "entropy": 5.632023048400879,
      "epoch": 0.62671982928062,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.5103,
      "mean_token_accuracy": 0.1853021651506424,
      "num_tokens": 12458739.0,
      "step": 5580
    },
    {
      "entropy": 5.576176357269287,
      "epoch": 0.6272814061885775,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.4511,
      "mean_token_accuracy": 0.18338645100593567,
      "num_tokens": 12469529.0,
      "step": 5585
    },
    {
      "entropy": 5.673983478546143,
      "epoch": 0.6278429830965351,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.4896,
      "mean_token_accuracy": 0.17851488143205643,
      "num_tokens": 12480067.0,
      "step": 5590
    },
    {
      "entropy": 5.605591201782227,
      "epoch": 0.6284045600044926,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.3631,
      "mean_token_accuracy": 0.19283511638641357,
      "num_tokens": 12491043.0,
      "step": 5595
    },
    {
      "entropy": 5.49311146736145,
      "epoch": 0.6289661369124502,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.2978,
      "mean_token_accuracy": 0.19157171100378037,
      "num_tokens": 12501852.0,
      "step": 5600
    },
    {
      "entropy": 5.592472171783447,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.6187,
      "mean_token_accuracy": 0.17990778535604476,
      "num_tokens": 12512696.0,
      "step": 5605
    },
    {
      "entropy": 5.623296737670898,
      "epoch": 0.6300892907283653,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.456,
      "mean_token_accuracy": 0.18712275922298433,
      "num_tokens": 12523643.0,
      "step": 5610
    },
    {
      "entropy": 5.598635911941528,
      "epoch": 0.6306508676363228,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.4289,
      "mean_token_accuracy": 0.18764980584383012,
      "num_tokens": 12534827.0,
      "step": 5615
    },
    {
      "entropy": 5.46979455947876,
      "epoch": 0.6312124445442804,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.3034,
      "mean_token_accuracy": 0.1917729213833809,
      "num_tokens": 12545977.0,
      "step": 5620
    },
    {
      "entropy": 5.4649755477905275,
      "epoch": 0.6317740214522379,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.3726,
      "mean_token_accuracy": 0.18026579469442366,
      "num_tokens": 12555890.0,
      "step": 5625
    },
    {
      "entropy": 5.544844627380371,
      "epoch": 0.6323355983601955,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.4302,
      "mean_token_accuracy": 0.193123060464859,
      "num_tokens": 12566412.0,
      "step": 5630
    },
    {
      "entropy": 5.634961366653442,
      "epoch": 0.6328971752681529,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.4747,
      "mean_token_accuracy": 0.18541150838136672,
      "num_tokens": 12577175.0,
      "step": 5635
    },
    {
      "entropy": 5.53953857421875,
      "epoch": 0.6334587521761105,
      "grad_norm": 0.8125,
      "learning_rate": 0.000496923567209739,
      "loss": 5.3978,
      "mean_token_accuracy": 0.18610223531723022,
      "num_tokens": 12588760.0,
      "step": 5640
    },
    {
      "entropy": 5.5577116966247555,
      "epoch": 0.634020329084068,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.4673,
      "mean_token_accuracy": 0.18638110607862474,
      "num_tokens": 12599899.0,
      "step": 5645
    },
    {
      "entropy": 5.646306705474854,
      "epoch": 0.6345819059920256,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.4182,
      "mean_token_accuracy": 0.19530787616968154,
      "num_tokens": 12610836.0,
      "step": 5650
    },
    {
      "entropy": 5.538104391098022,
      "epoch": 0.6351434828999831,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.3832,
      "mean_token_accuracy": 0.19196907579898834,
      "num_tokens": 12621855.0,
      "step": 5655
    },
    {
      "entropy": 5.597290754318237,
      "epoch": 0.6357050598079407,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.464,
      "mean_token_accuracy": 0.18715441077947617,
      "num_tokens": 12632964.0,
      "step": 5660
    },
    {
      "entropy": 5.509879541397095,
      "epoch": 0.6362666367158982,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000496890396295883,
      "loss": 5.3246,
      "mean_token_accuracy": 0.19128041863441467,
      "num_tokens": 12643717.0,
      "step": 5665
    },
    {
      "entropy": 5.588350248336792,
      "epoch": 0.6368282136238558,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.4352,
      "mean_token_accuracy": 0.1833639457821846,
      "num_tokens": 12654491.0,
      "step": 5670
    },
    {
      "entropy": 5.574241399765015,
      "epoch": 0.6373897905318133,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.4882,
      "mean_token_accuracy": 0.18664699345827102,
      "num_tokens": 12666128.0,
      "step": 5675
    },
    {
      "entropy": 5.667492341995239,
      "epoch": 0.6379513674397709,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.5725,
      "mean_token_accuracy": 0.1784846618771553,
      "num_tokens": 12677957.0,
      "step": 5680
    },
    {
      "entropy": 5.590038394927978,
      "epoch": 0.6385129443477284,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.4011,
      "mean_token_accuracy": 0.18758929818868636,
      "num_tokens": 12687964.0,
      "step": 5685
    },
    {
      "entropy": 5.492672872543335,
      "epoch": 0.639074521255686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.4048,
      "mean_token_accuracy": 0.18345636129379272,
      "num_tokens": 12698805.0,
      "step": 5690
    },
    {
      "entropy": 5.616511344909668,
      "epoch": 0.6396360981636435,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.4221,
      "mean_token_accuracy": 0.18310350477695464,
      "num_tokens": 12709942.0,
      "step": 5695
    },
    {
      "entropy": 5.685959434509277,
      "epoch": 0.6401976750716011,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.4875,
      "mean_token_accuracy": 0.18588088899850846,
      "num_tokens": 12721856.0,
      "step": 5700
    },
    {
      "entropy": 5.478527450561524,
      "epoch": 0.6407592519795586,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.3205,
      "mean_token_accuracy": 0.18528069853782653,
      "num_tokens": 12732619.0,
      "step": 5705
    },
    {
      "entropy": 5.545958375930786,
      "epoch": 0.6413208288875162,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.4237,
      "mean_token_accuracy": 0.18862492889165877,
      "num_tokens": 12743478.0,
      "step": 5710
    },
    {
      "entropy": 5.532248306274414,
      "epoch": 0.6418824057954737,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.3664,
      "mean_token_accuracy": 0.18406943678855897,
      "num_tokens": 12754761.0,
      "step": 5715
    },
    {
      "entropy": 5.665899085998535,
      "epoch": 0.6424439827034313,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.4834,
      "mean_token_accuracy": 0.18930073082447052,
      "num_tokens": 12765803.0,
      "step": 5720
    },
    {
      "entropy": 5.574135589599609,
      "epoch": 0.6430055596113888,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.5096,
      "mean_token_accuracy": 0.18169724941253662,
      "num_tokens": 12777882.0,
      "step": 5725
    },
    {
      "entropy": 5.643469667434692,
      "epoch": 0.6435671365193463,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.5135,
      "mean_token_accuracy": 0.18031821995973588,
      "num_tokens": 12788268.0,
      "step": 5730
    },
    {
      "entropy": 5.600995302200317,
      "epoch": 0.6441287134273038,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.503,
      "mean_token_accuracy": 0.18759284317493438,
      "num_tokens": 12800506.0,
      "step": 5735
    },
    {
      "entropy": 5.60623836517334,
      "epoch": 0.6446902903352614,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18512337803840637,
      "num_tokens": 12813040.0,
      "step": 5740
    },
    {
      "entropy": 5.637520360946655,
      "epoch": 0.6452518672432189,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.5197,
      "mean_token_accuracy": 0.18802948147058487,
      "num_tokens": 12825534.0,
      "step": 5745
    },
    {
      "entropy": 5.576848983764648,
      "epoch": 0.6458134441511765,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.3975,
      "mean_token_accuracy": 0.1857204630970955,
      "num_tokens": 12835882.0,
      "step": 5750
    },
    {
      "entropy": 5.585302686691284,
      "epoch": 0.646375021059134,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.6039,
      "mean_token_accuracy": 0.17597048431634904,
      "num_tokens": 12847663.0,
      "step": 5755
    },
    {
      "entropy": 5.744038200378418,
      "epoch": 0.6469365979670916,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.7125,
      "mean_token_accuracy": 0.17757612019777297,
      "num_tokens": 12860789.0,
      "step": 5760
    },
    {
      "entropy": 5.647675704956055,
      "epoch": 0.6474981748750491,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.4558,
      "mean_token_accuracy": 0.18649709075689316,
      "num_tokens": 12872509.0,
      "step": 5765
    },
    {
      "entropy": 5.555054569244385,
      "epoch": 0.6480597517830067,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.4534,
      "mean_token_accuracy": 0.18682979196310043,
      "num_tokens": 12884321.0,
      "step": 5770
    },
    {
      "entropy": 5.431756973266602,
      "epoch": 0.6486213286909642,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19240241646766662,
      "num_tokens": 12894559.0,
      "step": 5775
    },
    {
      "entropy": 5.564483499526977,
      "epoch": 0.6491829055989218,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.4319,
      "mean_token_accuracy": 0.18994510620832444,
      "num_tokens": 12904934.0,
      "step": 5780
    },
    {
      "entropy": 5.70449275970459,
      "epoch": 0.6497444825068793,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000496728717732762,
      "loss": 5.555,
      "mean_token_accuracy": 0.1803714379668236,
      "num_tokens": 12916551.0,
      "step": 5785
    },
    {
      "entropy": 5.609445714950562,
      "epoch": 0.6503060594148369,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.454,
      "mean_token_accuracy": 0.18270530104637145,
      "num_tokens": 12928097.0,
      "step": 5790
    },
    {
      "entropy": 5.558883285522461,
      "epoch": 0.6508676363227944,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19458029717206954,
      "num_tokens": 12939161.0,
      "step": 5795
    },
    {
      "entropy": 5.479016637802124,
      "epoch": 0.651429213230752,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.3878,
      "mean_token_accuracy": 0.18893582224845887,
      "num_tokens": 12950862.0,
      "step": 5800
    },
    {
      "entropy": 5.587749862670899,
      "epoch": 0.6519907901387095,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.4337,
      "mean_token_accuracy": 0.18156342059373856,
      "num_tokens": 12962333.0,
      "step": 5805
    },
    {
      "entropy": 5.59279727935791,
      "epoch": 0.6525523670466671,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.3942,
      "mean_token_accuracy": 0.18726360201835632,
      "num_tokens": 12973136.0,
      "step": 5810
    },
    {
      "entropy": 5.587766170501709,
      "epoch": 0.6531139439546246,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.4462,
      "mean_token_accuracy": 0.18316376209259033,
      "num_tokens": 12983602.0,
      "step": 5815
    },
    {
      "entropy": 5.575126361846924,
      "epoch": 0.6536755208625822,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.3571,
      "mean_token_accuracy": 0.19102459400892258,
      "num_tokens": 12994129.0,
      "step": 5820
    },
    {
      "entropy": 5.565567350387573,
      "epoch": 0.6542370977705396,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.4066,
      "mean_token_accuracy": 0.18423057496547698,
      "num_tokens": 13005071.0,
      "step": 5825
    },
    {
      "entropy": 5.66208438873291,
      "epoch": 0.6547986746784972,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.5974,
      "mean_token_accuracy": 0.18425355702638627,
      "num_tokens": 13016202.0,
      "step": 5830
    },
    {
      "entropy": 5.645590591430664,
      "epoch": 0.6553602515864547,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.5697,
      "mean_token_accuracy": 0.1739979788661003,
      "num_tokens": 13028468.0,
      "step": 5835
    },
    {
      "entropy": 5.606764221191407,
      "epoch": 0.6559218284944123,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496653255842169,
      "loss": 5.4487,
      "mean_token_accuracy": 0.18786526769399642,
      "num_tokens": 13039667.0,
      "step": 5840
    },
    {
      "entropy": 5.617080545425415,
      "epoch": 0.6564834054023698,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.5264,
      "mean_token_accuracy": 0.17802227288484573,
      "num_tokens": 13051583.0,
      "step": 5845
    },
    {
      "entropy": 5.606671905517578,
      "epoch": 0.6570449823103274,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.4708,
      "mean_token_accuracy": 0.17936455011367797,
      "num_tokens": 13062401.0,
      "step": 5850
    },
    {
      "entropy": 5.533967208862305,
      "epoch": 0.6576065592182849,
      "grad_norm": 0.875,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1948936700820923,
      "num_tokens": 13074219.0,
      "step": 5855
    },
    {
      "entropy": 5.618377780914306,
      "epoch": 0.6581681361262425,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.5585,
      "mean_token_accuracy": 0.18522871136665345,
      "num_tokens": 13086073.0,
      "step": 5860
    },
    {
      "entropy": 5.6113865852355955,
      "epoch": 0.6587297130342,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.4566,
      "mean_token_accuracy": 0.18118618726730346,
      "num_tokens": 13096340.0,
      "step": 5865
    },
    {
      "entropy": 5.520273065567016,
      "epoch": 0.6592912899421576,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.433,
      "mean_token_accuracy": 0.18932820707559586,
      "num_tokens": 13107273.0,
      "step": 5870
    },
    {
      "entropy": 5.5223218441009525,
      "epoch": 0.6598528668501151,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.4083,
      "mean_token_accuracy": 0.17953830510377883,
      "num_tokens": 13117388.0,
      "step": 5875
    },
    {
      "entropy": 5.585290336608887,
      "epoch": 0.6604144437580727,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496597838078686,
      "loss": 5.5291,
      "mean_token_accuracy": 0.18056883960962294,
      "num_tokens": 13128720.0,
      "step": 5880
    },
    {
      "entropy": 5.605383491516113,
      "epoch": 0.6609760206660302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496590879105005,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18524353951215744,
      "num_tokens": 13139405.0,
      "step": 5885
    },
    {
      "entropy": 5.6540992736816404,
      "epoch": 0.6615375975739878,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.5396,
      "mean_token_accuracy": 0.18290454149246216,
      "num_tokens": 13151408.0,
      "step": 5890
    },
    {
      "entropy": 5.585953855514527,
      "epoch": 0.6620991744819453,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.4388,
      "mean_token_accuracy": 0.18750968277454377,
      "num_tokens": 13162592.0,
      "step": 5895
    },
    {
      "entropy": 5.613091945648193,
      "epoch": 0.6626607513899029,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.4654,
      "mean_token_accuracy": 0.18831417411565782,
      "num_tokens": 13173040.0,
      "step": 5900
    },
    {
      "entropy": 5.580550241470337,
      "epoch": 0.6632223282978604,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.5408,
      "mean_token_accuracy": 0.18459512442350387,
      "num_tokens": 13185475.0,
      "step": 5905
    },
    {
      "entropy": 5.546143627166748,
      "epoch": 0.663783905205818,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.4001,
      "mean_token_accuracy": 0.19337214678525924,
      "num_tokens": 13196481.0,
      "step": 5910
    },
    {
      "entropy": 5.508285856246948,
      "epoch": 0.6643454821137755,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.2867,
      "mean_token_accuracy": 0.1948182225227356,
      "num_tokens": 13207947.0,
      "step": 5915
    },
    {
      "entropy": 5.468513250350952,
      "epoch": 0.664907059021733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.338,
      "mean_token_accuracy": 0.19859456717967988,
      "num_tokens": 13219392.0,
      "step": 5920
    },
    {
      "entropy": 5.487781715393067,
      "epoch": 0.6654686359296905,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.3837,
      "mean_token_accuracy": 0.19404632598161697,
      "num_tokens": 13231556.0,
      "step": 5925
    },
    {
      "entropy": 5.560096693038941,
      "epoch": 0.6660302128376481,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.3223,
      "mean_token_accuracy": 0.18837640136480333,
      "num_tokens": 13241287.0,
      "step": 5930
    },
    {
      "entropy": 5.512104034423828,
      "epoch": 0.6665917897456056,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.3776,
      "mean_token_accuracy": 0.18145681619644166,
      "num_tokens": 13254120.0,
      "step": 5935
    },
    {
      "entropy": 5.558088684082032,
      "epoch": 0.6671533666535632,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.4523,
      "mean_token_accuracy": 0.1826157882809639,
      "num_tokens": 13265745.0,
      "step": 5940
    },
    {
      "entropy": 5.666581439971924,
      "epoch": 0.6677149435615207,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000496506821152345,
      "loss": 5.512,
      "mean_token_accuracy": 0.18324743062257767,
      "num_tokens": 13277791.0,
      "step": 5945
    },
    {
      "entropy": 5.560856914520263,
      "epoch": 0.6682765204694783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18288409113883972,
      "num_tokens": 13288078.0,
      "step": 5950
    },
    {
      "entropy": 5.54308385848999,
      "epoch": 0.6688380973774358,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000496492712743046,
      "loss": 5.4164,
      "mean_token_accuracy": 0.18916729539632798,
      "num_tokens": 13300547.0,
      "step": 5955
    },
    {
      "entropy": 5.574636030197143,
      "epoch": 0.6693996742853934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.4712,
      "mean_token_accuracy": 0.1844284325838089,
      "num_tokens": 13310339.0,
      "step": 5960
    },
    {
      "entropy": 5.649972581863404,
      "epoch": 0.6699612511933509,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.4652,
      "mean_token_accuracy": 0.18339674770832062,
      "num_tokens": 13321294.0,
      "step": 5965
    },
    {
      "entropy": 5.5962268829345705,
      "epoch": 0.6705228281013085,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.4876,
      "mean_token_accuracy": 0.18522342145442963,
      "num_tokens": 13332608.0,
      "step": 5970
    },
    {
      "entropy": 5.588433122634887,
      "epoch": 0.671084405009266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.4135,
      "mean_token_accuracy": 0.1959644690155983,
      "num_tokens": 13344169.0,
      "step": 5975
    },
    {
      "entropy": 5.4729576587677,
      "epoch": 0.6716459819172236,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.3822,
      "mean_token_accuracy": 0.18695180267095565,
      "num_tokens": 13355713.0,
      "step": 5980
    },
    {
      "entropy": 5.600048303604126,
      "epoch": 0.6722075588251811,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.4743,
      "mean_token_accuracy": 0.1833113133907318,
      "num_tokens": 13366644.0,
      "step": 5985
    },
    {
      "entropy": 5.611146450042725,
      "epoch": 0.6727691357331387,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.447,
      "mean_token_accuracy": 0.18404945284128188,
      "num_tokens": 13378125.0,
      "step": 5990
    },
    {
      "entropy": 5.538123226165771,
      "epoch": 0.6733307126410962,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.4918,
      "mean_token_accuracy": 0.18231723010540007,
      "num_tokens": 13389681.0,
      "step": 5995
    },
    {
      "entropy": 5.638032388687134,
      "epoch": 0.6738922895490538,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.4689,
      "mean_token_accuracy": 0.18596884310245515,
      "num_tokens": 13399814.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.401589800347856,
      "eval_loss": 5.4581098556518555,
      "eval_mean_token_accuracy": 0.192657116930264,
      "eval_num_tokens": 13399814.0,
      "eval_runtime": 18.2338,
      "eval_samples_per_second": 1700.857,
      "eval_steps_per_second": 212.628,
      "step": 6000
    },
    {
      "entropy": 5.646923589706421,
      "epoch": 0.6744538664570113,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.4957,
      "mean_token_accuracy": 0.18415769040584565,
      "num_tokens": 13410614.0,
      "step": 6005
    },
    {
      "entropy": 5.57463698387146,
      "epoch": 0.6750154433649689,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000496414612292672,
      "loss": 5.5008,
      "mean_token_accuracy": 0.1825885832309723,
      "num_tokens": 13421392.0,
      "step": 6010
    },
    {
      "entropy": 5.6496422290802,
      "epoch": 0.6755770202729264,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.5291,
      "mean_token_accuracy": 0.18065254390239716,
      "num_tokens": 13432004.0,
      "step": 6015
    },
    {
      "entropy": 5.580326557159424,
      "epoch": 0.6761385971808839,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.3787,
      "mean_token_accuracy": 0.1912464365363121,
      "num_tokens": 13442450.0,
      "step": 6020
    },
    {
      "entropy": 5.587422704696655,
      "epoch": 0.6767001740888414,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.4337,
      "mean_token_accuracy": 0.18147221207618713,
      "num_tokens": 13453848.0,
      "step": 6025
    },
    {
      "entropy": 5.537183475494385,
      "epoch": 0.677261750996799,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.4354,
      "mean_token_accuracy": 0.18918716460466384,
      "num_tokens": 13465119.0,
      "step": 6030
    },
    {
      "entropy": 5.525205278396607,
      "epoch": 0.6778233279047565,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.4136,
      "mean_token_accuracy": 0.1885279580950737,
      "num_tokens": 13475937.0,
      "step": 6035
    },
    {
      "entropy": 5.559143924713135,
      "epoch": 0.6783849048127141,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.399,
      "mean_token_accuracy": 0.18962415903806687,
      "num_tokens": 13487430.0,
      "step": 6040
    },
    {
      "entropy": 5.604590988159179,
      "epoch": 0.6789464817206716,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.3828,
      "mean_token_accuracy": 0.19052744209766387,
      "num_tokens": 13499631.0,
      "step": 6045
    },
    {
      "entropy": 5.587384033203125,
      "epoch": 0.6795080586286292,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000496357276153527,
      "loss": 5.4668,
      "mean_token_accuracy": 0.18758317530155183,
      "num_tokens": 13511753.0,
      "step": 6050
    },
    {
      "entropy": 5.585440444946289,
      "epoch": 0.6800696355365867,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.522,
      "mean_token_accuracy": 0.17988150864839553,
      "num_tokens": 13524228.0,
      "step": 6055
    },
    {
      "entropy": 5.66345181465149,
      "epoch": 0.6806312124445443,
      "grad_norm": 0.828125,
      "learning_rate": 0.000496342871633153,
      "loss": 5.5442,
      "mean_token_accuracy": 0.1844257578253746,
      "num_tokens": 13535700.0,
      "step": 6060
    },
    {
      "entropy": 5.607192277908325,
      "epoch": 0.6811927893525018,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.5257,
      "mean_token_accuracy": 0.17659420073032378,
      "num_tokens": 13547337.0,
      "step": 6065
    },
    {
      "entropy": 5.487743520736695,
      "epoch": 0.6817543662604594,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.3796,
      "mean_token_accuracy": 0.18587132096290587,
      "num_tokens": 13558350.0,
      "step": 6070
    },
    {
      "entropy": 5.577818298339844,
      "epoch": 0.6823159431684169,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.4212,
      "mean_token_accuracy": 0.1904219791293144,
      "num_tokens": 13569230.0,
      "step": 6075
    },
    {
      "entropy": 5.64628233909607,
      "epoch": 0.6828775200763745,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.5055,
      "mean_token_accuracy": 0.18413326591253282,
      "num_tokens": 13580502.0,
      "step": 6080
    },
    {
      "entropy": 5.611775588989258,
      "epoch": 0.683439096984332,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.4251,
      "mean_token_accuracy": 0.19118367433547973,
      "num_tokens": 13591524.0,
      "step": 6085
    },
    {
      "entropy": 5.589556741714477,
      "epoch": 0.6840006738922896,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.4441,
      "mean_token_accuracy": 0.1964499443769455,
      "num_tokens": 13602432.0,
      "step": 6090
    },
    {
      "entropy": 5.5022101402282715,
      "epoch": 0.6845622508002471,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.373,
      "mean_token_accuracy": 0.19672563821077346,
      "num_tokens": 13613052.0,
      "step": 6095
    },
    {
      "entropy": 5.5472776889801025,
      "epoch": 0.6851238277082047,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.3863,
      "mean_token_accuracy": 0.19051288664340973,
      "num_tokens": 13623961.0,
      "step": 6100
    },
    {
      "entropy": 5.524657487869263,
      "epoch": 0.6856854046161622,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1882488965988159,
      "num_tokens": 13635285.0,
      "step": 6105
    },
    {
      "entropy": 5.666265821456909,
      "epoch": 0.6862469815241198,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17889947146177293,
      "num_tokens": 13647720.0,
      "step": 6110
    },
    {
      "entropy": 5.523880243301392,
      "epoch": 0.6868085584320772,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.3972,
      "mean_token_accuracy": 0.18586231619119645,
      "num_tokens": 13659726.0,
      "step": 6115
    },
    {
      "entropy": 5.662991333007812,
      "epoch": 0.6873701353400348,
      "grad_norm": 0.875,
      "learning_rate": 0.000496255852573374,
      "loss": 5.5763,
      "mean_token_accuracy": 0.1790321499109268,
      "num_tokens": 13670332.0,
      "step": 6120
    },
    {
      "entropy": 5.703384304046631,
      "epoch": 0.6879317122479923,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.5346,
      "mean_token_accuracy": 0.17642847299575806,
      "num_tokens": 13682419.0,
      "step": 6125
    },
    {
      "entropy": 5.4620270252227785,
      "epoch": 0.6884932891559499,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.2815,
      "mean_token_accuracy": 0.19445524513721466,
      "num_tokens": 13692698.0,
      "step": 6130
    },
    {
      "entropy": 5.50990047454834,
      "epoch": 0.6890548660639074,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.4666,
      "mean_token_accuracy": 0.18721719086170197,
      "num_tokens": 13704791.0,
      "step": 6135
    },
    {
      "entropy": 5.604409885406494,
      "epoch": 0.689616442971865,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496226620764454,
      "loss": 5.3977,
      "mean_token_accuracy": 0.19106096178293228,
      "num_tokens": 13715860.0,
      "step": 6140
    },
    {
      "entropy": 5.5842962741851805,
      "epoch": 0.6901780198798225,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496219295201788,
      "loss": 5.372,
      "mean_token_accuracy": 0.19463435113430022,
      "num_tokens": 13727029.0,
      "step": 6145
    },
    {
      "entropy": 5.5058893203735355,
      "epoch": 0.6907395967877801,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.3707,
      "mean_token_accuracy": 0.19573699980974196,
      "num_tokens": 13737168.0,
      "step": 6150
    },
    {
      "entropy": 5.589278888702393,
      "epoch": 0.6913011736957376,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.5219,
      "mean_token_accuracy": 0.17839339077472688,
      "num_tokens": 13749828.0,
      "step": 6155
    },
    {
      "entropy": 5.516400766372681,
      "epoch": 0.6918627506036952,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.3359,
      "mean_token_accuracy": 0.18713945597410203,
      "num_tokens": 13761617.0,
      "step": 6160
    },
    {
      "entropy": 5.611632585525513,
      "epoch": 0.6924243275116527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496189922516379,
      "loss": 5.3912,
      "mean_token_accuracy": 0.19397446513175964,
      "num_tokens": 13772721.0,
      "step": 6165
    },
    {
      "entropy": 5.482655096054077,
      "epoch": 0.6929859044196103,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1874512568116188,
      "num_tokens": 13783215.0,
      "step": 6170
    },
    {
      "entropy": 5.491256427764893,
      "epoch": 0.6935474813275678,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.3287,
      "mean_token_accuracy": 0.19310313314199448,
      "num_tokens": 13793496.0,
      "step": 6175
    },
    {
      "entropy": 5.462819814682007,
      "epoch": 0.6941090582355254,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496167819052389,
      "loss": 5.25,
      "mean_token_accuracy": 0.18821246027946473,
      "num_tokens": 13802818.0,
      "step": 6180
    },
    {
      "entropy": 5.680048990249634,
      "epoch": 0.6946706351434829,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.5542,
      "mean_token_accuracy": 0.1835671290755272,
      "num_tokens": 13814449.0,
      "step": 6185
    },
    {
      "entropy": 5.573808002471924,
      "epoch": 0.6952322120514405,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.4917,
      "mean_token_accuracy": 0.1829431265592575,
      "num_tokens": 13825369.0,
      "step": 6190
    },
    {
      "entropy": 5.480724239349366,
      "epoch": 0.695793788959398,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18638586699962617,
      "num_tokens": 13837992.0,
      "step": 6195
    },
    {
      "entropy": 5.568413543701172,
      "epoch": 0.6963553658673556,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1812524452805519,
      "num_tokens": 13849771.0,
      "step": 6200
    },
    {
      "entropy": 5.595166349411011,
      "epoch": 0.6969169427753131,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.42,
      "mean_token_accuracy": 0.18859976828098296,
      "num_tokens": 13861997.0,
      "step": 6205
    },
    {
      "entropy": 5.659422731399536,
      "epoch": 0.6974785196832706,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.5096,
      "mean_token_accuracy": 0.1829330876469612,
      "num_tokens": 13872972.0,
      "step": 6210
    },
    {
      "entropy": 5.4354277610778805,
      "epoch": 0.6980400965912281,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19414234906435013,
      "num_tokens": 13883139.0,
      "step": 6215
    },
    {
      "entropy": 5.582811975479126,
      "epoch": 0.6986016734991857,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.4405,
      "mean_token_accuracy": 0.18963438123464585,
      "num_tokens": 13894326.0,
      "step": 6220
    },
    {
      "entropy": 5.603238773345947,
      "epoch": 0.6991632504071432,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.4808,
      "mean_token_accuracy": 0.18860790580511094,
      "num_tokens": 13905525.0,
      "step": 6225
    },
    {
      "entropy": 5.5148704051971436,
      "epoch": 0.6997248273151008,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.4041,
      "mean_token_accuracy": 0.187268990278244,
      "num_tokens": 13915973.0,
      "step": 6230
    },
    {
      "entropy": 5.51510763168335,
      "epoch": 0.7002864042230583,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.2199,
      "mean_token_accuracy": 0.19681506752967834,
      "num_tokens": 13926380.0,
      "step": 6235
    },
    {
      "entropy": 5.55254430770874,
      "epoch": 0.7008479811310159,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.4171,
      "mean_token_accuracy": 0.18290770351886748,
      "num_tokens": 13938523.0,
      "step": 6240
    },
    {
      "entropy": 5.5069276809692385,
      "epoch": 0.7014095580389734,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.3924,
      "mean_token_accuracy": 0.19256162494421006,
      "num_tokens": 13949343.0,
      "step": 6245
    },
    {
      "entropy": 5.591718006134033,
      "epoch": 0.701971134946931,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.5111,
      "mean_token_accuracy": 0.18868376314640045,
      "num_tokens": 13960056.0,
      "step": 6250
    },
    {
      "entropy": 5.67092456817627,
      "epoch": 0.7025327118548885,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.4685,
      "mean_token_accuracy": 0.18847769796848296,
      "num_tokens": 13971932.0,
      "step": 6255
    },
    {
      "entropy": 5.629927158355713,
      "epoch": 0.7030942887628461,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.479,
      "mean_token_accuracy": 0.18567973971366883,
      "num_tokens": 13981992.0,
      "step": 6260
    },
    {
      "entropy": 5.566175651550293,
      "epoch": 0.7036558656708036,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.4326,
      "mean_token_accuracy": 0.1920413225889206,
      "num_tokens": 13992215.0,
      "step": 6265
    },
    {
      "entropy": 5.530323839187622,
      "epoch": 0.7042174425787612,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19865480661392212,
      "num_tokens": 14002689.0,
      "step": 6270
    },
    {
      "entropy": 5.531408071517944,
      "epoch": 0.7047790194867187,
      "grad_norm": 0.75,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.4472,
      "mean_token_accuracy": 0.18879526406526564,
      "num_tokens": 14014951.0,
      "step": 6275
    },
    {
      "entropy": 5.584560108184815,
      "epoch": 0.7053405963946763,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1890434667468071,
      "num_tokens": 14025502.0,
      "step": 6280
    },
    {
      "entropy": 5.607841062545776,
      "epoch": 0.7059021733026338,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.4389,
      "mean_token_accuracy": 0.19130444824695586,
      "num_tokens": 14036576.0,
      "step": 6285
    },
    {
      "entropy": 5.621018886566162,
      "epoch": 0.7064637502105914,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.4673,
      "mean_token_accuracy": 0.1888699486851692,
      "num_tokens": 14046493.0,
      "step": 6290
    },
    {
      "entropy": 5.609211587905884,
      "epoch": 0.7070253271185489,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.5076,
      "mean_token_accuracy": 0.18652400821447374,
      "num_tokens": 14057297.0,
      "step": 6295
    },
    {
      "entropy": 5.505354499816894,
      "epoch": 0.7075869040265065,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.29,
      "mean_token_accuracy": 0.19129664450883865,
      "num_tokens": 14068947.0,
      "step": 6300
    },
    {
      "entropy": 5.536684226989746,
      "epoch": 0.7081484809344639,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.3957,
      "mean_token_accuracy": 0.18915995955467224,
      "num_tokens": 14078979.0,
      "step": 6305
    },
    {
      "entropy": 5.5250598907470705,
      "epoch": 0.7087100578424215,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.3218,
      "mean_token_accuracy": 0.19602209478616714,
      "num_tokens": 14089785.0,
      "step": 6310
    },
    {
      "entropy": 5.566288661956787,
      "epoch": 0.709271634750379,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.4172,
      "mean_token_accuracy": 0.18718542009592057,
      "num_tokens": 14099921.0,
      "step": 6315
    },
    {
      "entropy": 5.455599546432495,
      "epoch": 0.7098332116583366,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.3229,
      "mean_token_accuracy": 0.18730655461549758,
      "num_tokens": 14110965.0,
      "step": 6320
    },
    {
      "entropy": 5.619172811508179,
      "epoch": 0.7103947885662941,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.476,
      "mean_token_accuracy": 0.18679166734218597,
      "num_tokens": 14122409.0,
      "step": 6325
    },
    {
      "entropy": 5.585164785385132,
      "epoch": 0.7109563654742517,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.4064,
      "mean_token_accuracy": 0.19553910940885544,
      "num_tokens": 14134017.0,
      "step": 6330
    },
    {
      "entropy": 5.550099849700928,
      "epoch": 0.7115179423822092,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.5294,
      "mean_token_accuracy": 0.1850521206855774,
      "num_tokens": 14144843.0,
      "step": 6335
    },
    {
      "entropy": 5.584148693084717,
      "epoch": 0.7120795192901668,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.3801,
      "mean_token_accuracy": 0.18981605917215347,
      "num_tokens": 14155944.0,
      "step": 6340
    },
    {
      "entropy": 5.5727903842926025,
      "epoch": 0.7126410961981243,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.4403,
      "mean_token_accuracy": 0.1895127698779106,
      "num_tokens": 14167064.0,
      "step": 6345
    },
    {
      "entropy": 5.584324741363526,
      "epoch": 0.7132026731060819,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.432,
      "mean_token_accuracy": 0.18523438423871993,
      "num_tokens": 14178583.0,
      "step": 6350
    },
    {
      "entropy": 5.50538740158081,
      "epoch": 0.7137642500140394,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.322,
      "mean_token_accuracy": 0.19202942103147508,
      "num_tokens": 14188752.0,
      "step": 6355
    },
    {
      "entropy": 5.520774888992309,
      "epoch": 0.714325826921997,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495897635781793,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1939898669719696,
      "num_tokens": 14199962.0,
      "step": 6360
    },
    {
      "entropy": 5.529936647415161,
      "epoch": 0.7148874038299545,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.3526,
      "mean_token_accuracy": 0.19180352091789246,
      "num_tokens": 14209877.0,
      "step": 6365
    },
    {
      "entropy": 5.508715867996216,
      "epoch": 0.7154489807379121,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.4061,
      "mean_token_accuracy": 0.19456677287817,
      "num_tokens": 14220672.0,
      "step": 6370
    },
    {
      "entropy": 5.635096693038941,
      "epoch": 0.7160105576458696,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.3994,
      "mean_token_accuracy": 0.19460146725177765,
      "num_tokens": 14232348.0,
      "step": 6375
    },
    {
      "entropy": 5.544111728668213,
      "epoch": 0.7165721345538272,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.4144,
      "mean_token_accuracy": 0.18092965185642243,
      "num_tokens": 14243985.0,
      "step": 6380
    },
    {
      "entropy": 5.451342964172364,
      "epoch": 0.7171337114617847,
      "grad_norm": 0.84375,
      "learning_rate": 0.000495859389142044,
      "loss": 5.29,
      "mean_token_accuracy": 0.19636288434267044,
      "num_tokens": 14253548.0,
      "step": 6385
    },
    {
      "entropy": 5.595607852935791,
      "epoch": 0.7176952883697423,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495851718716373,
      "loss": 5.4383,
      "mean_token_accuracy": 0.18815089911222457,
      "num_tokens": 14264757.0,
      "step": 6390
    },
    {
      "entropy": 5.4588829517364506,
      "epoch": 0.7182568652776998,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.3494,
      "mean_token_accuracy": 0.1987864926457405,
      "num_tokens": 14275946.0,
      "step": 6395
    },
    {
      "entropy": 5.505002546310425,
      "epoch": 0.7188184421856574,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.4124,
      "mean_token_accuracy": 0.18201731890439987,
      "num_tokens": 14287006.0,
      "step": 6400
    },
    {
      "entropy": 5.654749965667724,
      "epoch": 0.7193800190936148,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.4488,
      "mean_token_accuracy": 0.1832731172442436,
      "num_tokens": 14299328.0,
      "step": 6405
    },
    {
      "entropy": 5.603736782073975,
      "epoch": 0.7199415960015724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.4416,
      "mean_token_accuracy": 0.1916926309466362,
      "num_tokens": 14310048.0,
      "step": 6410
    },
    {
      "entropy": 5.469794273376465,
      "epoch": 0.7205031729095299,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.3665,
      "mean_token_accuracy": 0.19476278126239777,
      "num_tokens": 14321107.0,
      "step": 6415
    },
    {
      "entropy": 5.511231136322022,
      "epoch": 0.7210647498174875,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.2868,
      "mean_token_accuracy": 0.1940134882926941,
      "num_tokens": 14332561.0,
      "step": 6420
    },
    {
      "entropy": 5.6382426738739015,
      "epoch": 0.721626326725445,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.4774,
      "mean_token_accuracy": 0.1871576800942421,
      "num_tokens": 14343238.0,
      "step": 6425
    },
    {
      "entropy": 5.563603067398072,
      "epoch": 0.7221879036334026,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.3769,
      "mean_token_accuracy": 0.19246648401021957,
      "num_tokens": 14356067.0,
      "step": 6430
    },
    {
      "entropy": 5.504949760437012,
      "epoch": 0.7227494805413601,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.4069,
      "mean_token_accuracy": 0.19316048324108123,
      "num_tokens": 14367302.0,
      "step": 6435
    },
    {
      "entropy": 5.519736480712891,
      "epoch": 0.7233110574493177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19798117876052856,
      "num_tokens": 14377535.0,
      "step": 6440
    },
    {
      "entropy": 5.597304964065552,
      "epoch": 0.7238726343572752,
      "grad_norm": 0.875,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.4703,
      "mean_token_accuracy": 0.18569201678037645,
      "num_tokens": 14389493.0,
      "step": 6445
    },
    {
      "entropy": 5.513818645477295,
      "epoch": 0.7244342112652328,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.38,
      "mean_token_accuracy": 0.19683949947357177,
      "num_tokens": 14401564.0,
      "step": 6450
    },
    {
      "entropy": 5.583502721786499,
      "epoch": 0.7249957881731903,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.4452,
      "mean_token_accuracy": 0.1950697973370552,
      "num_tokens": 14412123.0,
      "step": 6455
    },
    {
      "entropy": 5.582270765304566,
      "epoch": 0.7255573650811479,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.4358,
      "mean_token_accuracy": 0.19213946908712387,
      "num_tokens": 14423614.0,
      "step": 6460
    },
    {
      "entropy": 5.5610638618469235,
      "epoch": 0.7261189419891054,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.4092,
      "mean_token_accuracy": 0.18669372648000718,
      "num_tokens": 14434830.0,
      "step": 6465
    },
    {
      "entropy": 5.5573282718658445,
      "epoch": 0.726680518897063,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.443,
      "mean_token_accuracy": 0.18869293481111526,
      "num_tokens": 14445015.0,
      "step": 6470
    },
    {
      "entropy": 5.619374179840088,
      "epoch": 0.7272420958050205,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.545,
      "mean_token_accuracy": 0.17871332168579102,
      "num_tokens": 14455809.0,
      "step": 6475
    },
    {
      "entropy": 5.612972593307495,
      "epoch": 0.7278036727129781,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.3051,
      "mean_token_accuracy": 0.20037926733493805,
      "num_tokens": 14466565.0,
      "step": 6480
    },
    {
      "entropy": 5.583424758911133,
      "epoch": 0.7283652496209356,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.4051,
      "mean_token_accuracy": 0.19647229909896852,
      "num_tokens": 14477383.0,
      "step": 6485
    },
    {
      "entropy": 5.5140275955200195,
      "epoch": 0.7289268265288932,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.3961,
      "mean_token_accuracy": 0.19006068855524064,
      "num_tokens": 14489968.0,
      "step": 6490
    },
    {
      "entropy": 5.6003580570220945,
      "epoch": 0.7294884034368507,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.4535,
      "mean_token_accuracy": 0.1879394993185997,
      "num_tokens": 14500808.0,
      "step": 6495
    },
    {
      "entropy": 5.547910356521607,
      "epoch": 0.7300499803448082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.4572,
      "mean_token_accuracy": 0.1888415828347206,
      "num_tokens": 14511444.0,
      "step": 6500
    },
    {
      "entropy": 5.5708457946777346,
      "epoch": 0.7306115572527657,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.3589,
      "mean_token_accuracy": 0.19092436581850053,
      "num_tokens": 14522562.0,
      "step": 6505
    },
    {
      "entropy": 5.524218368530273,
      "epoch": 0.7311731341607233,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.4033,
      "mean_token_accuracy": 0.18657224774360656,
      "num_tokens": 14533497.0,
      "step": 6510
    },
    {
      "entropy": 5.559765386581421,
      "epoch": 0.7317347110686808,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004956576733121,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18306263387203217,
      "num_tokens": 14545637.0,
      "step": 6515
    },
    {
      "entropy": 5.603921270370483,
      "epoch": 0.7322962879766384,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.4155,
      "mean_token_accuracy": 0.18643435686826706,
      "num_tokens": 14556826.0,
      "step": 6520
    },
    {
      "entropy": 5.49514331817627,
      "epoch": 0.7328578648845959,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000495641959931013,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18760323375463486,
      "num_tokens": 14567315.0,
      "step": 6525
    },
    {
      "entropy": 5.547469234466552,
      "epoch": 0.7334194417925535,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.4247,
      "mean_token_accuracy": 0.18974489718675613,
      "num_tokens": 14577685.0,
      "step": 6530
    },
    {
      "entropy": 5.572555875778198,
      "epoch": 0.733981018700511,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000495626218448624,
      "loss": 5.3799,
      "mean_token_accuracy": 0.19248069375753402,
      "num_tokens": 14588885.0,
      "step": 6535
    },
    {
      "entropy": 5.600017261505127,
      "epoch": 0.7345425956084686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.4081,
      "mean_token_accuracy": 0.1935748502612114,
      "num_tokens": 14599774.0,
      "step": 6540
    },
    {
      "entropy": 5.628725242614746,
      "epoch": 0.7351041725164261,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18519861549139022,
      "num_tokens": 14612101.0,
      "step": 6545
    },
    {
      "entropy": 5.661167860031128,
      "epoch": 0.7356657494243837,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.519,
      "mean_token_accuracy": 0.1750835120677948,
      "num_tokens": 14622151.0,
      "step": 6550
    },
    {
      "entropy": 5.5682800769805905,
      "epoch": 0.7362273263323412,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18312456905841829,
      "num_tokens": 14634631.0,
      "step": 6555
    },
    {
      "entropy": 5.542213344573975,
      "epoch": 0.7367889032402988,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1886059269309044,
      "num_tokens": 14645429.0,
      "step": 6560
    },
    {
      "entropy": 5.521846008300781,
      "epoch": 0.7373504801482563,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.3583,
      "mean_token_accuracy": 0.19451422840356827,
      "num_tokens": 14657309.0,
      "step": 6565
    },
    {
      "entropy": 5.505011129379272,
      "epoch": 0.7379120570562139,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.3065,
      "mean_token_accuracy": 0.1937410607933998,
      "num_tokens": 14668578.0,
      "step": 6570
    },
    {
      "entropy": 5.569357442855835,
      "epoch": 0.7384736339641714,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.5221,
      "mean_token_accuracy": 0.18763448297977448,
      "num_tokens": 14679120.0,
      "step": 6575
    },
    {
      "entropy": 5.629948091506958,
      "epoch": 0.739035210872129,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.4717,
      "mean_token_accuracy": 0.18317111879587172,
      "num_tokens": 14690849.0,
      "step": 6580
    },
    {
      "entropy": 5.589046144485474,
      "epoch": 0.7395967877800865,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.4297,
      "mean_token_accuracy": 0.18859443962574005,
      "num_tokens": 14702431.0,
      "step": 6585
    },
    {
      "entropy": 5.465069723129273,
      "epoch": 0.7401583646880441,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.3306,
      "mean_token_accuracy": 0.1946323871612549,
      "num_tokens": 14714271.0,
      "step": 6590
    },
    {
      "entropy": 5.518150568008423,
      "epoch": 0.7407199415960015,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.4504,
      "mean_token_accuracy": 0.19020374566316606,
      "num_tokens": 14724796.0,
      "step": 6595
    },
    {
      "entropy": 5.5341894149780275,
      "epoch": 0.7412815185039591,
      "grad_norm": 0.890625,
      "learning_rate": 0.000495523213982657,
      "loss": 5.3861,
      "mean_token_accuracy": 0.18682342320680617,
      "num_tokens": 14735331.0,
      "step": 6600
    },
    {
      "entropy": 5.659159135818482,
      "epoch": 0.7418430954119166,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.4613,
      "mean_token_accuracy": 0.18522186428308487,
      "num_tokens": 14748595.0,
      "step": 6605
    },
    {
      "entropy": 5.600618886947632,
      "epoch": 0.7424046723198742,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.3706,
      "mean_token_accuracy": 0.18925037682056428,
      "num_tokens": 14759501.0,
      "step": 6610
    },
    {
      "entropy": 5.455843830108643,
      "epoch": 0.7429662492278317,
      "grad_norm": 0.875,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.4301,
      "mean_token_accuracy": 0.19183726012706756,
      "num_tokens": 14769752.0,
      "step": 6615
    },
    {
      "entropy": 5.5726597785949705,
      "epoch": 0.7435278261357893,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19258251041173935,
      "num_tokens": 14781166.0,
      "step": 6620
    },
    {
      "entropy": 5.4954887390136715,
      "epoch": 0.7440894030437468,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495483280880511,
      "loss": 5.2698,
      "mean_token_accuracy": 0.19690830111503602,
      "num_tokens": 14790536.0,
      "step": 6625
    },
    {
      "entropy": 5.588681983947754,
      "epoch": 0.7446509799517044,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000495475273198244,
      "loss": 5.4871,
      "mean_token_accuracy": 0.18325650095939636,
      "num_tokens": 14801673.0,
      "step": 6630
    },
    {
      "entropy": 5.573399448394776,
      "epoch": 0.7452125568596619,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.3765,
      "mean_token_accuracy": 0.18690255433320999,
      "num_tokens": 14812345.0,
      "step": 6635
    },
    {
      "entropy": 5.62178111076355,
      "epoch": 0.7457741337676195,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.5528,
      "mean_token_accuracy": 0.18060370683670043,
      "num_tokens": 14823620.0,
      "step": 6640
    },
    {
      "entropy": 5.566000509262085,
      "epoch": 0.746335710675577,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19653813689947128,
      "num_tokens": 14833487.0,
      "step": 6645
    },
    {
      "entropy": 5.434810543060303,
      "epoch": 0.7468972875835346,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.2261,
      "mean_token_accuracy": 0.19704335033893586,
      "num_tokens": 14844421.0,
      "step": 6650
    },
    {
      "entropy": 5.472637557983399,
      "epoch": 0.7474588644914921,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.4071,
      "mean_token_accuracy": 0.19750991612672805,
      "num_tokens": 14855525.0,
      "step": 6655
    },
    {
      "entropy": 5.626046895980835,
      "epoch": 0.7480204413994497,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.4507,
      "mean_token_accuracy": 0.1893661677837372,
      "num_tokens": 14866717.0,
      "step": 6660
    },
    {
      "entropy": 5.528974771499634,
      "epoch": 0.7485820183074072,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.3667,
      "mean_token_accuracy": 0.1851828545331955,
      "num_tokens": 14877372.0,
      "step": 6665
    },
    {
      "entropy": 5.5370550632476805,
      "epoch": 0.7491435952153648,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.4143,
      "mean_token_accuracy": 0.18978495597839357,
      "num_tokens": 14888618.0,
      "step": 6670
    },
    {
      "entropy": 5.5310461044311525,
      "epoch": 0.7497051721233223,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18488093167543412,
      "num_tokens": 14900421.0,
      "step": 6675
    },
    {
      "entropy": 5.569034290313721,
      "epoch": 0.7502667490312799,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.395,
      "mean_token_accuracy": 0.1861965164542198,
      "num_tokens": 14910634.0,
      "step": 6680
    },
    {
      "entropy": 5.6285967350006105,
      "epoch": 0.7508283259392374,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.4555,
      "mean_token_accuracy": 0.18458913266658783,
      "num_tokens": 14921408.0,
      "step": 6685
    },
    {
      "entropy": 5.571818542480469,
      "epoch": 0.7513899028471949,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.4237,
      "mean_token_accuracy": 0.19003868997097015,
      "num_tokens": 14933435.0,
      "step": 6690
    },
    {
      "entropy": 5.537566995620727,
      "epoch": 0.7519514797551524,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.4524,
      "mean_token_accuracy": 0.18646232038736343,
      "num_tokens": 14946488.0,
      "step": 6695
    },
    {
      "entropy": 5.551724195480347,
      "epoch": 0.75251305666311,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.3194,
      "mean_token_accuracy": 0.19105692356824874,
      "num_tokens": 14958053.0,
      "step": 6700
    },
    {
      "entropy": 5.557274389266968,
      "epoch": 0.7530746335710675,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.3849,
      "mean_token_accuracy": 0.18780029714107513,
      "num_tokens": 14968922.0,
      "step": 6705
    },
    {
      "entropy": 5.550119066238404,
      "epoch": 0.7536362104790251,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.4083,
      "mean_token_accuracy": 0.1934783563017845,
      "num_tokens": 14979541.0,
      "step": 6710
    },
    {
      "entropy": 5.5849934101104735,
      "epoch": 0.7541977873869826,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.4127,
      "mean_token_accuracy": 0.18646025508642197,
      "num_tokens": 14990009.0,
      "step": 6715
    },
    {
      "entropy": 5.512436628341675,
      "epoch": 0.7547593642949402,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.4081,
      "mean_token_accuracy": 0.1883482962846756,
      "num_tokens": 15001712.0,
      "step": 6720
    },
    {
      "entropy": 5.5713903427124025,
      "epoch": 0.7553209412028977,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.367,
      "mean_token_accuracy": 0.19454659521579742,
      "num_tokens": 15012574.0,
      "step": 6725
    },
    {
      "entropy": 5.48911452293396,
      "epoch": 0.7558825181108553,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.3695,
      "mean_token_accuracy": 0.1913204848766327,
      "num_tokens": 15023034.0,
      "step": 6730
    },
    {
      "entropy": 5.623052501678467,
      "epoch": 0.7564440950188128,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.4206,
      "mean_token_accuracy": 0.18837956637144088,
      "num_tokens": 15033925.0,
      "step": 6735
    },
    {
      "entropy": 5.5848432064056395,
      "epoch": 0.7570056719267704,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.4134,
      "mean_token_accuracy": 0.19637832492589952,
      "num_tokens": 15044947.0,
      "step": 6740
    },
    {
      "entropy": 5.5618785381317135,
      "epoch": 0.7575672488347279,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.3211,
      "mean_token_accuracy": 0.1985709175467491,
      "num_tokens": 15054784.0,
      "step": 6745
    },
    {
      "entropy": 5.5245020389556885,
      "epoch": 0.7581288257426855,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.4005,
      "mean_token_accuracy": 0.18477832973003389,
      "num_tokens": 15066272.0,
      "step": 6750
    },
    {
      "entropy": 5.5647424221038815,
      "epoch": 0.758690402650643,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.4776,
      "mean_token_accuracy": 0.18543320745229722,
      "num_tokens": 15079504.0,
      "step": 6755
    },
    {
      "entropy": 5.603858995437622,
      "epoch": 0.7592519795586006,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.4143,
      "mean_token_accuracy": 0.1918395146727562,
      "num_tokens": 15089623.0,
      "step": 6760
    },
    {
      "entropy": 5.50961480140686,
      "epoch": 0.7598135564665581,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.3154,
      "mean_token_accuracy": 0.1919368550181389,
      "num_tokens": 15100816.0,
      "step": 6765
    },
    {
      "entropy": 5.524628400802612,
      "epoch": 0.7603751333745157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.4237,
      "mean_token_accuracy": 0.19635622650384904,
      "num_tokens": 15111697.0,
      "step": 6770
    },
    {
      "entropy": 5.548503875732422,
      "epoch": 0.7609367102824732,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.4877,
      "mean_token_accuracy": 0.18239143043756484,
      "num_tokens": 15122869.0,
      "step": 6775
    },
    {
      "entropy": 5.499809646606446,
      "epoch": 0.7614982871904308,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.2861,
      "mean_token_accuracy": 0.1911429449915886,
      "num_tokens": 15134709.0,
      "step": 6780
    },
    {
      "entropy": 5.527266597747802,
      "epoch": 0.7620598640983883,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.2965,
      "mean_token_accuracy": 0.1974724605679512,
      "num_tokens": 15145319.0,
      "step": 6785
    },
    {
      "entropy": 5.520294809341431,
      "epoch": 0.7626214410063458,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000495215322197552,
      "loss": 5.352,
      "mean_token_accuracy": 0.18900242298841477,
      "num_tokens": 15156463.0,
      "step": 6790
    },
    {
      "entropy": 5.620303964614868,
      "epoch": 0.7631830179143033,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.4938,
      "mean_token_accuracy": 0.17575843557715415,
      "num_tokens": 15167618.0,
      "step": 6795
    },
    {
      "entropy": 5.608112096786499,
      "epoch": 0.7637445948222609,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1902905896306038,
      "num_tokens": 15179312.0,
      "step": 6800
    },
    {
      "entropy": 5.493063926696777,
      "epoch": 0.7643061717302184,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.3797,
      "mean_token_accuracy": 0.1939145416021347,
      "num_tokens": 15191370.0,
      "step": 6805
    },
    {
      "entropy": 5.606846714019776,
      "epoch": 0.764867748638176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.3275,
      "mean_token_accuracy": 0.19638228565454482,
      "num_tokens": 15201060.0,
      "step": 6810
    },
    {
      "entropy": 5.475592184066772,
      "epoch": 0.7654293255461335,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.2656,
      "mean_token_accuracy": 0.19386058747768403,
      "num_tokens": 15212186.0,
      "step": 6815
    },
    {
      "entropy": 5.489868545532227,
      "epoch": 0.7659909024540911,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19804201722145082,
      "num_tokens": 15223301.0,
      "step": 6820
    },
    {
      "entropy": 5.52301549911499,
      "epoch": 0.7665524793620486,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.4391,
      "mean_token_accuracy": 0.18734809756278992,
      "num_tokens": 15234475.0,
      "step": 6825
    },
    {
      "entropy": 5.575947284698486,
      "epoch": 0.7671140562700062,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1875208780169487,
      "num_tokens": 15245877.0,
      "step": 6830
    },
    {
      "entropy": 5.627669906616211,
      "epoch": 0.7676756331779637,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.4073,
      "mean_token_accuracy": 0.190565824508667,
      "num_tokens": 15257375.0,
      "step": 6835
    },
    {
      "entropy": 5.575936651229858,
      "epoch": 0.7682372100859213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.4183,
      "mean_token_accuracy": 0.1895638108253479,
      "num_tokens": 15268079.0,
      "step": 6840
    },
    {
      "entropy": 5.491517400741577,
      "epoch": 0.7687987869938788,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18927390575408937,
      "num_tokens": 15278678.0,
      "step": 6845
    },
    {
      "entropy": 5.618205642700195,
      "epoch": 0.7693603639018364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.5076,
      "mean_token_accuracy": 0.1865157887339592,
      "num_tokens": 15290080.0,
      "step": 6850
    },
    {
      "entropy": 5.583620357513428,
      "epoch": 0.7699219408097939,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.432,
      "mean_token_accuracy": 0.18702781349420547,
      "num_tokens": 15300602.0,
      "step": 6855
    },
    {
      "entropy": 5.594201231002808,
      "epoch": 0.7704835177177515,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.4349,
      "mean_token_accuracy": 0.19131208211183548,
      "num_tokens": 15312823.0,
      "step": 6860
    },
    {
      "entropy": 5.634102821350098,
      "epoch": 0.771045094625709,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000495090998008418,
      "loss": 5.4441,
      "mean_token_accuracy": 0.18723303079605103,
      "num_tokens": 15323870.0,
      "step": 6865
    },
    {
      "entropy": 5.4997388362884525,
      "epoch": 0.7716066715336666,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000495082653656997,
      "loss": 5.3302,
      "mean_token_accuracy": 0.1965235248208046,
      "num_tokens": 15334887.0,
      "step": 6870
    },
    {
      "entropy": 5.572580862045288,
      "epoch": 0.7721682484416241,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.4373,
      "mean_token_accuracy": 0.18693511039018632,
      "num_tokens": 15345871.0,
      "step": 6875
    },
    {
      "entropy": 5.509525918960572,
      "epoch": 0.7727298253495817,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1993497222661972,
      "num_tokens": 15356530.0,
      "step": 6880
    },
    {
      "entropy": 5.530187654495239,
      "epoch": 0.7732914022575391,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.364,
      "mean_token_accuracy": 0.19166866093873977,
      "num_tokens": 15368756.0,
      "step": 6885
    },
    {
      "entropy": 5.612720966339111,
      "epoch": 0.7738529791654967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.4484,
      "mean_token_accuracy": 0.1904335215687752,
      "num_tokens": 15379450.0,
      "step": 6890
    },
    {
      "entropy": 5.519256973266602,
      "epoch": 0.7744145560734542,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.3537,
      "mean_token_accuracy": 0.1858827590942383,
      "num_tokens": 15389501.0,
      "step": 6895
    },
    {
      "entropy": 5.385013151168823,
      "epoch": 0.7749761329814118,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.3646,
      "mean_token_accuracy": 0.19009243696928024,
      "num_tokens": 15400608.0,
      "step": 6900
    },
    {
      "entropy": 5.608634376525879,
      "epoch": 0.7755377098893693,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.4393,
      "mean_token_accuracy": 0.1848304584622383,
      "num_tokens": 15411709.0,
      "step": 6905
    },
    {
      "entropy": 5.53405122756958,
      "epoch": 0.7760992867973269,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.2999,
      "mean_token_accuracy": 0.19948293417692184,
      "num_tokens": 15422292.0,
      "step": 6910
    },
    {
      "entropy": 5.550871086120606,
      "epoch": 0.7766608637052844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495007239187896,
      "loss": 5.4532,
      "mean_token_accuracy": 0.18661819994449616,
      "num_tokens": 15434439.0,
      "step": 6915
    },
    {
      "entropy": 5.514196968078613,
      "epoch": 0.777222440613242,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.3966,
      "mean_token_accuracy": 0.19032733142375946,
      "num_tokens": 15445480.0,
      "step": 6920
    },
    {
      "entropy": 5.508719158172608,
      "epoch": 0.7777840175211995,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.3857,
      "mean_token_accuracy": 0.19126745760440828,
      "num_tokens": 15456960.0,
      "step": 6925
    },
    {
      "entropy": 5.510855197906494,
      "epoch": 0.7783455944291571,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000494981974929824,
      "loss": 5.3858,
      "mean_token_accuracy": 0.18876940459012986,
      "num_tokens": 15468169.0,
      "step": 6930
    },
    {
      "entropy": 5.550881004333496,
      "epoch": 0.7789071713371146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494973539501486,
      "loss": 5.3383,
      "mean_token_accuracy": 0.1927856683731079,
      "num_tokens": 15478740.0,
      "step": 6935
    },
    {
      "entropy": 5.555187034606933,
      "epoch": 0.7794687482450722,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.4513,
      "mean_token_accuracy": 0.18629500418901443,
      "num_tokens": 15489705.0,
      "step": 6940
    },
    {
      "entropy": 5.4992107391357425,
      "epoch": 0.7800303251530297,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.2956,
      "mean_token_accuracy": 0.18771727234125138,
      "num_tokens": 15500549.0,
      "step": 6945
    },
    {
      "entropy": 5.647176933288574,
      "epoch": 0.7805919020609873,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.4289,
      "mean_token_accuracy": 0.18746083974838257,
      "num_tokens": 15511124.0,
      "step": 6950
    },
    {
      "entropy": 5.5008502960205075,
      "epoch": 0.7811534789689448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.3351,
      "mean_token_accuracy": 0.1971758633852005,
      "num_tokens": 15521810.0,
      "step": 6955
    },
    {
      "entropy": 5.530869197845459,
      "epoch": 0.7817150558769024,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.3992,
      "mean_token_accuracy": 0.1918172210454941,
      "num_tokens": 15533108.0,
      "step": 6960
    },
    {
      "entropy": 5.530841445922851,
      "epoch": 0.7822766327848599,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.4111,
      "mean_token_accuracy": 0.1919243112206459,
      "num_tokens": 15544841.0,
      "step": 6965
    },
    {
      "entropy": 5.560785245895386,
      "epoch": 0.7828382096928175,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.4083,
      "mean_token_accuracy": 0.18217880874872208,
      "num_tokens": 15556297.0,
      "step": 6970
    },
    {
      "entropy": 5.5459887981414795,
      "epoch": 0.783399786600775,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.2868,
      "mean_token_accuracy": 0.19351261705160142,
      "num_tokens": 15566607.0,
      "step": 6975
    },
    {
      "entropy": 5.55396466255188,
      "epoch": 0.7839613635087325,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.4688,
      "mean_token_accuracy": 0.18487369418144226,
      "num_tokens": 15577435.0,
      "step": 6980
    },
    {
      "entropy": 5.538680505752564,
      "epoch": 0.78452294041669,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.3663,
      "mean_token_accuracy": 0.1931474909186363,
      "num_tokens": 15587844.0,
      "step": 6985
    },
    {
      "entropy": 5.637773132324218,
      "epoch": 0.7850845173246476,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.4324,
      "mean_token_accuracy": 0.18827223032712936,
      "num_tokens": 15599586.0,
      "step": 6990
    },
    {
      "entropy": 5.513357830047608,
      "epoch": 0.7856460942326051,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.318,
      "mean_token_accuracy": 0.18972791731357574,
      "num_tokens": 15610838.0,
      "step": 6995
    },
    {
      "entropy": 5.547962617874146,
      "epoch": 0.7862076711405627,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.4026,
      "mean_token_accuracy": 0.19327963143587112,
      "num_tokens": 15622652.0,
      "step": 7000
    },
    {
      "entropy": 5.64856448173523,
      "epoch": 0.7867692480485202,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.4897,
      "mean_token_accuracy": 0.18318869769573212,
      "num_tokens": 15634280.0,
      "step": 7005
    },
    {
      "entropy": 5.4800700664520265,
      "epoch": 0.7873308249564778,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.2215,
      "mean_token_accuracy": 0.199824021756649,
      "num_tokens": 15644088.0,
      "step": 7010
    },
    {
      "entropy": 5.486963939666748,
      "epoch": 0.7878924018644353,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.4564,
      "mean_token_accuracy": 0.1799951523542404,
      "num_tokens": 15654968.0,
      "step": 7015
    },
    {
      "entropy": 5.549936771392822,
      "epoch": 0.7884539787723929,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.3667,
      "mean_token_accuracy": 0.19336676746606826,
      "num_tokens": 15666758.0,
      "step": 7020
    },
    {
      "entropy": 5.424612140655517,
      "epoch": 0.7890155556803504,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.213,
      "mean_token_accuracy": 0.19932689517736435,
      "num_tokens": 15677295.0,
      "step": 7025
    },
    {
      "entropy": 5.5242737293243405,
      "epoch": 0.789577132588308,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.4261,
      "mean_token_accuracy": 0.19343287497758865,
      "num_tokens": 15689508.0,
      "step": 7030
    },
    {
      "entropy": 5.58857364654541,
      "epoch": 0.7901387094962655,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.3808,
      "mean_token_accuracy": 0.19007019251585006,
      "num_tokens": 15701289.0,
      "step": 7035
    },
    {
      "entropy": 5.494377374649048,
      "epoch": 0.7907002864042231,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.336,
      "mean_token_accuracy": 0.19051492512226104,
      "num_tokens": 15711579.0,
      "step": 7040
    },
    {
      "entropy": 5.527608489990234,
      "epoch": 0.7912618633121806,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.4217,
      "mean_token_accuracy": 0.19214471131563188,
      "num_tokens": 15722629.0,
      "step": 7045
    },
    {
      "entropy": 5.557682180404663,
      "epoch": 0.7918234402201382,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18356509059667586,
      "num_tokens": 15735818.0,
      "step": 7050
    },
    {
      "entropy": 5.512237167358398,
      "epoch": 0.7923850171280957,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.4147,
      "mean_token_accuracy": 0.1895158275961876,
      "num_tokens": 15746914.0,
      "step": 7055
    },
    {
      "entropy": 5.513495969772339,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.0625,
      "learning_rate": 0.000494760378181364,
      "loss": 5.3162,
      "mean_token_accuracy": 0.18762257397174836,
      "num_tokens": 15757597.0,
      "step": 7060
    },
    {
      "entropy": 5.458051300048828,
      "epoch": 0.7935081709440108,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19967395961284637,
      "num_tokens": 15768860.0,
      "step": 7065
    },
    {
      "entropy": 5.538003635406494,
      "epoch": 0.7940697478519684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.3602,
      "mean_token_accuracy": 0.19036328196525573,
      "num_tokens": 15779476.0,
      "step": 7070
    },
    {
      "entropy": 5.519798755645752,
      "epoch": 0.7946313247599258,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.4229,
      "mean_token_accuracy": 0.185899318754673,
      "num_tokens": 15792142.0,
      "step": 7075
    },
    {
      "entropy": 5.565739440917969,
      "epoch": 0.7951929016678834,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.3904,
      "mean_token_accuracy": 0.1878409206867218,
      "num_tokens": 15802307.0,
      "step": 7080
    },
    {
      "entropy": 5.5143050193786625,
      "epoch": 0.7957544785758409,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18899307996034623,
      "num_tokens": 15812986.0,
      "step": 7085
    },
    {
      "entropy": 5.46642804145813,
      "epoch": 0.7963160554837985,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19560994058847428,
      "num_tokens": 15823938.0,
      "step": 7090
    },
    {
      "entropy": 5.463511753082275,
      "epoch": 0.796877632391756,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.294,
      "mean_token_accuracy": 0.19440711587667464,
      "num_tokens": 15835813.0,
      "step": 7095
    },
    {
      "entropy": 5.563611745834351,
      "epoch": 0.7974392092997136,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494691242717217,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19473447799682617,
      "num_tokens": 15846203.0,
      "step": 7100
    },
    {
      "entropy": 5.548427248001099,
      "epoch": 0.7980007862076711,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.3116,
      "mean_token_accuracy": 0.1960406556725502,
      "num_tokens": 15856919.0,
      "step": 7105
    },
    {
      "entropy": 5.528464078903198,
      "epoch": 0.7985623631156287,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494673888895125,
      "loss": 5.4381,
      "mean_token_accuracy": 0.18583116829395294,
      "num_tokens": 15867149.0,
      "step": 7110
    },
    {
      "entropy": 5.451193571090698,
      "epoch": 0.7991239400235862,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.3199,
      "mean_token_accuracy": 0.1911962777376175,
      "num_tokens": 15877229.0,
      "step": 7115
    },
    {
      "entropy": 5.534007453918457,
      "epoch": 0.7996855169315438,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1942991554737091,
      "num_tokens": 15888799.0,
      "step": 7120
    },
    {
      "entropy": 5.453420400619507,
      "epoch": 0.8002470938395013,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.2998,
      "mean_token_accuracy": 0.20614445805549622,
      "num_tokens": 15900144.0,
      "step": 7125
    },
    {
      "entropy": 5.5438392639160154,
      "epoch": 0.8008086707474589,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.4826,
      "mean_token_accuracy": 0.18744038939476013,
      "num_tokens": 15912030.0,
      "step": 7130
    },
    {
      "entropy": 5.532735061645508,
      "epoch": 0.8013702476554164,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.3432,
      "mean_token_accuracy": 0.19349053651094436,
      "num_tokens": 15923606.0,
      "step": 7135
    },
    {
      "entropy": 5.513996839523315,
      "epoch": 0.801931824563374,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.3457,
      "mean_token_accuracy": 0.19366424530744553,
      "num_tokens": 15935077.0,
      "step": 7140
    },
    {
      "entropy": 5.463584470748901,
      "epoch": 0.8024934014713315,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.2792,
      "mean_token_accuracy": 0.19784001410007476,
      "num_tokens": 15945331.0,
      "step": 7145
    },
    {
      "entropy": 5.484343242645264,
      "epoch": 0.8030549783792891,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.3751,
      "mean_token_accuracy": 0.1876963496208191,
      "num_tokens": 15957031.0,
      "step": 7150
    },
    {
      "entropy": 5.553222417831421,
      "epoch": 0.8036165552872466,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.3959,
      "mean_token_accuracy": 0.19386137425899505,
      "num_tokens": 15967545.0,
      "step": 7155
    },
    {
      "entropy": 5.667354869842529,
      "epoch": 0.8041781321952042,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.4384,
      "mean_token_accuracy": 0.18039540499448775,
      "num_tokens": 15978700.0,
      "step": 7160
    },
    {
      "entropy": 5.5349548816680905,
      "epoch": 0.8047397091031617,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.3387,
      "mean_token_accuracy": 0.1968977376818657,
      "num_tokens": 15989712.0,
      "step": 7165
    },
    {
      "entropy": 5.606640815734863,
      "epoch": 0.8053012860111193,
      "grad_norm": 0.78125,
      "learning_rate": 0.000494569178502031,
      "loss": 5.551,
      "mean_token_accuracy": 0.18528392165899277,
      "num_tokens": 16002380.0,
      "step": 7170
    },
    {
      "entropy": 5.566968727111816,
      "epoch": 0.8058628629190767,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.3175,
      "mean_token_accuracy": 0.1930129960179329,
      "num_tokens": 16013749.0,
      "step": 7175
    },
    {
      "entropy": 5.561373567581176,
      "epoch": 0.8064244398270343,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.5018,
      "mean_token_accuracy": 0.1882968232035637,
      "num_tokens": 16025131.0,
      "step": 7180
    },
    {
      "entropy": 5.568534517288208,
      "epoch": 0.8069860167349918,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.3688,
      "mean_token_accuracy": 0.1933572143316269,
      "num_tokens": 16037269.0,
      "step": 7185
    },
    {
      "entropy": 5.671428966522217,
      "epoch": 0.8075475936429494,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.5666,
      "mean_token_accuracy": 0.17810628414154053,
      "num_tokens": 16050992.0,
      "step": 7190
    },
    {
      "entropy": 5.507680368423462,
      "epoch": 0.8081091705509069,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.3812,
      "mean_token_accuracy": 0.18824538588523865,
      "num_tokens": 16062309.0,
      "step": 7195
    },
    {
      "entropy": 5.600071954727173,
      "epoch": 0.8086707474588645,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.4184,
      "mean_token_accuracy": 0.18918824046850205,
      "num_tokens": 16073079.0,
      "step": 7200
    },
    {
      "entropy": 5.570611572265625,
      "epoch": 0.809232324366822,
      "grad_norm": 0.875,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.3146,
      "mean_token_accuracy": 0.1991833671927452,
      "num_tokens": 16082885.0,
      "step": 7205
    },
    {
      "entropy": 5.431540012359619,
      "epoch": 0.8097939012747796,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.2685,
      "mean_token_accuracy": 0.20003439486026764,
      "num_tokens": 16094173.0,
      "step": 7210
    },
    {
      "entropy": 5.469393920898438,
      "epoch": 0.8103554781827371,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494489985015675,
      "loss": 5.3326,
      "mean_token_accuracy": 0.19339679181575775,
      "num_tokens": 16106351.0,
      "step": 7215
    },
    {
      "entropy": 5.4503538608551025,
      "epoch": 0.8109170550906947,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.3145,
      "mean_token_accuracy": 0.1912689507007599,
      "num_tokens": 16116892.0,
      "step": 7220
    },
    {
      "entropy": 5.6118786334991455,
      "epoch": 0.8114786319986522,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.4661,
      "mean_token_accuracy": 0.18370114117860795,
      "num_tokens": 16128316.0,
      "step": 7225
    },
    {
      "entropy": 5.521116590499878,
      "epoch": 0.8120402089066098,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19552281200885774,
      "num_tokens": 16139245.0,
      "step": 7230
    },
    {
      "entropy": 5.5180247783660885,
      "epoch": 0.8126017858145673,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049445460619367,
      "loss": 5.3374,
      "mean_token_accuracy": 0.1948014661669731,
      "num_tokens": 16150893.0,
      "step": 7235
    },
    {
      "entropy": 5.5881579399108885,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.4639,
      "mean_token_accuracy": 0.17190439850091935,
      "num_tokens": 16163430.0,
      "step": 7240
    },
    {
      "entropy": 5.57142825126648,
      "epoch": 0.8137249396304824,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.3525,
      "mean_token_accuracy": 0.1935686409473419,
      "num_tokens": 16173954.0,
      "step": 7245
    },
    {
      "entropy": 5.552434825897217,
      "epoch": 0.81428651653844,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.4205,
      "mean_token_accuracy": 0.185828697681427,
      "num_tokens": 16185491.0,
      "step": 7250
    },
    {
      "entropy": 5.450665426254273,
      "epoch": 0.8148480934463975,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.267,
      "mean_token_accuracy": 0.20237178206443787,
      "num_tokens": 16195301.0,
      "step": 7255
    },
    {
      "entropy": 5.556926202774048,
      "epoch": 0.8154096703543551,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1841839298605919,
      "num_tokens": 16206620.0,
      "step": 7260
    },
    {
      "entropy": 5.5046991348266605,
      "epoch": 0.8159712472623126,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.4839,
      "mean_token_accuracy": 0.1835168182849884,
      "num_tokens": 16218693.0,
      "step": 7265
    },
    {
      "entropy": 5.570673894882202,
      "epoch": 0.8165328241702701,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.3636,
      "mean_token_accuracy": 0.192826946079731,
      "num_tokens": 16229124.0,
      "step": 7270
    },
    {
      "entropy": 5.52437915802002,
      "epoch": 0.8170944010782276,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.3274,
      "mean_token_accuracy": 0.19266424775123597,
      "num_tokens": 16239999.0,
      "step": 7275
    },
    {
      "entropy": 5.519512367248535,
      "epoch": 0.8176559779861852,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.4209,
      "mean_token_accuracy": 0.19051078855991363,
      "num_tokens": 16251027.0,
      "step": 7280
    },
    {
      "entropy": 5.564484930038452,
      "epoch": 0.8182175548941427,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.3595,
      "mean_token_accuracy": 0.19417764544486998,
      "num_tokens": 16262924.0,
      "step": 7285
    },
    {
      "entropy": 5.636598253250122,
      "epoch": 0.8187791318021003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.4695,
      "mean_token_accuracy": 0.1896255284547806,
      "num_tokens": 16274481.0,
      "step": 7290
    },
    {
      "entropy": 5.530969762802124,
      "epoch": 0.8193407087100578,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.4005,
      "mean_token_accuracy": 0.19529114067554473,
      "num_tokens": 16285214.0,
      "step": 7295
    },
    {
      "entropy": 5.5609289646148685,
      "epoch": 0.8199022856180154,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.4643,
      "mean_token_accuracy": 0.18784563839435578,
      "num_tokens": 16296077.0,
      "step": 7300
    },
    {
      "entropy": 5.588407897949219,
      "epoch": 0.8204638625259729,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.3795,
      "mean_token_accuracy": 0.19368068128824234,
      "num_tokens": 16307514.0,
      "step": 7305
    },
    {
      "entropy": 5.652513551712036,
      "epoch": 0.8210254394339305,
      "grad_norm": 0.875,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.3754,
      "mean_token_accuracy": 0.19173796027898787,
      "num_tokens": 16318734.0,
      "step": 7310
    },
    {
      "entropy": 5.497101211547852,
      "epoch": 0.821587016341888,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494311973083621,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18982843905687333,
      "num_tokens": 16329057.0,
      "step": 7315
    },
    {
      "entropy": 5.501382112503052,
      "epoch": 0.8221485932498456,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.3257,
      "mean_token_accuracy": 0.19335218369960785,
      "num_tokens": 16339646.0,
      "step": 7320
    },
    {
      "entropy": 5.582084655761719,
      "epoch": 0.8227101701578031,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.4173,
      "mean_token_accuracy": 0.18968148976564408,
      "num_tokens": 16349783.0,
      "step": 7325
    },
    {
      "entropy": 5.5929333686828615,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.331,
      "mean_token_accuracy": 0.1926446095108986,
      "num_tokens": 16361078.0,
      "step": 7330
    },
    {
      "entropy": 5.478531408309936,
      "epoch": 0.8238333239737182,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.3506,
      "mean_token_accuracy": 0.19307311475276948,
      "num_tokens": 16371308.0,
      "step": 7335
    },
    {
      "entropy": 5.527521324157715,
      "epoch": 0.8243949008816758,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.3868,
      "mean_token_accuracy": 0.19338780492544175,
      "num_tokens": 16382682.0,
      "step": 7340
    },
    {
      "entropy": 5.605612945556641,
      "epoch": 0.8249564777896333,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.4086,
      "mean_token_accuracy": 0.19141978174448013,
      "num_tokens": 16393715.0,
      "step": 7345
    },
    {
      "entropy": 5.468640565872192,
      "epoch": 0.8255180546975909,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.3949,
      "mean_token_accuracy": 0.19251279532909393,
      "num_tokens": 16406332.0,
      "step": 7350
    },
    {
      "entropy": 5.5208453178405765,
      "epoch": 0.8260796316055484,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.3301,
      "mean_token_accuracy": 0.18916616588830948,
      "num_tokens": 16417702.0,
      "step": 7355
    },
    {
      "entropy": 5.563840389251709,
      "epoch": 0.826641208513506,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.3709,
      "mean_token_accuracy": 0.19190021306276323,
      "num_tokens": 16428497.0,
      "step": 7360
    },
    {
      "entropy": 5.543086099624634,
      "epoch": 0.8272027854214634,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.3661,
      "mean_token_accuracy": 0.1900161996483803,
      "num_tokens": 16440637.0,
      "step": 7365
    },
    {
      "entropy": 5.521610069274902,
      "epoch": 0.827764362329421,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18313750624656677,
      "num_tokens": 16453213.0,
      "step": 7370
    },
    {
      "entropy": 5.528475093841553,
      "epoch": 0.8283259392373785,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.3975,
      "mean_token_accuracy": 0.19446578472852707,
      "num_tokens": 16464345.0,
      "step": 7375
    },
    {
      "entropy": 5.535910940170288,
      "epoch": 0.8288875161453361,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19729151874780654,
      "num_tokens": 16475399.0,
      "step": 7380
    },
    {
      "entropy": 5.48125319480896,
      "epoch": 0.8294490930532936,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.2577,
      "mean_token_accuracy": 0.1967956691980362,
      "num_tokens": 16485439.0,
      "step": 7385
    },
    {
      "entropy": 5.482007074356079,
      "epoch": 0.8300106699612512,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.263,
      "mean_token_accuracy": 0.1964390233159065,
      "num_tokens": 16496765.0,
      "step": 7390
    },
    {
      "entropy": 5.475945377349854,
      "epoch": 0.8305722468692087,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.3035,
      "mean_token_accuracy": 0.1950630024075508,
      "num_tokens": 16509199.0,
      "step": 7395
    },
    {
      "entropy": 5.546600580215454,
      "epoch": 0.8311338237771663,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.4583,
      "mean_token_accuracy": 0.1897977501153946,
      "num_tokens": 16520073.0,
      "step": 7400
    },
    {
      "entropy": 5.472220230102539,
      "epoch": 0.8316954006851238,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.331,
      "mean_token_accuracy": 0.19148144870996475,
      "num_tokens": 16530919.0,
      "step": 7405
    },
    {
      "entropy": 5.536686372756958,
      "epoch": 0.8322569775930814,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.3571,
      "mean_token_accuracy": 0.19261872321367263,
      "num_tokens": 16542591.0,
      "step": 7410
    },
    {
      "entropy": 5.57855257987976,
      "epoch": 0.8328185545010389,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.3693,
      "mean_token_accuracy": 0.19039538949728013,
      "num_tokens": 16554984.0,
      "step": 7415
    },
    {
      "entropy": 5.433529186248779,
      "epoch": 0.8333801314089965,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000494122054427151,
      "loss": 5.2631,
      "mean_token_accuracy": 0.19918903708457947,
      "num_tokens": 16566598.0,
      "step": 7420
    },
    {
      "entropy": 5.60313572883606,
      "epoch": 0.833941708316954,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18586488366127013,
      "num_tokens": 16577868.0,
      "step": 7425
    },
    {
      "entropy": 5.516066837310791,
      "epoch": 0.8345032852249116,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.3602,
      "mean_token_accuracy": 0.19123763740062713,
      "num_tokens": 16587891.0,
      "step": 7430
    },
    {
      "entropy": 5.518604183197022,
      "epoch": 0.8350648621328691,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494094671958544,
      "loss": 5.3008,
      "mean_token_accuracy": 0.2017410770058632,
      "num_tokens": 16599924.0,
      "step": 7435
    },
    {
      "entropy": 5.476879024505616,
      "epoch": 0.8356264390408267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.2494,
      "mean_token_accuracy": 0.19790668934583663,
      "num_tokens": 16609958.0,
      "step": 7440
    },
    {
      "entropy": 5.483253622055054,
      "epoch": 0.8361880159487842,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.3777,
      "mean_token_accuracy": 0.18868407756090164,
      "num_tokens": 16621159.0,
      "step": 7445
    },
    {
      "entropy": 5.592933893203735,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.468,
      "mean_token_accuracy": 0.19059434980154039,
      "num_tokens": 16632263.0,
      "step": 7450
    },
    {
      "entropy": 5.529058313369751,
      "epoch": 0.8373111697646993,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.3614,
      "mean_token_accuracy": 0.18601302355527877,
      "num_tokens": 16644117.0,
      "step": 7455
    },
    {
      "entropy": 5.427464723587036,
      "epoch": 0.8378727466726568,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19728361517190934,
      "num_tokens": 16654529.0,
      "step": 7460
    },
    {
      "entropy": 5.473336267471313,
      "epoch": 0.8384343235806143,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.3,
      "mean_token_accuracy": 0.2006448820233345,
      "num_tokens": 16666191.0,
      "step": 7465
    },
    {
      "entropy": 5.529743385314942,
      "epoch": 0.8389959004885719,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19333176612854003,
      "num_tokens": 16677252.0,
      "step": 7470
    },
    {
      "entropy": 5.492028474807739,
      "epoch": 0.8395574773965294,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494021345117024,
      "loss": 5.3378,
      "mean_token_accuracy": 0.19180602729320526,
      "num_tokens": 16689487.0,
      "step": 7475
    },
    {
      "entropy": 5.461004590988159,
      "epoch": 0.840119054304487,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.2332,
      "mean_token_accuracy": 0.19861174374818802,
      "num_tokens": 16699955.0,
      "step": 7480
    },
    {
      "entropy": 5.393486642837525,
      "epoch": 0.8406806312124445,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.2896,
      "mean_token_accuracy": 0.18961472660303116,
      "num_tokens": 16711071.0,
      "step": 7485
    },
    {
      "entropy": 5.430171346664428,
      "epoch": 0.8412422081204021,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.2964,
      "mean_token_accuracy": 0.194288931787014,
      "num_tokens": 16722115.0,
      "step": 7490
    },
    {
      "entropy": 5.485861444473267,
      "epoch": 0.8418037850283596,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18838993161916734,
      "num_tokens": 16733146.0,
      "step": 7495
    },
    {
      "entropy": 5.486392593383789,
      "epoch": 0.8423653619363172,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.319,
      "mean_token_accuracy": 0.19011340141296387,
      "num_tokens": 16745543.0,
      "step": 7500
    },
    {
      "entropy": 5.540941953659058,
      "epoch": 0.8429269388442747,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.3722,
      "mean_token_accuracy": 0.1917041152715683,
      "num_tokens": 16755782.0,
      "step": 7505
    },
    {
      "entropy": 5.617510795593262,
      "epoch": 0.8434885157522323,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.4143,
      "mean_token_accuracy": 0.18830805867910386,
      "num_tokens": 16767509.0,
      "step": 7510
    },
    {
      "entropy": 5.4642973899841305,
      "epoch": 0.8440500926601898,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.281,
      "mean_token_accuracy": 0.19337462335824968,
      "num_tokens": 16779257.0,
      "step": 7515
    },
    {
      "entropy": 5.441992902755738,
      "epoch": 0.8446116695681474,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.3159,
      "mean_token_accuracy": 0.1949651435017586,
      "num_tokens": 16790769.0,
      "step": 7520
    },
    {
      "entropy": 5.51636118888855,
      "epoch": 0.8451732464761049,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.31,
      "mean_token_accuracy": 0.19314562678337097,
      "num_tokens": 16800054.0,
      "step": 7525
    },
    {
      "entropy": 5.59525785446167,
      "epoch": 0.8457348233840625,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.4189,
      "mean_token_accuracy": 0.1875070497393608,
      "num_tokens": 16811485.0,
      "step": 7530
    },
    {
      "entropy": 5.621941375732422,
      "epoch": 0.84629640029202,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.4679,
      "mean_token_accuracy": 0.19378382563591004,
      "num_tokens": 16822683.0,
      "step": 7535
    },
    {
      "entropy": 5.537042951583862,
      "epoch": 0.8468579771999776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.3388,
      "mean_token_accuracy": 0.19330107271671296,
      "num_tokens": 16834200.0,
      "step": 7540
    },
    {
      "entropy": 5.543708467483521,
      "epoch": 0.8474195541079351,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.3785,
      "mean_token_accuracy": 0.18811353892087937,
      "num_tokens": 16846399.0,
      "step": 7545
    },
    {
      "entropy": 5.649693870544434,
      "epoch": 0.8479811310158927,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.5236,
      "mean_token_accuracy": 0.1862836241722107,
      "num_tokens": 16858322.0,
      "step": 7550
    },
    {
      "entropy": 5.560483121871949,
      "epoch": 0.8485427079238502,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.3538,
      "mean_token_accuracy": 0.19695565700531006,
      "num_tokens": 16869537.0,
      "step": 7555
    },
    {
      "entropy": 5.6301332950592045,
      "epoch": 0.8491042848318077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.3907,
      "mean_token_accuracy": 0.19266645312309266,
      "num_tokens": 16881566.0,
      "step": 7560
    },
    {
      "entropy": 5.458097124099732,
      "epoch": 0.8496658617397652,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.3145,
      "mean_token_accuracy": 0.19759692400693893,
      "num_tokens": 16892959.0,
      "step": 7565
    },
    {
      "entropy": 5.530508184432984,
      "epoch": 0.8502274386477228,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.3778,
      "mean_token_accuracy": 0.18589012324810028,
      "num_tokens": 16904809.0,
      "step": 7570
    },
    {
      "entropy": 5.57429575920105,
      "epoch": 0.8507890155556803,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.3971,
      "mean_token_accuracy": 0.1847073659300804,
      "num_tokens": 16916969.0,
      "step": 7575
    },
    {
      "entropy": 5.461697816848755,
      "epoch": 0.8513505924636379,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.2724,
      "mean_token_accuracy": 0.19852736294269563,
      "num_tokens": 16928253.0,
      "step": 7580
    },
    {
      "entropy": 5.4836688995361325,
      "epoch": 0.8519121693715954,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.3268,
      "mean_token_accuracy": 0.19043341279029846,
      "num_tokens": 16939875.0,
      "step": 7585
    },
    {
      "entropy": 5.629465341567993,
      "epoch": 0.852473746279553,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.4481,
      "mean_token_accuracy": 0.1888473078608513,
      "num_tokens": 16950173.0,
      "step": 7590
    },
    {
      "entropy": 5.478756999969482,
      "epoch": 0.8530353231875105,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.3698,
      "mean_token_accuracy": 0.19270646274089814,
      "num_tokens": 16960933.0,
      "step": 7595
    },
    {
      "entropy": 5.494446277618408,
      "epoch": 0.8535969000954681,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.268,
      "mean_token_accuracy": 0.19606069177389146,
      "num_tokens": 16972284.0,
      "step": 7600
    },
    {
      "entropy": 5.510715007781982,
      "epoch": 0.8541584770034256,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.2819,
      "mean_token_accuracy": 0.19838723242282869,
      "num_tokens": 16983541.0,
      "step": 7605
    },
    {
      "entropy": 5.545347261428833,
      "epoch": 0.8547200539113832,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049377057282307,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18634025305509566,
      "num_tokens": 16993773.0,
      "step": 7610
    },
    {
      "entropy": 5.615330791473388,
      "epoch": 0.8552816308193407,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000493761187403909,
      "loss": 5.4084,
      "mean_token_accuracy": 0.19078975915908813,
      "num_tokens": 17005287.0,
      "step": 7615
    },
    {
      "entropy": 5.548895788192749,
      "epoch": 0.8558432077272983,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000493751795019307,
      "loss": 5.3438,
      "mean_token_accuracy": 0.18888139426708223,
      "num_tokens": 17015664.0,
      "step": 7620
    },
    {
      "entropy": 5.490134954452515,
      "epoch": 0.8564047846352558,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.2514,
      "mean_token_accuracy": 0.19612780809402466,
      "num_tokens": 17026876.0,
      "step": 7625
    },
    {
      "entropy": 5.466113615036011,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.3004,
      "mean_token_accuracy": 0.2042296215891838,
      "num_tokens": 17037321.0,
      "step": 7630
    },
    {
      "entropy": 5.507684326171875,
      "epoch": 0.8575279384511709,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.3795,
      "mean_token_accuracy": 0.19242687225341798,
      "num_tokens": 17047506.0,
      "step": 7635
    },
    {
      "entropy": 5.560714054107666,
      "epoch": 0.8580895153591285,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.4205,
      "mean_token_accuracy": 0.19133085161447524,
      "num_tokens": 17059710.0,
      "step": 7640
    },
    {
      "entropy": 5.57714581489563,
      "epoch": 0.858651092267086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.4412,
      "mean_token_accuracy": 0.1824363186955452,
      "num_tokens": 17070956.0,
      "step": 7645
    },
    {
      "entropy": 5.607796049118042,
      "epoch": 0.8592126691750436,
      "grad_norm": 0.875,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.4052,
      "mean_token_accuracy": 0.19116396307945252,
      "num_tokens": 17081404.0,
      "step": 7650
    },
    {
      "entropy": 5.643889713287353,
      "epoch": 0.859774246083001,
      "grad_norm": 0.875,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.4151,
      "mean_token_accuracy": 0.19118166118860244,
      "num_tokens": 17092692.0,
      "step": 7655
    },
    {
      "entropy": 5.4890522956848145,
      "epoch": 0.8603358229909586,
      "grad_norm": 0.90625,
      "learning_rate": 0.000493676405222546,
      "loss": 5.2731,
      "mean_token_accuracy": 0.1976029172539711,
      "num_tokens": 17103730.0,
      "step": 7660
    },
    {
      "entropy": 5.532904243469238,
      "epoch": 0.8608973998989161,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19136169403791428,
      "num_tokens": 17115086.0,
      "step": 7665
    },
    {
      "entropy": 5.5359259128570555,
      "epoch": 0.8614589768068737,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19453381299972533,
      "num_tokens": 17125928.0,
      "step": 7670
    },
    {
      "entropy": 5.523212480545044,
      "epoch": 0.8620205537148312,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.3739,
      "mean_token_accuracy": 0.18908844143152237,
      "num_tokens": 17137328.0,
      "step": 7675
    },
    {
      "entropy": 5.524524784088134,
      "epoch": 0.8625821306227888,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.3495,
      "mean_token_accuracy": 0.19175111055374144,
      "num_tokens": 17148274.0,
      "step": 7680
    },
    {
      "entropy": 5.544976091384887,
      "epoch": 0.8631437075307463,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.3815,
      "mean_token_accuracy": 0.18837463706731797,
      "num_tokens": 17158834.0,
      "step": 7685
    },
    {
      "entropy": 5.5258111476898195,
      "epoch": 0.8637052844387039,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1950663670897484,
      "num_tokens": 17169276.0,
      "step": 7690
    },
    {
      "entropy": 5.503770923614502,
      "epoch": 0.8642668613466614,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.2914,
      "mean_token_accuracy": 0.19853373765945434,
      "num_tokens": 17180642.0,
      "step": 7695
    },
    {
      "entropy": 5.494513034820557,
      "epoch": 0.864828438254619,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004936005698104,
      "loss": 5.3445,
      "mean_token_accuracy": 0.197983980178833,
      "num_tokens": 17192477.0,
      "step": 7700
    },
    {
      "entropy": 5.569464302062988,
      "epoch": 0.8653900151625765,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.4847,
      "mean_token_accuracy": 0.18918235301971437,
      "num_tokens": 17202653.0,
      "step": 7705
    },
    {
      "entropy": 5.447111320495606,
      "epoch": 0.8659515920705341,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.2162,
      "mean_token_accuracy": 0.2023095965385437,
      "num_tokens": 17212607.0,
      "step": 7710
    },
    {
      "entropy": 5.541468286514283,
      "epoch": 0.8665131689784916,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000493572016677141,
      "loss": 5.3531,
      "mean_token_accuracy": 0.19717216789722442,
      "num_tokens": 17223701.0,
      "step": 7715
    },
    {
      "entropy": 5.535778760910034,
      "epoch": 0.8670747458864492,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19667365849018098,
      "num_tokens": 17234878.0,
      "step": 7720
    },
    {
      "entropy": 5.508067464828491,
      "epoch": 0.8676363227944067,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.346,
      "mean_token_accuracy": 0.19576495438814162,
      "num_tokens": 17245813.0,
      "step": 7725
    },
    {
      "entropy": 5.515953826904297,
      "epoch": 0.8681978997023643,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.4193,
      "mean_token_accuracy": 0.19034817367792128,
      "num_tokens": 17257378.0,
      "step": 7730
    },
    {
      "entropy": 5.48284945487976,
      "epoch": 0.8687594766103218,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.3412,
      "mean_token_accuracy": 0.19484313279390336,
      "num_tokens": 17269108.0,
      "step": 7735
    },
    {
      "entropy": 5.550235271453857,
      "epoch": 0.8693210535182794,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.3306,
      "mean_token_accuracy": 0.18691225796937944,
      "num_tokens": 17280643.0,
      "step": 7740
    },
    {
      "entropy": 5.593494558334351,
      "epoch": 0.8698826304262369,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.3916,
      "mean_token_accuracy": 0.19205151349306107,
      "num_tokens": 17291236.0,
      "step": 7745
    },
    {
      "entropy": 5.604566144943237,
      "epoch": 0.8704442073341944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.463,
      "mean_token_accuracy": 0.18148287236690522,
      "num_tokens": 17303547.0,
      "step": 7750
    },
    {
      "entropy": 5.613951539993286,
      "epoch": 0.8710057842421519,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.4302,
      "mean_token_accuracy": 0.18613469749689102,
      "num_tokens": 17314803.0,
      "step": 7755
    },
    {
      "entropy": 5.53216519355774,
      "epoch": 0.8715673611501095,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.298,
      "mean_token_accuracy": 0.19373864829540252,
      "num_tokens": 17326009.0,
      "step": 7760
    },
    {
      "entropy": 5.569486904144287,
      "epoch": 0.872128938058067,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.2991,
      "mean_token_accuracy": 0.20181606709957123,
      "num_tokens": 17337078.0,
      "step": 7765
    },
    {
      "entropy": 5.547098207473755,
      "epoch": 0.8726905149660246,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.4276,
      "mean_token_accuracy": 0.18845681548118592,
      "num_tokens": 17349266.0,
      "step": 7770
    },
    {
      "entropy": 5.401117372512817,
      "epoch": 0.8732520918739821,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.2293,
      "mean_token_accuracy": 0.1980017215013504,
      "num_tokens": 17360327.0,
      "step": 7775
    },
    {
      "entropy": 5.618751907348633,
      "epoch": 0.8738136687819397,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.4021,
      "mean_token_accuracy": 0.18749536275863649,
      "num_tokens": 17372727.0,
      "step": 7780
    },
    {
      "entropy": 5.631286334991455,
      "epoch": 0.8743752456898972,
      "grad_norm": 0.875,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.4525,
      "mean_token_accuracy": 0.183826045691967,
      "num_tokens": 17384160.0,
      "step": 7785
    },
    {
      "entropy": 5.523481798171997,
      "epoch": 0.8749368225978548,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.3731,
      "mean_token_accuracy": 0.18825365453958512,
      "num_tokens": 17396009.0,
      "step": 7790
    },
    {
      "entropy": 5.521708822250366,
      "epoch": 0.8754983995058123,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.3739,
      "mean_token_accuracy": 0.1929594859480858,
      "num_tokens": 17406903.0,
      "step": 7795
    },
    {
      "entropy": 5.470612192153931,
      "epoch": 0.8760599764137699,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.2255,
      "mean_token_accuracy": 0.1995358482003212,
      "num_tokens": 17417694.0,
      "step": 7800
    },
    {
      "entropy": 5.503048610687256,
      "epoch": 0.8766215533217274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.4652,
      "mean_token_accuracy": 0.19427064955234527,
      "num_tokens": 17428731.0,
      "step": 7805
    },
    {
      "entropy": 5.5018659114837645,
      "epoch": 0.877183130229685,
      "grad_norm": 0.875,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19879435896873474,
      "num_tokens": 17439242.0,
      "step": 7810
    },
    {
      "entropy": 5.567990493774414,
      "epoch": 0.8777447071376425,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.3333,
      "mean_token_accuracy": 0.19243992865085602,
      "num_tokens": 17449980.0,
      "step": 7815
    },
    {
      "entropy": 5.596957969665527,
      "epoch": 0.8783062840456001,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.425,
      "mean_token_accuracy": 0.18374176025390626,
      "num_tokens": 17464140.0,
      "step": 7820
    },
    {
      "entropy": 5.615205717086792,
      "epoch": 0.8788678609535576,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.4335,
      "mean_token_accuracy": 0.18940581679344176,
      "num_tokens": 17476246.0,
      "step": 7825
    },
    {
      "entropy": 5.596604347229004,
      "epoch": 0.8794294378615152,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.4443,
      "mean_token_accuracy": 0.18022418767213821,
      "num_tokens": 17488169.0,
      "step": 7830
    },
    {
      "entropy": 5.505037450790406,
      "epoch": 0.8799910147694727,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.2575,
      "mean_token_accuracy": 0.19681808799505235,
      "num_tokens": 17498465.0,
      "step": 7835
    },
    {
      "entropy": 5.558221769332886,
      "epoch": 0.8805525916774303,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.254,
      "mean_token_accuracy": 0.2049960345029831,
      "num_tokens": 17509217.0,
      "step": 7840
    },
    {
      "entropy": 5.472606801986695,
      "epoch": 0.8811141685853877,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19729041755199433,
      "num_tokens": 17519851.0,
      "step": 7845
    },
    {
      "entropy": 5.502962446212768,
      "epoch": 0.8816757454933453,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.3673,
      "mean_token_accuracy": 0.19381319284439086,
      "num_tokens": 17532144.0,
      "step": 7850
    },
    {
      "entropy": 5.499823713302613,
      "epoch": 0.8822373224013028,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.2621,
      "mean_token_accuracy": 0.20211952477693557,
      "num_tokens": 17542929.0,
      "step": 7855
    },
    {
      "entropy": 5.6443767070770265,
      "epoch": 0.8827988993092604,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.4472,
      "mean_token_accuracy": 0.18478454500436783,
      "num_tokens": 17554931.0,
      "step": 7860
    },
    {
      "entropy": 5.512378692626953,
      "epoch": 0.8833604762172179,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.2632,
      "mean_token_accuracy": 0.19460886418819429,
      "num_tokens": 17566026.0,
      "step": 7865
    },
    {
      "entropy": 5.502492427825928,
      "epoch": 0.8839220531251755,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049327330152104,
      "loss": 5.3605,
      "mean_token_accuracy": 0.19323073625564574,
      "num_tokens": 17578396.0,
      "step": 7870
    },
    {
      "entropy": 5.5544038772583,
      "epoch": 0.884483630033133,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.3439,
      "mean_token_accuracy": 0.19812433272600175,
      "num_tokens": 17590559.0,
      "step": 7875
    },
    {
      "entropy": 5.51351408958435,
      "epoch": 0.8850452069410906,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.3202,
      "mean_token_accuracy": 0.19301193207502365,
      "num_tokens": 17600833.0,
      "step": 7880
    },
    {
      "entropy": 5.487991523742676,
      "epoch": 0.8856067838490481,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.3149,
      "mean_token_accuracy": 0.19400409758090972,
      "num_tokens": 17612166.0,
      "step": 7885
    },
    {
      "entropy": 5.564039659500122,
      "epoch": 0.8861683607570057,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.4208,
      "mean_token_accuracy": 0.18780704736709594,
      "num_tokens": 17623070.0,
      "step": 7890
    },
    {
      "entropy": 5.62103009223938,
      "epoch": 0.8867299376649632,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493224495924948,
      "loss": 5.4862,
      "mean_token_accuracy": 0.18929917961359025,
      "num_tokens": 17633319.0,
      "step": 7895
    },
    {
      "entropy": 5.550469541549683,
      "epoch": 0.8872915145729208,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.2996,
      "mean_token_accuracy": 0.19384047985076905,
      "num_tokens": 17643818.0,
      "step": 7900
    },
    {
      "entropy": 5.468474674224853,
      "epoch": 0.8878530914808783,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19358717501163483,
      "num_tokens": 17655665.0,
      "step": 7905
    },
    {
      "entropy": 5.454145526885986,
      "epoch": 0.8884146683888359,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.2182,
      "mean_token_accuracy": 0.20388979613780975,
      "num_tokens": 17666200.0,
      "step": 7910
    },
    {
      "entropy": 5.485030221939087,
      "epoch": 0.8889762452967934,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.4095,
      "mean_token_accuracy": 0.1872565269470215,
      "num_tokens": 17677299.0,
      "step": 7915
    },
    {
      "entropy": 5.572393131256104,
      "epoch": 0.889537822204751,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.2841,
      "mean_token_accuracy": 0.20393956452608109,
      "num_tokens": 17687870.0,
      "step": 7920
    },
    {
      "entropy": 5.52224178314209,
      "epoch": 0.8900993991127085,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.2995,
      "mean_token_accuracy": 0.19359648674726487,
      "num_tokens": 17698853.0,
      "step": 7925
    },
    {
      "entropy": 5.513133859634399,
      "epoch": 0.8906609760206661,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.3475,
      "mean_token_accuracy": 0.19691725075244904,
      "num_tokens": 17709897.0,
      "step": 7930
    },
    {
      "entropy": 5.555172634124756,
      "epoch": 0.8912225529286236,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.352,
      "mean_token_accuracy": 0.19076764583587646,
      "num_tokens": 17720974.0,
      "step": 7935
    },
    {
      "entropy": 5.640217590332031,
      "epoch": 0.8917841298365812,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.5075,
      "mean_token_accuracy": 0.18322578966617584,
      "num_tokens": 17734449.0,
      "step": 7940
    },
    {
      "entropy": 5.438530492782593,
      "epoch": 0.8923457067445386,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.2417,
      "mean_token_accuracy": 0.19710720777511598,
      "num_tokens": 17744735.0,
      "step": 7945
    },
    {
      "entropy": 5.491062259674072,
      "epoch": 0.8929072836524962,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.2637,
      "mean_token_accuracy": 0.19604025185108184,
      "num_tokens": 17755641.0,
      "step": 7950
    },
    {
      "entropy": 5.5673990726470945,
      "epoch": 0.8934688605604537,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.3708,
      "mean_token_accuracy": 0.187679785490036,
      "num_tokens": 17767621.0,
      "step": 7955
    },
    {
      "entropy": 5.640732288360596,
      "epoch": 0.8940304374684113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.4709,
      "mean_token_accuracy": 0.18887109756469728,
      "num_tokens": 17779507.0,
      "step": 7960
    },
    {
      "entropy": 5.499023199081421,
      "epoch": 0.8945920143763688,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.3152,
      "mean_token_accuracy": 0.18959518522024155,
      "num_tokens": 17790371.0,
      "step": 7965
    },
    {
      "entropy": 5.525818395614624,
      "epoch": 0.8951535912843264,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.3162,
      "mean_token_accuracy": 0.19837681353092193,
      "num_tokens": 17800903.0,
      "step": 7970
    },
    {
      "entropy": 5.58081464767456,
      "epoch": 0.8957151681922839,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.5155,
      "mean_token_accuracy": 0.17849943935871124,
      "num_tokens": 17812053.0,
      "step": 7975
    },
    {
      "entropy": 5.508196544647217,
      "epoch": 0.8962767451002415,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.2358,
      "mean_token_accuracy": 0.19567595422267914,
      "num_tokens": 17822113.0,
      "step": 7980
    },
    {
      "entropy": 5.510505151748657,
      "epoch": 0.896838322008199,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19738962948322297,
      "num_tokens": 17832741.0,
      "step": 7985
    },
    {
      "entropy": 5.3847143173217775,
      "epoch": 0.8973998989161566,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.2142,
      "mean_token_accuracy": 0.1925610765814781,
      "num_tokens": 17842968.0,
      "step": 7990
    },
    {
      "entropy": 5.636124515533448,
      "epoch": 0.8979614758241141,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.4743,
      "mean_token_accuracy": 0.18087823539972306,
      "num_tokens": 17854295.0,
      "step": 7995
    },
    {
      "entropy": 5.648864507675171,
      "epoch": 0.8985230527320717,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.396,
      "mean_token_accuracy": 0.19515135735273362,
      "num_tokens": 17865597.0,
      "step": 8000
    },
    {
      "entropy": 5.421992588043213,
      "epoch": 0.8990846296400292,
      "grad_norm": 0.875,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1959100067615509,
      "num_tokens": 17877405.0,
      "step": 8005
    },
    {
      "entropy": 5.612342739105225,
      "epoch": 0.8996462065479868,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.3702,
      "mean_token_accuracy": 0.19155756682157515,
      "num_tokens": 17889273.0,
      "step": 8010
    },
    {
      "entropy": 5.464116716384888,
      "epoch": 0.9002077834559443,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.2856,
      "mean_token_accuracy": 0.19995818585157393,
      "num_tokens": 17899761.0,
      "step": 8015
    },
    {
      "entropy": 5.476109457015991,
      "epoch": 0.9007693603639019,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.2957,
      "mean_token_accuracy": 0.1960954025387764,
      "num_tokens": 17910607.0,
      "step": 8020
    },
    {
      "entropy": 5.558165597915649,
      "epoch": 0.9013309372718594,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.3559,
      "mean_token_accuracy": 0.19106128215789794,
      "num_tokens": 17922538.0,
      "step": 8025
    },
    {
      "entropy": 5.52228741645813,
      "epoch": 0.901892514179817,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1919267013669014,
      "num_tokens": 17933313.0,
      "step": 8030
    },
    {
      "entropy": 5.626115608215332,
      "epoch": 0.9024540910877745,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.4825,
      "mean_token_accuracy": 0.18904066383838652,
      "num_tokens": 17944893.0,
      "step": 8035
    },
    {
      "entropy": 5.547903442382813,
      "epoch": 0.903015667995732,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.3435,
      "mean_token_accuracy": 0.19203057736158372,
      "num_tokens": 17955019.0,
      "step": 8040
    },
    {
      "entropy": 5.620864963531494,
      "epoch": 0.9035772449036895,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.3799,
      "mean_token_accuracy": 0.18826183080673217,
      "num_tokens": 17965793.0,
      "step": 8045
    },
    {
      "entropy": 5.515917253494263,
      "epoch": 0.9041388218116471,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000492918025562663,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19581181704998016,
      "num_tokens": 17975709.0,
      "step": 8050
    },
    {
      "entropy": 5.554681015014649,
      "epoch": 0.9047003987196046,
      "grad_norm": 0.875,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.4439,
      "mean_token_accuracy": 0.18521264046430588,
      "num_tokens": 17985405.0,
      "step": 8055
    },
    {
      "entropy": 5.572683048248291,
      "epoch": 0.9052619756275622,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1860238254070282,
      "num_tokens": 17997105.0,
      "step": 8060
    },
    {
      "entropy": 5.557096385955811,
      "epoch": 0.9058235525355197,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.3486,
      "mean_token_accuracy": 0.19937967360019684,
      "num_tokens": 18007914.0,
      "step": 8065
    },
    {
      "entropy": 5.490323162078857,
      "epoch": 0.9063851294434773,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.3503,
      "mean_token_accuracy": 0.19325897693634034,
      "num_tokens": 18018328.0,
      "step": 8070
    },
    {
      "entropy": 5.569067096710205,
      "epoch": 0.9069467063514348,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000492867970138744,
      "loss": 5.312,
      "mean_token_accuracy": 0.19093905091285707,
      "num_tokens": 18028568.0,
      "step": 8075
    },
    {
      "entropy": 5.548278379440307,
      "epoch": 0.9075082832593924,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.3293,
      "mean_token_accuracy": 0.1936284586787224,
      "num_tokens": 18038655.0,
      "step": 8080
    },
    {
      "entropy": 5.490902996063232,
      "epoch": 0.9080698601673499,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492847899407939,
      "loss": 5.2668,
      "mean_token_accuracy": 0.19545679688453674,
      "num_tokens": 18049716.0,
      "step": 8085
    },
    {
      "entropy": 5.447275638580322,
      "epoch": 0.9086314370753075,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19852099865674971,
      "num_tokens": 18062530.0,
      "step": 8090
    },
    {
      "entropy": 5.503797626495361,
      "epoch": 0.909193013983265,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.3548,
      "mean_token_accuracy": 0.19214667081832887,
      "num_tokens": 18073076.0,
      "step": 8095
    },
    {
      "entropy": 5.587497901916504,
      "epoch": 0.9097545908912226,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.3025,
      "mean_token_accuracy": 0.19379561096429826,
      "num_tokens": 18083759.0,
      "step": 8100
    },
    {
      "entropy": 5.458155870437622,
      "epoch": 0.9103161677991801,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.2375,
      "mean_token_accuracy": 0.1974729210138321,
      "num_tokens": 18095980.0,
      "step": 8105
    },
    {
      "entropy": 5.5262823581695555,
      "epoch": 0.9108777447071377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.3323,
      "mean_token_accuracy": 0.19940765500068663,
      "num_tokens": 18106672.0,
      "step": 8110
    },
    {
      "entropy": 5.514932918548584,
      "epoch": 0.9114393216150952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.353,
      "mean_token_accuracy": 0.19236091673374175,
      "num_tokens": 18117481.0,
      "step": 8115
    },
    {
      "entropy": 5.4810340881347654,
      "epoch": 0.9120008985230528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.2941,
      "mean_token_accuracy": 0.19350982159376146,
      "num_tokens": 18127602.0,
      "step": 8120
    },
    {
      "entropy": 5.562128257751465,
      "epoch": 0.9125624754310103,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.3671,
      "mean_token_accuracy": 0.19208624809980393,
      "num_tokens": 18140109.0,
      "step": 8125
    },
    {
      "entropy": 5.476376247406006,
      "epoch": 0.9131240523389679,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.2397,
      "mean_token_accuracy": 0.20164853185415268,
      "num_tokens": 18150271.0,
      "step": 8130
    },
    {
      "entropy": 5.486606884002685,
      "epoch": 0.9136856292469253,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.3604,
      "mean_token_accuracy": 0.1954893186688423,
      "num_tokens": 18160746.0,
      "step": 8135
    },
    {
      "entropy": 5.599887132644653,
      "epoch": 0.9142472061548829,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19265573769807814,
      "num_tokens": 18172271.0,
      "step": 8140
    },
    {
      "entropy": 5.508764982223511,
      "epoch": 0.9148087830628404,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19398891925811768,
      "num_tokens": 18183116.0,
      "step": 8145
    },
    {
      "entropy": 5.459906101226807,
      "epoch": 0.915370359970798,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.2578,
      "mean_token_accuracy": 0.20310339629650115,
      "num_tokens": 18195147.0,
      "step": 8150
    },
    {
      "entropy": 5.526568651199341,
      "epoch": 0.9159319368787555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000492706627563404,
      "loss": 5.3078,
      "mean_token_accuracy": 0.19455216079950333,
      "num_tokens": 18206217.0,
      "step": 8155
    },
    {
      "entropy": 5.495534801483155,
      "epoch": 0.9164935137867131,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.3289,
      "mean_token_accuracy": 0.19275714606046676,
      "num_tokens": 18217434.0,
      "step": 8160
    },
    {
      "entropy": 5.534586429595947,
      "epoch": 0.9170550906946706,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19082320630550384,
      "num_tokens": 18229277.0,
      "step": 8165
    },
    {
      "entropy": 5.451659727096557,
      "epoch": 0.9176166676026282,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.1895,
      "mean_token_accuracy": 0.2019288346171379,
      "num_tokens": 18240320.0,
      "step": 8170
    },
    {
      "entropy": 5.464824676513672,
      "epoch": 0.9181782445105857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19750059694051741,
      "num_tokens": 18250380.0,
      "step": 8175
    },
    {
      "entropy": 5.529764413833618,
      "epoch": 0.9187398214185433,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.4003,
      "mean_token_accuracy": 0.1937176465988159,
      "num_tokens": 18261703.0,
      "step": 8180
    },
    {
      "entropy": 5.512107563018799,
      "epoch": 0.9193013983265008,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1919119492173195,
      "num_tokens": 18272472.0,
      "step": 8185
    },
    {
      "entropy": 5.627255821228028,
      "epoch": 0.9198629752344584,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.4242,
      "mean_token_accuracy": 0.18777076303958892,
      "num_tokens": 18282168.0,
      "step": 8190
    },
    {
      "entropy": 5.64281005859375,
      "epoch": 0.9204245521424159,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.4103,
      "mean_token_accuracy": 0.1880524516105652,
      "num_tokens": 18293750.0,
      "step": 8195
    },
    {
      "entropy": 5.509887409210205,
      "epoch": 0.9209861290503735,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.3149,
      "mean_token_accuracy": 0.1946457579731941,
      "num_tokens": 18304568.0,
      "step": 8200
    },
    {
      "entropy": 5.473498296737671,
      "epoch": 0.921547705958331,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.2652,
      "mean_token_accuracy": 0.19549497812986374,
      "num_tokens": 18315129.0,
      "step": 8205
    },
    {
      "entropy": 5.43020396232605,
      "epoch": 0.9221092828662886,
      "grad_norm": 1.0,
      "learning_rate": 0.000492594675081777,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1941726803779602,
      "num_tokens": 18324628.0,
      "step": 8210
    },
    {
      "entropy": 5.391623592376709,
      "epoch": 0.9226708597742461,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492584456007161,
      "loss": 5.2183,
      "mean_token_accuracy": 0.20514661818742752,
      "num_tokens": 18334899.0,
      "step": 8215
    },
    {
      "entropy": 5.519819974899292,
      "epoch": 0.9232324366822037,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.2346,
      "mean_token_accuracy": 0.20070850551128389,
      "num_tokens": 18345719.0,
      "step": 8220
    },
    {
      "entropy": 5.5702324390411375,
      "epoch": 0.9237940135901612,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.3399,
      "mean_token_accuracy": 0.19450607001781464,
      "num_tokens": 18355814.0,
      "step": 8225
    },
    {
      "entropy": 5.55308952331543,
      "epoch": 0.9243555904981187,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18684204518795014,
      "num_tokens": 18366746.0,
      "step": 8230
    },
    {
      "entropy": 5.463734531402588,
      "epoch": 0.9249171674060762,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.2273,
      "mean_token_accuracy": 0.2022406905889511,
      "num_tokens": 18375846.0,
      "step": 8235
    },
    {
      "entropy": 5.446729707717895,
      "epoch": 0.9254787443140338,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000492533256720994,
      "loss": 5.3524,
      "mean_token_accuracy": 0.19247244745492936,
      "num_tokens": 18386850.0,
      "step": 8240
    },
    {
      "entropy": 5.469273376464844,
      "epoch": 0.9260403212219913,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.2508,
      "mean_token_accuracy": 0.1982171654701233,
      "num_tokens": 18396347.0,
      "step": 8245
    },
    {
      "entropy": 5.466515398025512,
      "epoch": 0.9266018981299489,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19833889603614807,
      "num_tokens": 18406583.0,
      "step": 8250
    },
    {
      "entropy": 5.59253306388855,
      "epoch": 0.9271634750379064,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492502454030566,
      "loss": 5.412,
      "mean_token_accuracy": 0.18874627351760864,
      "num_tokens": 18418586.0,
      "step": 8255
    },
    {
      "entropy": 5.5294488906860355,
      "epoch": 0.927725051945864,
      "grad_norm": 0.875,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.2392,
      "mean_token_accuracy": 0.19591175466775895,
      "num_tokens": 18429752.0,
      "step": 8260
    },
    {
      "entropy": 5.529174566268921,
      "epoch": 0.9282866288538215,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.3355,
      "mean_token_accuracy": 0.18996217846870422,
      "num_tokens": 18440663.0,
      "step": 8265
    },
    {
      "entropy": 5.4571661949157715,
      "epoch": 0.9288482057617791,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.2721,
      "mean_token_accuracy": 0.20011357814073563,
      "num_tokens": 18451497.0,
      "step": 8270
    },
    {
      "entropy": 5.499194192886352,
      "epoch": 0.9294097826697366,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18895938694477082,
      "num_tokens": 18462794.0,
      "step": 8275
    },
    {
      "entropy": 5.5031129837036135,
      "epoch": 0.9299713595776942,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.2639,
      "mean_token_accuracy": 0.19696242958307267,
      "num_tokens": 18472973.0,
      "step": 8280
    },
    {
      "entropy": 5.546626949310303,
      "epoch": 0.9305329364856517,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.4316,
      "mean_token_accuracy": 0.1955415740609169,
      "num_tokens": 18484451.0,
      "step": 8285
    },
    {
      "entropy": 5.5921026229858395,
      "epoch": 0.9310945133936093,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.2841,
      "mean_token_accuracy": 0.2020363301038742,
      "num_tokens": 18496412.0,
      "step": 8290
    },
    {
      "entropy": 5.539563989639282,
      "epoch": 0.9316560903015668,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.3189,
      "mean_token_accuracy": 0.196001236140728,
      "num_tokens": 18508547.0,
      "step": 8295
    },
    {
      "entropy": 5.460558557510376,
      "epoch": 0.9322176672095244,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492409672025112,
      "loss": 5.3446,
      "mean_token_accuracy": 0.19378729164600372,
      "num_tokens": 18518954.0,
      "step": 8300
    },
    {
      "entropy": 5.474789333343506,
      "epoch": 0.9327792441174819,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.2074,
      "mean_token_accuracy": 0.20229743272066117,
      "num_tokens": 18530444.0,
      "step": 8305
    },
    {
      "entropy": 5.518152236938477,
      "epoch": 0.9333408210254395,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.2782,
      "mean_token_accuracy": 0.19407936334609985,
      "num_tokens": 18541612.0,
      "step": 8310
    },
    {
      "entropy": 5.528368091583252,
      "epoch": 0.933902397933397,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.4168,
      "mean_token_accuracy": 0.18776628524065017,
      "num_tokens": 18553316.0,
      "step": 8315
    },
    {
      "entropy": 5.479777479171753,
      "epoch": 0.9344639748413546,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049236825558139,
      "loss": 5.3017,
      "mean_token_accuracy": 0.1947462663054466,
      "num_tokens": 18564468.0,
      "step": 8320
    },
    {
      "entropy": 5.465587186813354,
      "epoch": 0.9350255517493121,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.2305,
      "mean_token_accuracy": 0.20041596591472627,
      "num_tokens": 18575536.0,
      "step": 8325
    },
    {
      "entropy": 5.48757004737854,
      "epoch": 0.9355871286572696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.3602,
      "mean_token_accuracy": 0.18073692172765732,
      "num_tokens": 18586906.0,
      "step": 8330
    },
    {
      "entropy": 5.583889770507812,
      "epoch": 0.9361487055652271,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19059130996465684,
      "num_tokens": 18599064.0,
      "step": 8335
    },
    {
      "entropy": 5.495828866958618,
      "epoch": 0.9367102824731847,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.3661,
      "mean_token_accuracy": 0.18868516385555267,
      "num_tokens": 18610057.0,
      "step": 8340
    },
    {
      "entropy": 5.404728841781616,
      "epoch": 0.9372718593811422,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.1775,
      "mean_token_accuracy": 0.2008477121591568,
      "num_tokens": 18620008.0,
      "step": 8345
    },
    {
      "entropy": 5.441278791427612,
      "epoch": 0.9378334362890998,
      "grad_norm": 0.875,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.1993,
      "mean_token_accuracy": 0.19684551060199737,
      "num_tokens": 18630599.0,
      "step": 8350
    },
    {
      "entropy": 5.506535768508911,
      "epoch": 0.9383950131970573,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.39,
      "mean_token_accuracy": 0.19271948784589768,
      "num_tokens": 18643898.0,
      "step": 8355
    },
    {
      "entropy": 5.556987667083741,
      "epoch": 0.9389565901050149,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492285090502816,
      "loss": 5.3672,
      "mean_token_accuracy": 0.18643654584884645,
      "num_tokens": 18654760.0,
      "step": 8360
    },
    {
      "entropy": 5.556431579589844,
      "epoch": 0.9395181670129724,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.3275,
      "mean_token_accuracy": 0.19593957364559172,
      "num_tokens": 18666036.0,
      "step": 8365
    },
    {
      "entropy": 5.543078088760376,
      "epoch": 0.94007974392093,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.3809,
      "mean_token_accuracy": 0.19456374645233154,
      "num_tokens": 18677806.0,
      "step": 8370
    },
    {
      "entropy": 5.600285768508911,
      "epoch": 0.9406413208288875,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000492253789434953,
      "loss": 5.4998,
      "mean_token_accuracy": 0.18756832480430602,
      "num_tokens": 18688971.0,
      "step": 8375
    },
    {
      "entropy": 5.489480304718017,
      "epoch": 0.9412028977368451,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19015772342681886,
      "num_tokens": 18700153.0,
      "step": 8380
    },
    {
      "entropy": 5.630396795272827,
      "epoch": 0.9417644746448026,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.4173,
      "mean_token_accuracy": 0.19243549704551696,
      "num_tokens": 18711935.0,
      "step": 8385
    },
    {
      "entropy": 5.44732403755188,
      "epoch": 0.9423260515527602,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.2239,
      "mean_token_accuracy": 0.1980170026421547,
      "num_tokens": 18722694.0,
      "step": 8390
    },
    {
      "entropy": 5.544261121749878,
      "epoch": 0.9428876284607177,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.3842,
      "mean_token_accuracy": 0.19403104931116105,
      "num_tokens": 18733443.0,
      "step": 8395
    },
    {
      "entropy": 5.579034423828125,
      "epoch": 0.9434492053686753,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.2937,
      "mean_token_accuracy": 0.19047629088163376,
      "num_tokens": 18743965.0,
      "step": 8400
    },
    {
      "entropy": 5.508858251571655,
      "epoch": 0.9440107822766328,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.3837,
      "mean_token_accuracy": 0.19500262141227723,
      "num_tokens": 18755712.0,
      "step": 8405
    },
    {
      "entropy": 5.510873889923095,
      "epoch": 0.9445723591845904,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.2907,
      "mean_token_accuracy": 0.19899307936429977,
      "num_tokens": 18766756.0,
      "step": 8410
    },
    {
      "entropy": 5.527020215988159,
      "epoch": 0.9451339360925479,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.3986,
      "mean_token_accuracy": 0.1874764159321785,
      "num_tokens": 18776932.0,
      "step": 8415
    },
    {
      "entropy": 5.586201143264771,
      "epoch": 0.9456955130005055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.32,
      "mean_token_accuracy": 0.1923366293311119,
      "num_tokens": 18787197.0,
      "step": 8420
    },
    {
      "entropy": 5.472904920578003,
      "epoch": 0.9462570899084629,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.2833,
      "mean_token_accuracy": 0.19781346023082733,
      "num_tokens": 18797861.0,
      "step": 8425
    },
    {
      "entropy": 5.489264488220215,
      "epoch": 0.9468186668164205,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19531847834587096,
      "num_tokens": 18809111.0,
      "step": 8430
    },
    {
      "entropy": 5.502651834487915,
      "epoch": 0.947380243724378,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.292,
      "mean_token_accuracy": 0.20616334527730942,
      "num_tokens": 18820236.0,
      "step": 8435
    },
    {
      "entropy": 5.4803853034973145,
      "epoch": 0.9479418206323356,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.291,
      "mean_token_accuracy": 0.1944515287876129,
      "num_tokens": 18831942.0,
      "step": 8440
    },
    {
      "entropy": 5.515991640090943,
      "epoch": 0.9485033975402931,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.3232,
      "mean_token_accuracy": 0.19548382312059404,
      "num_tokens": 18843271.0,
      "step": 8445
    },
    {
      "entropy": 5.54376802444458,
      "epoch": 0.9490649744482507,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.3731,
      "mean_token_accuracy": 0.19469914436340333,
      "num_tokens": 18856285.0,
      "step": 8450
    },
    {
      "entropy": 5.5682103633880615,
      "epoch": 0.9496265513562082,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049208579917436,
      "loss": 5.3826,
      "mean_token_accuracy": 0.19510563164949418,
      "num_tokens": 18867203.0,
      "step": 8455
    },
    {
      "entropy": 5.530523061752319,
      "epoch": 0.9501881282641658,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.421,
      "mean_token_accuracy": 0.1850555270910263,
      "num_tokens": 18878598.0,
      "step": 8460
    },
    {
      "entropy": 5.586306476593018,
      "epoch": 0.9507497051721233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.3313,
      "mean_token_accuracy": 0.1959939107298851,
      "num_tokens": 18890233.0,
      "step": 8465
    },
    {
      "entropy": 5.59341025352478,
      "epoch": 0.9513112820800809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.4251,
      "mean_token_accuracy": 0.18987570852041244,
      "num_tokens": 18900362.0,
      "step": 8470
    },
    {
      "entropy": 5.574532175064087,
      "epoch": 0.9518728589880384,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.3468,
      "mean_token_accuracy": 0.19702188074588775,
      "num_tokens": 18911924.0,
      "step": 8475
    },
    {
      "entropy": 5.514659261703491,
      "epoch": 0.952434435895996,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.3575,
      "mean_token_accuracy": 0.18938423544168473,
      "num_tokens": 18923931.0,
      "step": 8480
    },
    {
      "entropy": 5.496541690826416,
      "epoch": 0.9529960128039535,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.2642,
      "mean_token_accuracy": 0.19337952136993408,
      "num_tokens": 18935859.0,
      "step": 8485
    },
    {
      "entropy": 5.534044551849365,
      "epoch": 0.9535575897119111,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.3994,
      "mean_token_accuracy": 0.18973634093999864,
      "num_tokens": 18946939.0,
      "step": 8490
    },
    {
      "entropy": 5.581312084197998,
      "epoch": 0.9541191666198686,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.3517,
      "mean_token_accuracy": 0.19256744980812074,
      "num_tokens": 18958067.0,
      "step": 8495
    },
    {
      "entropy": 5.56809253692627,
      "epoch": 0.9546807435278262,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.2723,
      "mean_token_accuracy": 0.20111051946878433,
      "num_tokens": 18967920.0,
      "step": 8500
    },
    {
      "entropy": 5.532199048995972,
      "epoch": 0.9552423204357837,
      "grad_norm": 0.890625,
      "learning_rate": 0.000491979906602092,
      "loss": 5.2783,
      "mean_token_accuracy": 0.2051227569580078,
      "num_tokens": 18978244.0,
      "step": 8505
    },
    {
      "entropy": 5.5091753005981445,
      "epoch": 0.9558038973437413,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.2677,
      "mean_token_accuracy": 0.19475196450948715,
      "num_tokens": 18988464.0,
      "step": 8510
    },
    {
      "entropy": 5.440777397155761,
      "epoch": 0.9563654742516988,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.2181,
      "mean_token_accuracy": 0.19491596817970275,
      "num_tokens": 18998318.0,
      "step": 8515
    },
    {
      "entropy": 5.45198802947998,
      "epoch": 0.9569270511596563,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19848695695400237,
      "num_tokens": 19009802.0,
      "step": 8520
    },
    {
      "entropy": 5.58954553604126,
      "epoch": 0.9574886280676138,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.4242,
      "mean_token_accuracy": 0.1954422414302826,
      "num_tokens": 19020174.0,
      "step": 8525
    },
    {
      "entropy": 5.538468980789185,
      "epoch": 0.9580502049755714,
      "grad_norm": 0.828125,
      "learning_rate": 0.000491926701217703,
      "loss": 5.3552,
      "mean_token_accuracy": 0.18940722346305847,
      "num_tokens": 19031505.0,
      "step": 8530
    },
    {
      "entropy": 5.492196702957154,
      "epoch": 0.9586117818835289,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.242,
      "mean_token_accuracy": 0.20092486590147018,
      "num_tokens": 19042052.0,
      "step": 8535
    },
    {
      "entropy": 5.636459827423096,
      "epoch": 0.9591733587914865,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491905370712362,
      "loss": 5.4543,
      "mean_token_accuracy": 0.1879579856991768,
      "num_tokens": 19052050.0,
      "step": 8540
    },
    {
      "entropy": 5.5843945980072025,
      "epoch": 0.959734935699444,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.3641,
      "mean_token_accuracy": 0.1980499505996704,
      "num_tokens": 19062053.0,
      "step": 8545
    },
    {
      "entropy": 5.445322275161743,
      "epoch": 0.9602965126074016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.3066,
      "mean_token_accuracy": 0.19347066879272462,
      "num_tokens": 19073095.0,
      "step": 8550
    },
    {
      "entropy": 5.566421699523926,
      "epoch": 0.9608580895153591,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.3776,
      "mean_token_accuracy": 0.1968414694070816,
      "num_tokens": 19084521.0,
      "step": 8555
    },
    {
      "entropy": 5.569588613510132,
      "epoch": 0.9614196664233167,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.358,
      "mean_token_accuracy": 0.19272718727588653,
      "num_tokens": 19096372.0,
      "step": 8560
    },
    {
      "entropy": 5.594934177398682,
      "epoch": 0.9619812433312742,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.4749,
      "mean_token_accuracy": 0.18462338149547577,
      "num_tokens": 19108535.0,
      "step": 8565
    },
    {
      "entropy": 5.618961954116822,
      "epoch": 0.9625428202392318,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.3066,
      "mean_token_accuracy": 0.20059431940317154,
      "num_tokens": 19119480.0,
      "step": 8570
    },
    {
      "entropy": 5.506146192550659,
      "epoch": 0.9631043971471893,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1992769494652748,
      "num_tokens": 19131168.0,
      "step": 8575
    },
    {
      "entropy": 5.556567049026489,
      "epoch": 0.9636659740551469,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.4679,
      "mean_token_accuracy": 0.18865093141794204,
      "num_tokens": 19142454.0,
      "step": 8580
    },
    {
      "entropy": 5.569630146026611,
      "epoch": 0.9642275509631044,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000491809041613254,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1951150804758072,
      "num_tokens": 19154682.0,
      "step": 8585
    },
    {
      "entropy": 5.5051126956939695,
      "epoch": 0.964789127871062,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.2585,
      "mean_token_accuracy": 0.19430696368217468,
      "num_tokens": 19165190.0,
      "step": 8590
    },
    {
      "entropy": 5.533736228942871,
      "epoch": 0.9653507047790195,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.386,
      "mean_token_accuracy": 0.19778713434934617,
      "num_tokens": 19176075.0,
      "step": 8595
    },
    {
      "entropy": 5.5400341033935545,
      "epoch": 0.9659122816869771,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.3942,
      "mean_token_accuracy": 0.18972357511520385,
      "num_tokens": 19188450.0,
      "step": 8600
    },
    {
      "entropy": 5.449521780014038,
      "epoch": 0.9664738585949346,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.2448,
      "mean_token_accuracy": 0.20544234812259674,
      "num_tokens": 19201254.0,
      "step": 8605
    },
    {
      "entropy": 5.483716821670532,
      "epoch": 0.9670354355028922,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.339,
      "mean_token_accuracy": 0.19322220981121063,
      "num_tokens": 19214062.0,
      "step": 8610
    },
    {
      "entropy": 5.514604806900024,
      "epoch": 0.9675970124108496,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.2301,
      "mean_token_accuracy": 0.19938719719648362,
      "num_tokens": 19224844.0,
      "step": 8615
    },
    {
      "entropy": 5.517232751846313,
      "epoch": 0.9681585893188072,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.3884,
      "mean_token_accuracy": 0.19019622951745987,
      "num_tokens": 19237347.0,
      "step": 8620
    },
    {
      "entropy": 5.568839979171753,
      "epoch": 0.9687201662267647,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.2929,
      "mean_token_accuracy": 0.19349850863218307,
      "num_tokens": 19248743.0,
      "step": 8625
    },
    {
      "entropy": 5.5409348487854,
      "epoch": 0.9692817431347223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.2811,
      "mean_token_accuracy": 0.19889605045318604,
      "num_tokens": 19259904.0,
      "step": 8630
    },
    {
      "entropy": 5.476102256774903,
      "epoch": 0.9698433200426798,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.2135,
      "mean_token_accuracy": 0.1982925906777382,
      "num_tokens": 19271388.0,
      "step": 8635
    },
    {
      "entropy": 5.472508955001831,
      "epoch": 0.9704048969506374,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.3459,
      "mean_token_accuracy": 0.19611824601888656,
      "num_tokens": 19282790.0,
      "step": 8640
    },
    {
      "entropy": 5.476872730255127,
      "epoch": 0.9709664738585949,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.2542,
      "mean_token_accuracy": 0.20165022611618041,
      "num_tokens": 19293389.0,
      "step": 8645
    },
    {
      "entropy": 5.516427230834961,
      "epoch": 0.9715280507665525,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19275973290205,
      "num_tokens": 19304593.0,
      "step": 8650
    },
    {
      "entropy": 5.521431970596313,
      "epoch": 0.97208962767451,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19730097353458403,
      "num_tokens": 19317388.0,
      "step": 8655
    },
    {
      "entropy": 5.5133246898651125,
      "epoch": 0.9726512045824676,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.3082,
      "mean_token_accuracy": 0.19863904714584352,
      "num_tokens": 19329228.0,
      "step": 8660
    },
    {
      "entropy": 5.557421541213989,
      "epoch": 0.9732127814904251,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.2904,
      "mean_token_accuracy": 0.19996307790279388,
      "num_tokens": 19339421.0,
      "step": 8665
    },
    {
      "entropy": 5.405920171737671,
      "epoch": 0.9737743583983827,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.1978,
      "mean_token_accuracy": 0.19828363209962846,
      "num_tokens": 19349618.0,
      "step": 8670
    },
    {
      "entropy": 5.4068992137908936,
      "epoch": 0.9743359353063402,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.274,
      "mean_token_accuracy": 0.19820062071084976,
      "num_tokens": 19361324.0,
      "step": 8675
    },
    {
      "entropy": 5.5136371612548825,
      "epoch": 0.9748975122142978,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19429763406515121,
      "num_tokens": 19373631.0,
      "step": 8680
    },
    {
      "entropy": 5.512303686141967,
      "epoch": 0.9754590891222553,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.2596,
      "mean_token_accuracy": 0.1894065722823143,
      "num_tokens": 19384682.0,
      "step": 8685
    },
    {
      "entropy": 5.577161693572998,
      "epoch": 0.9760206660302129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.3596,
      "mean_token_accuracy": 0.19214255213737488,
      "num_tokens": 19396344.0,
      "step": 8690
    },
    {
      "entropy": 5.491068601608276,
      "epoch": 0.9765822429381704,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.2228,
      "mean_token_accuracy": 0.19514360427856445,
      "num_tokens": 19406921.0,
      "step": 8695
    },
    {
      "entropy": 5.4801679134368895,
      "epoch": 0.977143819846128,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491560327430032,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19372628778219222,
      "num_tokens": 19416851.0,
      "step": 8700
    },
    {
      "entropy": 5.533003902435302,
      "epoch": 0.9777053967540855,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.3108,
      "mean_token_accuracy": 0.19475093483924866,
      "num_tokens": 19428180.0,
      "step": 8705
    },
    {
      "entropy": 5.630847930908203,
      "epoch": 0.9782669736620431,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.3742,
      "mean_token_accuracy": 0.19478792548179627,
      "num_tokens": 19439575.0,
      "step": 8710
    },
    {
      "entropy": 5.536477088928223,
      "epoch": 0.9788285505700005,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.3302,
      "mean_token_accuracy": 0.19357409924268723,
      "num_tokens": 19452074.0,
      "step": 8715
    },
    {
      "entropy": 5.425706577301026,
      "epoch": 0.9793901274779581,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.2486,
      "mean_token_accuracy": 0.20036822855472564,
      "num_tokens": 19463085.0,
      "step": 8720
    },
    {
      "entropy": 5.623198986053467,
      "epoch": 0.9799517043859156,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.4146,
      "mean_token_accuracy": 0.1868090495467186,
      "num_tokens": 19472726.0,
      "step": 8725
    },
    {
      "entropy": 5.487967443466187,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049149484540214,
      "loss": 5.2552,
      "mean_token_accuracy": 0.1935631588101387,
      "num_tokens": 19482275.0,
      "step": 8730
    },
    {
      "entropy": 5.505958414077758,
      "epoch": 0.9810748582018307,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.3595,
      "mean_token_accuracy": 0.19699829518795015,
      "num_tokens": 19492782.0,
      "step": 8735
    },
    {
      "entropy": 5.446452713012695,
      "epoch": 0.9816364351097883,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.2307,
      "mean_token_accuracy": 0.1958414539694786,
      "num_tokens": 19505709.0,
      "step": 8740
    },
    {
      "entropy": 5.541211843490601,
      "epoch": 0.9821980120177458,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1925818592309952,
      "num_tokens": 19516269.0,
      "step": 8745
    },
    {
      "entropy": 5.496512413024902,
      "epoch": 0.9827595889257034,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19538244009017944,
      "num_tokens": 19526786.0,
      "step": 8750
    },
    {
      "entropy": 5.492437362670898,
      "epoch": 0.9833211658336609,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.3183,
      "mean_token_accuracy": 0.18935514986515045,
      "num_tokens": 19537865.0,
      "step": 8755
    },
    {
      "entropy": 5.424366855621338,
      "epoch": 0.9838827427416185,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.2566,
      "mean_token_accuracy": 0.2032148703932762,
      "num_tokens": 19548320.0,
      "step": 8760
    },
    {
      "entropy": 5.526584100723267,
      "epoch": 0.984444319649576,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000491418136064015,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19547163248062133,
      "num_tokens": 19558856.0,
      "step": 8765
    },
    {
      "entropy": 5.554690217971801,
      "epoch": 0.9850058965575336,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491407150020996,
      "loss": 5.3689,
      "mean_token_accuracy": 0.20002047270536422,
      "num_tokens": 19570792.0,
      "step": 8770
    },
    {
      "entropy": 5.530953502655029,
      "epoch": 0.9855674734654911,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.2727,
      "mean_token_accuracy": 0.2007561966776848,
      "num_tokens": 19581982.0,
      "step": 8775
    },
    {
      "entropy": 5.435000228881836,
      "epoch": 0.9861290503734487,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.241,
      "mean_token_accuracy": 0.19601088762283325,
      "num_tokens": 19592882.0,
      "step": 8780
    },
    {
      "entropy": 5.551732063293457,
      "epoch": 0.9866906272814062,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.4594,
      "mean_token_accuracy": 0.19115378260612487,
      "num_tokens": 19605967.0,
      "step": 8785
    },
    {
      "entropy": 5.597033596038818,
      "epoch": 0.9872522041893638,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.2916,
      "mean_token_accuracy": 0.19472094923257827,
      "num_tokens": 19616494.0,
      "step": 8790
    },
    {
      "entropy": 5.554978656768799,
      "epoch": 0.9878137810973213,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.3866,
      "mean_token_accuracy": 0.19440087527036667,
      "num_tokens": 19627013.0,
      "step": 8795
    },
    {
      "entropy": 5.5084312915802,
      "epoch": 0.9883753580052789,
      "grad_norm": 0.796875,
      "learning_rate": 0.000491341089074901,
      "loss": 5.2808,
      "mean_token_accuracy": 0.19979849457740784,
      "num_tokens": 19639099.0,
      "step": 8800
    },
    {
      "entropy": 5.423909091949463,
      "epoch": 0.9889369349132364,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491330054805827,
      "loss": 5.2667,
      "mean_token_accuracy": 0.20079948902130126,
      "num_tokens": 19650650.0,
      "step": 8805
    },
    {
      "entropy": 5.513259363174439,
      "epoch": 0.9894985118211939,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.3168,
      "mean_token_accuracy": 0.19340629875659943,
      "num_tokens": 19662869.0,
      "step": 8810
    },
    {
      "entropy": 5.486497497558593,
      "epoch": 0.9900600887291514,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1968105748295784,
      "num_tokens": 19673232.0,
      "step": 8815
    },
    {
      "entropy": 5.498173379898072,
      "epoch": 0.990621665637109,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.324,
      "mean_token_accuracy": 0.19287081211805343,
      "num_tokens": 19685069.0,
      "step": 8820
    },
    {
      "entropy": 5.54580078125,
      "epoch": 0.9911832425450665,
      "grad_norm": 0.8125,
      "learning_rate": 0.000491285848852719,
      "loss": 5.4141,
      "mean_token_accuracy": 0.18451459109783172,
      "num_tokens": 19697018.0,
      "step": 8825
    },
    {
      "entropy": 5.523934412002563,
      "epoch": 0.9917448194530241,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.2722,
      "mean_token_accuracy": 0.2014525279402733,
      "num_tokens": 19707922.0,
      "step": 8830
    },
    {
      "entropy": 5.423628091812134,
      "epoch": 0.9923063963609816,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491263704555005,
      "loss": 5.2151,
      "mean_token_accuracy": 0.19532766938209534,
      "num_tokens": 19718693.0,
      "step": 8835
    },
    {
      "entropy": 5.465753889083862,
      "epoch": 0.9928679732689392,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.2905,
      "mean_token_accuracy": 0.1943553179502487,
      "num_tokens": 19730846.0,
      "step": 8840
    },
    {
      "entropy": 5.500577354431153,
      "epoch": 0.9934295501768967,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.3017,
      "mean_token_accuracy": 0.20023206174373626,
      "num_tokens": 19742099.0,
      "step": 8845
    },
    {
      "entropy": 5.545119380950927,
      "epoch": 0.9939911270848543,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491230436464915,
      "loss": 5.3017,
      "mean_token_accuracy": 0.1975068509578705,
      "num_tokens": 19753216.0,
      "step": 8850
    },
    {
      "entropy": 5.490866613388062,
      "epoch": 0.9945527039928118,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.2465,
      "mean_token_accuracy": 0.19712071716785431,
      "num_tokens": 19765316.0,
      "step": 8855
    },
    {
      "entropy": 5.4881330013275145,
      "epoch": 0.9951142809007694,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19463828206062317,
      "num_tokens": 19776208.0,
      "step": 8860
    },
    {
      "entropy": 5.560824251174926,
      "epoch": 0.9956758578087269,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.301,
      "mean_token_accuracy": 0.19955289661884307,
      "num_tokens": 19788065.0,
      "step": 8865
    },
    {
      "entropy": 5.509456396102905,
      "epoch": 0.9962374347166845,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19480395019054414,
      "num_tokens": 19799388.0,
      "step": 8870
    },
    {
      "entropy": 5.5303130626678465,
      "epoch": 0.996799011624642,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.3259,
      "mean_token_accuracy": 0.1891831636428833,
      "num_tokens": 19809933.0,
      "step": 8875
    },
    {
      "entropy": 5.620634746551514,
      "epoch": 0.9973605885325996,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.4136,
      "mean_token_accuracy": 0.18932687044143676,
      "num_tokens": 19821566.0,
      "step": 8880
    },
    {
      "entropy": 5.571918249130249,
      "epoch": 0.9979221654405571,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.3724,
      "mean_token_accuracy": 0.19498956203460693,
      "num_tokens": 19833727.0,
      "step": 8885
    },
    {
      "entropy": 5.572908639907837,
      "epoch": 0.9984837423485147,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.392,
      "mean_token_accuracy": 0.1903043270111084,
      "num_tokens": 19845176.0,
      "step": 8890
    },
    {
      "entropy": 5.5138904571533205,
      "epoch": 0.9990453192564722,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.2911,
      "mean_token_accuracy": 0.18927495181560516,
      "num_tokens": 19857528.0,
      "step": 8895
    },
    {
      "entropy": 5.549451923370361,
      "epoch": 0.9996068961644298,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.2114,
      "mean_token_accuracy": 0.19732483476400375,
      "num_tokens": 19868232.0,
      "step": 8900
    },
    {
      "entropy": 5.4804858101738825,
      "epoch": 1.0001123153815914,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20015125473340353,
      "num_tokens": 19877881.0,
      "step": 8905
    },
    {
      "entropy": 5.396058654785156,
      "epoch": 1.000673892289549,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.2504,
      "mean_token_accuracy": 0.20254514813423158,
      "num_tokens": 19888398.0,
      "step": 8910
    },
    {
      "entropy": 5.541640901565552,
      "epoch": 1.0012354691975065,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.2093,
      "mean_token_accuracy": 0.20296085327863694,
      "num_tokens": 19899661.0,
      "step": 8915
    },
    {
      "entropy": 5.555482578277588,
      "epoch": 1.0017970461054642,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.2615,
      "mean_token_accuracy": 0.20101720690727234,
      "num_tokens": 19911073.0,
      "step": 8920
    },
    {
      "entropy": 5.463236618041992,
      "epoch": 1.0023586230134216,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.2093,
      "mean_token_accuracy": 0.20411583483219148,
      "num_tokens": 19922567.0,
      "step": 8925
    },
    {
      "entropy": 5.560108327865601,
      "epoch": 1.0029201999213793,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.296,
      "mean_token_accuracy": 0.19743030667304992,
      "num_tokens": 19934547.0,
      "step": 8930
    },
    {
      "entropy": 5.4949757099151615,
      "epoch": 1.0034817768293367,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.2888,
      "mean_token_accuracy": 0.19609466642141343,
      "num_tokens": 19947659.0,
      "step": 8935
    },
    {
      "entropy": 5.531194591522217,
      "epoch": 1.0040433537372944,
      "grad_norm": 0.96875,
      "learning_rate": 0.000491029527020594,
      "loss": 5.2054,
      "mean_token_accuracy": 0.19615259468555452,
      "num_tokens": 19958150.0,
      "step": 8940
    },
    {
      "entropy": 5.472798204421997,
      "epoch": 1.0046049306452518,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.1853,
      "mean_token_accuracy": 0.20172982066869735,
      "num_tokens": 19968173.0,
      "step": 8945
    },
    {
      "entropy": 5.453453397750854,
      "epoch": 1.0051665075532095,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.2407,
      "mean_token_accuracy": 0.19461157023906708,
      "num_tokens": 19978543.0,
      "step": 8950
    },
    {
      "entropy": 5.512823152542114,
      "epoch": 1.005728084461167,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19809273034334182,
      "num_tokens": 19989586.0,
      "step": 8955
    },
    {
      "entropy": 5.5184050559997555,
      "epoch": 1.0062896613691246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.2273,
      "mean_token_accuracy": 0.19661420583724976,
      "num_tokens": 20000524.0,
      "step": 8960
    },
    {
      "entropy": 5.567474222183227,
      "epoch": 1.006851238277082,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1945196881890297,
      "num_tokens": 20012935.0,
      "step": 8965
    },
    {
      "entropy": 5.499795770645141,
      "epoch": 1.0074128151850397,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.2742,
      "mean_token_accuracy": 0.19491551965475082,
      "num_tokens": 20024435.0,
      "step": 8970
    },
    {
      "entropy": 5.528020858764648,
      "epoch": 1.007974392092997,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.2646,
      "mean_token_accuracy": 0.1961827978491783,
      "num_tokens": 20036017.0,
      "step": 8975
    },
    {
      "entropy": 5.547647142410279,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.2806,
      "mean_token_accuracy": 0.19958003759384155,
      "num_tokens": 20047216.0,
      "step": 8980
    },
    {
      "entropy": 5.481222105026245,
      "epoch": 1.0090975459089122,
      "grad_norm": 0.875,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.1365,
      "mean_token_accuracy": 0.20559441447257995,
      "num_tokens": 20057439.0,
      "step": 8985
    },
    {
      "entropy": 5.515385723114013,
      "epoch": 1.0096591228168699,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19677991718053817,
      "num_tokens": 20069602.0,
      "step": 8990
    },
    {
      "entropy": 5.465009069442749,
      "epoch": 1.0102206997248273,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.2476,
      "mean_token_accuracy": 0.2002618208527565,
      "num_tokens": 20081366.0,
      "step": 8995
    },
    {
      "entropy": 5.524226188659668,
      "epoch": 1.010782276632785,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.2482,
      "mean_token_accuracy": 0.20423384308815,
      "num_tokens": 20091384.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.3189411881599105,
      "eval_loss": 5.32230806350708,
      "eval_mean_token_accuracy": 0.20194207033181363,
      "eval_num_tokens": 20091384.0,
      "eval_runtime": 18.4866,
      "eval_samples_per_second": 1677.59,
      "eval_steps_per_second": 209.719,
      "step": 9000
    },
    {
      "entropy": 5.583727216720581,
      "epoch": 1.0113438535407424,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908830396257045,
      "loss": 5.2568,
      "mean_token_accuracy": 0.19337106347084046,
      "num_tokens": 20102953.0,
      "step": 9005
    },
    {
      "entropy": 5.4839965343475345,
      "epoch": 1.0119054304486998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004908717232378953,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20201504230499268,
      "num_tokens": 20113672.0,
      "step": 9010
    },
    {
      "entropy": 5.496430397033691,
      "epoch": 1.0124670073566575,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004908603999766467,
      "loss": 5.2308,
      "mean_token_accuracy": 0.19618422091007232,
      "num_tokens": 20125143.0,
      "step": 9015
    },
    {
      "entropy": 5.522008228302002,
      "epoch": 1.013028584264615,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004908490698423191,
      "loss": 5.2353,
      "mean_token_accuracy": 0.19966784119606018,
      "num_tokens": 20135489.0,
      "step": 9020
    },
    {
      "entropy": 5.463982200622558,
      "epoch": 1.0135901611725726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908377328352736,
      "loss": 5.2354,
      "mean_token_accuracy": 0.20298486202955246,
      "num_tokens": 20146768.0,
      "step": 9025
    },
    {
      "entropy": 5.323273515701294,
      "epoch": 1.01415173808053,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004908263889558709,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20798300206661224,
      "num_tokens": 20156894.0,
      "step": 9030
    },
    {
      "entropy": 5.513165903091431,
      "epoch": 1.0147133149884877,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004908150382044722,
      "loss": 5.2579,
      "mean_token_accuracy": 0.20556545108556748,
      "num_tokens": 20168174.0,
      "step": 9035
    },
    {
      "entropy": 5.5859458446502686,
      "epoch": 1.0152748918964452,
      "grad_norm": 0.890625,
      "learning_rate": 0.000490803680581439,
      "loss": 5.3437,
      "mean_token_accuracy": 0.18951404690742493,
      "num_tokens": 20179962.0,
      "step": 9040
    },
    {
      "entropy": 5.51582899093628,
      "epoch": 1.0158364688044028,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490792316087133,
      "loss": 5.2449,
      "mean_token_accuracy": 0.198522213101387,
      "num_tokens": 20189842.0,
      "step": 9045
    },
    {
      "entropy": 5.469941806793213,
      "epoch": 1.0163980457123603,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004907809447219158,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20657957941293717,
      "num_tokens": 20200471.0,
      "step": 9050
    },
    {
      "entropy": 5.495346593856811,
      "epoch": 1.016959622620318,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004907695664861498,
      "loss": 5.2226,
      "mean_token_accuracy": 0.1934772327542305,
      "num_tokens": 20212145.0,
      "step": 9055
    },
    {
      "entropy": 5.508553457260132,
      "epoch": 1.0175211995282754,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004907581813801971,
      "loss": 5.1937,
      "mean_token_accuracy": 0.20555748790502548,
      "num_tokens": 20222454.0,
      "step": 9060
    },
    {
      "entropy": 5.605180501937866,
      "epoch": 1.018082776436233,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004907467894044202,
      "loss": 5.341,
      "mean_token_accuracy": 0.19292936325073243,
      "num_tokens": 20233402.0,
      "step": 9065
    },
    {
      "entropy": 5.429851055145264,
      "epoch": 1.0186443533441905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004907353905591817,
      "loss": 5.1479,
      "mean_token_accuracy": 0.20717176645994187,
      "num_tokens": 20244102.0,
      "step": 9070
    },
    {
      "entropy": 5.569330358505249,
      "epoch": 1.0192059302521481,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000490723984844845,
      "loss": 5.3466,
      "mean_token_accuracy": 0.195126311480999,
      "num_tokens": 20255003.0,
      "step": 9075
    },
    {
      "entropy": 5.512718629837036,
      "epoch": 1.0197675071601056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004907125722617727,
      "loss": 5.184,
      "mean_token_accuracy": 0.19991266429424287,
      "num_tokens": 20265211.0,
      "step": 9080
    },
    {
      "entropy": 5.477661800384522,
      "epoch": 1.0203290840680632,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004907011528103285,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20236938893795015,
      "num_tokens": 20276552.0,
      "step": 9085
    },
    {
      "entropy": 5.538715934753418,
      "epoch": 1.0208906609760207,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004906897264908759,
      "loss": 5.2848,
      "mean_token_accuracy": 0.19699635207653046,
      "num_tokens": 20286805.0,
      "step": 9090
    },
    {
      "entropy": 5.54463152885437,
      "epoch": 1.0214522378839783,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004906782933037789,
      "loss": 5.2308,
      "mean_token_accuracy": 0.20046712905168534,
      "num_tokens": 20298536.0,
      "step": 9095
    },
    {
      "entropy": 5.477355098724365,
      "epoch": 1.0220138147919358,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906668532494012,
      "loss": 5.2267,
      "mean_token_accuracy": 0.1972913846373558,
      "num_tokens": 20309842.0,
      "step": 9100
    },
    {
      "entropy": 5.463448143005371,
      "epoch": 1.0225753916998932,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906554063281074,
      "loss": 5.1551,
      "mean_token_accuracy": 0.20000912994146347,
      "num_tokens": 20320413.0,
      "step": 9105
    },
    {
      "entropy": 5.4874673843383786,
      "epoch": 1.0231369686078509,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906439525402618,
      "loss": 5.21,
      "mean_token_accuracy": 0.19910569936037065,
      "num_tokens": 20331812.0,
      "step": 9110
    },
    {
      "entropy": 5.464101123809814,
      "epoch": 1.0236985455158083,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004906324918862291,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20653669387102128,
      "num_tokens": 20342148.0,
      "step": 9115
    },
    {
      "entropy": 5.5220417976379395,
      "epoch": 1.024260122423766,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906210243663742,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1975223734974861,
      "num_tokens": 20353843.0,
      "step": 9120
    },
    {
      "entropy": 5.507647180557251,
      "epoch": 1.0248216993317234,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004906095499810622,
      "loss": 5.2507,
      "mean_token_accuracy": 0.1995355397462845,
      "num_tokens": 20364846.0,
      "step": 9125
    },
    {
      "entropy": 5.5044286251068115,
      "epoch": 1.025383276239681,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004905980687306586,
      "loss": 5.1992,
      "mean_token_accuracy": 0.20238448083400726,
      "num_tokens": 20376874.0,
      "step": 9130
    },
    {
      "entropy": 5.488220834732056,
      "epoch": 1.0259448531476385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905865806155288,
      "loss": 5.1963,
      "mean_token_accuracy": 0.20079576671123506,
      "num_tokens": 20388189.0,
      "step": 9135
    },
    {
      "entropy": 5.5182538509368895,
      "epoch": 1.0265064300555962,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905750856360387,
      "loss": 5.2382,
      "mean_token_accuracy": 0.2024891421198845,
      "num_tokens": 20399155.0,
      "step": 9140
    },
    {
      "entropy": 5.4669781684875485,
      "epoch": 1.0270680069635536,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004905635837925543,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20529061406850815,
      "num_tokens": 20410569.0,
      "step": 9145
    },
    {
      "entropy": 5.404199934005737,
      "epoch": 1.0276295838715113,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004905520750854418,
      "loss": 5.1762,
      "mean_token_accuracy": 0.20634560585021972,
      "num_tokens": 20420438.0,
      "step": 9150
    },
    {
      "entropy": 5.500921154022217,
      "epoch": 1.0281911607794687,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004905405595150674,
      "loss": 5.2257,
      "mean_token_accuracy": 0.20004838705062866,
      "num_tokens": 20431874.0,
      "step": 9155
    },
    {
      "entropy": 5.472363758087158,
      "epoch": 1.0287527376874264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905290370817982,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19746422320604323,
      "num_tokens": 20442792.0,
      "step": 9160
    },
    {
      "entropy": 5.5148707866668705,
      "epoch": 1.0293143145953838,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490517507786001,
      "loss": 5.2288,
      "mean_token_accuracy": 0.19690819084644318,
      "num_tokens": 20454410.0,
      "step": 9165
    },
    {
      "entropy": 5.598886346817016,
      "epoch": 1.0298758915033415,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004905059716280425,
      "loss": 5.3649,
      "mean_token_accuracy": 0.1899912029504776,
      "num_tokens": 20467072.0,
      "step": 9170
    },
    {
      "entropy": 5.537892818450928,
      "epoch": 1.030437468411299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904944286082904,
      "loss": 5.283,
      "mean_token_accuracy": 0.20207869708538057,
      "num_tokens": 20477697.0,
      "step": 9175
    },
    {
      "entropy": 5.48917145729065,
      "epoch": 1.0309990453192566,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004904828787271121,
      "loss": 5.1154,
      "mean_token_accuracy": 0.20494226962327958,
      "num_tokens": 20489024.0,
      "step": 9180
    },
    {
      "entropy": 5.473089170455933,
      "epoch": 1.031560622227214,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904713219848753,
      "loss": 5.2037,
      "mean_token_accuracy": 0.1965711459517479,
      "num_tokens": 20499928.0,
      "step": 9185
    },
    {
      "entropy": 5.461671495437622,
      "epoch": 1.0321221991351717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904597583819481,
      "loss": 5.2574,
      "mean_token_accuracy": 0.19312940090894698,
      "num_tokens": 20511085.0,
      "step": 9190
    },
    {
      "entropy": 5.492776775360108,
      "epoch": 1.032683776043129,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004904481879186987,
      "loss": 5.1024,
      "mean_token_accuracy": 0.20232841968536378,
      "num_tokens": 20522164.0,
      "step": 9195
    },
    {
      "entropy": 5.521366453170776,
      "epoch": 1.0332453529510865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904366105954954,
      "loss": 5.2945,
      "mean_token_accuracy": 0.20026582926511766,
      "num_tokens": 20534187.0,
      "step": 9200
    },
    {
      "entropy": 5.576307392120361,
      "epoch": 1.0338069298590442,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004904250264127067,
      "loss": 5.2574,
      "mean_token_accuracy": 0.202733638882637,
      "num_tokens": 20545819.0,
      "step": 9205
    },
    {
      "entropy": 5.588125324249267,
      "epoch": 1.0343685067670016,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004904134353707018,
      "loss": 5.2712,
      "mean_token_accuracy": 0.19890021830797194,
      "num_tokens": 20556803.0,
      "step": 9210
    },
    {
      "entropy": 5.535311460494995,
      "epoch": 1.0349300836749593,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004904018374698496,
      "loss": 5.2618,
      "mean_token_accuracy": 0.198699426651001,
      "num_tokens": 20567366.0,
      "step": 9215
    },
    {
      "entropy": 5.507753562927246,
      "epoch": 1.0354916605829168,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004903902327105191,
      "loss": 5.2962,
      "mean_token_accuracy": 0.1953882098197937,
      "num_tokens": 20579182.0,
      "step": 9220
    },
    {
      "entropy": 5.466451406478882,
      "epoch": 1.0360532374908744,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903786210930803,
      "loss": 5.1512,
      "mean_token_accuracy": 0.20047435015439988,
      "num_tokens": 20590346.0,
      "step": 9225
    },
    {
      "entropy": 5.525947904586792,
      "epoch": 1.0366148143988319,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004903670026179025,
      "loss": 5.2595,
      "mean_token_accuracy": 0.2004369929432869,
      "num_tokens": 20601279.0,
      "step": 9230
    },
    {
      "entropy": 5.495059537887573,
      "epoch": 1.0371763913067895,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490355377285356,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19908430725336074,
      "num_tokens": 20612163.0,
      "step": 9235
    },
    {
      "entropy": 5.486514663696289,
      "epoch": 1.037737968214747,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004903437450958106,
      "loss": 5.1392,
      "mean_token_accuracy": 0.20307854413986207,
      "num_tokens": 20622587.0,
      "step": 9240
    },
    {
      "entropy": 5.505575513839721,
      "epoch": 1.0382995451227046,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004903321060496369,
      "loss": 5.3125,
      "mean_token_accuracy": 0.18896912038326263,
      "num_tokens": 20633341.0,
      "step": 9245
    },
    {
      "entropy": 5.44008207321167,
      "epoch": 1.038861122030662,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004903204601472054,
      "loss": 5.1862,
      "mean_token_accuracy": 0.20291981101036072,
      "num_tokens": 20644582.0,
      "step": 9250
    },
    {
      "entropy": 5.502528667449951,
      "epoch": 1.0394226989386197,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004903088073888869,
      "loss": 5.1559,
      "mean_token_accuracy": 0.2060237005352974,
      "num_tokens": 20654936.0,
      "step": 9255
    },
    {
      "entropy": 5.479387903213501,
      "epoch": 1.0399842758465772,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004902971477750526,
      "loss": 5.18,
      "mean_token_accuracy": 0.19857182800769807,
      "num_tokens": 20666348.0,
      "step": 9260
    },
    {
      "entropy": 5.576429319381714,
      "epoch": 1.0405458527545348,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004902854813060736,
      "loss": 5.3527,
      "mean_token_accuracy": 0.19512815177440643,
      "num_tokens": 20677482.0,
      "step": 9265
    },
    {
      "entropy": 5.556309175491333,
      "epoch": 1.0411074296624923,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004902738079823213,
      "loss": 5.2584,
      "mean_token_accuracy": 0.20312489569187164,
      "num_tokens": 20688369.0,
      "step": 9270
    },
    {
      "entropy": 5.447544813156128,
      "epoch": 1.04166900657045,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004902621278041674,
      "loss": 5.2744,
      "mean_token_accuracy": 0.19016608595848083,
      "num_tokens": 20698933.0,
      "step": 9275
    },
    {
      "entropy": 5.490622854232788,
      "epoch": 1.0422305834784074,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004902504407719839,
      "loss": 5.1313,
      "mean_token_accuracy": 0.20354401618242263,
      "num_tokens": 20710192.0,
      "step": 9280
    },
    {
      "entropy": 5.493754100799561,
      "epoch": 1.042792160386365,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004902387468861429,
      "loss": 5.211,
      "mean_token_accuracy": 0.198832930624485,
      "num_tokens": 20721408.0,
      "step": 9285
    },
    {
      "entropy": 5.438203191757202,
      "epoch": 1.0433537372943225,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004902270461470169,
      "loss": 5.1104,
      "mean_token_accuracy": 0.2063740983605385,
      "num_tokens": 20732155.0,
      "step": 9290
    },
    {
      "entropy": 5.441533136367798,
      "epoch": 1.04391531420228,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000490215338554978,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20939861536026,
      "num_tokens": 20741819.0,
      "step": 9295
    },
    {
      "entropy": 5.513307857513428,
      "epoch": 1.0444768911102376,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004902036241103994,
      "loss": 5.2478,
      "mean_token_accuracy": 0.19722046107053756,
      "num_tokens": 20754931.0,
      "step": 9300
    },
    {
      "entropy": 5.499553060531616,
      "epoch": 1.045038468018195,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004901919028136539,
      "loss": 5.3198,
      "mean_token_accuracy": 0.18906616121530534,
      "num_tokens": 20766533.0,
      "step": 9305
    },
    {
      "entropy": 5.471915340423584,
      "epoch": 1.0456000449261527,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004901801746651147,
      "loss": 5.1732,
      "mean_token_accuracy": 0.20560022741556166,
      "num_tokens": 20777055.0,
      "step": 9310
    },
    {
      "entropy": 5.426851415634156,
      "epoch": 1.04616162183411,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004901684396651554,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20422170162200928,
      "num_tokens": 20786871.0,
      "step": 9315
    },
    {
      "entropy": 5.541410255432129,
      "epoch": 1.0467231987420678,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004901566978141494,
      "loss": 5.2817,
      "mean_token_accuracy": 0.1956240639090538,
      "num_tokens": 20798546.0,
      "step": 9320
    },
    {
      "entropy": 5.525883913040161,
      "epoch": 1.0472847756500252,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004901449491124708,
      "loss": 5.2088,
      "mean_token_accuracy": 0.19941770881414414,
      "num_tokens": 20809657.0,
      "step": 9325
    },
    {
      "entropy": 5.440621995925904,
      "epoch": 1.0478463525579829,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901331935604936,
      "loss": 5.2298,
      "mean_token_accuracy": 0.194198140501976,
      "num_tokens": 20819662.0,
      "step": 9330
    },
    {
      "entropy": 5.3771546363830565,
      "epoch": 1.0484079294659403,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000490121431158592,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21119288653135299,
      "num_tokens": 20830554.0,
      "step": 9335
    },
    {
      "entropy": 5.50520920753479,
      "epoch": 1.048969506373898,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004901096619071407,
      "loss": 5.2674,
      "mean_token_accuracy": 0.1961747795343399,
      "num_tokens": 20842003.0,
      "step": 9340
    },
    {
      "entropy": 5.487845182418823,
      "epoch": 1.0495310832818554,
      "grad_norm": 0.875,
      "learning_rate": 0.0004900978858065142,
      "loss": 5.2625,
      "mean_token_accuracy": 0.20116258412599564,
      "num_tokens": 20853453.0,
      "step": 9345
    },
    {
      "entropy": 5.585537052154541,
      "epoch": 1.050092660189813,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004900861028570877,
      "loss": 5.3101,
      "mean_token_accuracy": 0.19838019460439682,
      "num_tokens": 20867296.0,
      "step": 9350
    },
    {
      "entropy": 5.447028160095215,
      "epoch": 1.0506542370977705,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004900743130592362,
      "loss": 5.1996,
      "mean_token_accuracy": 0.20565367192029954,
      "num_tokens": 20878951.0,
      "step": 9355
    },
    {
      "entropy": 5.543765926361084,
      "epoch": 1.0512158140057282,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004900625164133351,
      "loss": 5.2999,
      "mean_token_accuracy": 0.1965576171875,
      "num_tokens": 20891103.0,
      "step": 9360
    },
    {
      "entropy": 5.5436663150787355,
      "epoch": 1.0517773909136856,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004900507129197601,
      "loss": 5.3617,
      "mean_token_accuracy": 0.19579455703496934,
      "num_tokens": 20903570.0,
      "step": 9365
    },
    {
      "entropy": 5.555352687835693,
      "epoch": 1.0523389678216433,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004900389025788871,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19468713104724883,
      "num_tokens": 20914156.0,
      "step": 9370
    },
    {
      "entropy": 5.464226150512696,
      "epoch": 1.0529005447296007,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000490027085391092,
      "loss": 5.212,
      "mean_token_accuracy": 0.20149350464344024,
      "num_tokens": 20925192.0,
      "step": 9375
    },
    {
      "entropy": 5.481942415237427,
      "epoch": 1.0534621216375584,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000490015261356751,
      "loss": 5.0943,
      "mean_token_accuracy": 0.21118777990341187,
      "num_tokens": 20935281.0,
      "step": 9380
    },
    {
      "entropy": 5.567909908294678,
      "epoch": 1.0540236985455158,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004900034304762408,
      "loss": 5.309,
      "mean_token_accuracy": 0.19850397258996963,
      "num_tokens": 20947397.0,
      "step": 9385
    },
    {
      "entropy": 5.432170438766479,
      "epoch": 1.0545852754534732,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000489991592749938,
      "loss": 5.1667,
      "mean_token_accuracy": 0.2089316129684448,
      "num_tokens": 20958473.0,
      "step": 9390
    },
    {
      "entropy": 5.603392505645752,
      "epoch": 1.055146852361431,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004899797481782193,
      "loss": 5.2826,
      "mean_token_accuracy": 0.19170767068862915,
      "num_tokens": 20970112.0,
      "step": 9395
    },
    {
      "entropy": 5.4718832015991214,
      "epoch": 1.0557084292693883,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004899678967614623,
      "loss": 5.191,
      "mean_token_accuracy": 0.20091700702905654,
      "num_tokens": 20980762.0,
      "step": 9400
    },
    {
      "entropy": 5.404077100753784,
      "epoch": 1.056270006177346,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004899560385000439,
      "loss": 5.0995,
      "mean_token_accuracy": 0.20930145382881166,
      "num_tokens": 20991663.0,
      "step": 9405
    },
    {
      "entropy": 5.462149143218994,
      "epoch": 1.0568315830853034,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000489944173394342,
      "loss": 5.1289,
      "mean_token_accuracy": 0.2062255322933197,
      "num_tokens": 21002121.0,
      "step": 9410
    },
    {
      "entropy": 5.473354530334473,
      "epoch": 1.0573931599932611,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004899323014447342,
      "loss": 5.1957,
      "mean_token_accuracy": 0.19004318863153458,
      "num_tokens": 21013791.0,
      "step": 9415
    },
    {
      "entropy": 5.440091562271118,
      "epoch": 1.0579547369012186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899204226515985,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20271998196840285,
      "num_tokens": 21022977.0,
      "step": 9420
    },
    {
      "entropy": 5.492621946334839,
      "epoch": 1.0585163138091762,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004899085370153133,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19832295030355454,
      "num_tokens": 21036372.0,
      "step": 9425
    },
    {
      "entropy": 5.609709692001343,
      "epoch": 1.0590778907171337,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004898966445362569,
      "loss": 5.3715,
      "mean_token_accuracy": 0.18739080280065537,
      "num_tokens": 21047984.0,
      "step": 9430
    },
    {
      "entropy": 5.559036779403686,
      "epoch": 1.0596394676250913,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004898847452148079,
      "loss": 5.2385,
      "mean_token_accuracy": 0.19740052074193953,
      "num_tokens": 21060127.0,
      "step": 9435
    },
    {
      "entropy": 5.469058179855347,
      "epoch": 1.0602010445330488,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898728390513454,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19429519325494765,
      "num_tokens": 21070659.0,
      "step": 9440
    },
    {
      "entropy": 5.467495346069336,
      "epoch": 1.0607626214410064,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898609260462483,
      "loss": 5.1467,
      "mean_token_accuracy": 0.20710503607988356,
      "num_tokens": 21081667.0,
      "step": 9445
    },
    {
      "entropy": 5.452310180664062,
      "epoch": 1.0613241983489639,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004898490061998961,
      "loss": 5.1669,
      "mean_token_accuracy": 0.2016437292098999,
      "num_tokens": 21094275.0,
      "step": 9450
    },
    {
      "entropy": 5.497532081604004,
      "epoch": 1.0618857752569215,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004898370795126681,
      "loss": 5.2257,
      "mean_token_accuracy": 0.20021361410617827,
      "num_tokens": 21106360.0,
      "step": 9455
    },
    {
      "entropy": 5.5165211200714115,
      "epoch": 1.062447352164879,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004898251459849441,
      "loss": 5.194,
      "mean_token_accuracy": 0.2019632026553154,
      "num_tokens": 21117647.0,
      "step": 9460
    },
    {
      "entropy": 5.380741357803345,
      "epoch": 1.0630089290728366,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898132056171041,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21173667162656784,
      "num_tokens": 21129599.0,
      "step": 9465
    },
    {
      "entropy": 5.475966739654541,
      "epoch": 1.063570505980794,
      "grad_norm": 0.875,
      "learning_rate": 0.0004898012584095285,
      "loss": 5.1514,
      "mean_token_accuracy": 0.2001177340745926,
      "num_tokens": 21140573.0,
      "step": 9470
    },
    {
      "entropy": 5.49524245262146,
      "epoch": 1.0641320828887517,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004897893043625975,
      "loss": 5.2613,
      "mean_token_accuracy": 0.20271366387605666,
      "num_tokens": 21151899.0,
      "step": 9475
    },
    {
      "entropy": 5.561719799041748,
      "epoch": 1.0646936597967092,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004897773434766917,
      "loss": 5.3344,
      "mean_token_accuracy": 0.189687779545784,
      "num_tokens": 21162748.0,
      "step": 9480
    },
    {
      "entropy": 5.508446550369262,
      "epoch": 1.0652552367046666,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000489765375752192,
      "loss": 5.2299,
      "mean_token_accuracy": 0.2017131581902504,
      "num_tokens": 21172962.0,
      "step": 9485
    },
    {
      "entropy": 5.532747745513916,
      "epoch": 1.0658168136126243,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004897534011894793,
      "loss": 5.2009,
      "mean_token_accuracy": 0.20353374034166336,
      "num_tokens": 21183805.0,
      "step": 9490
    },
    {
      "entropy": 5.423833847045898,
      "epoch": 1.0663783905205817,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004897414197889352,
      "loss": 5.1253,
      "mean_token_accuracy": 0.20723408013582229,
      "num_tokens": 21195129.0,
      "step": 9495
    },
    {
      "entropy": 5.404111719131469,
      "epoch": 1.0669399674285394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000489729431550941,
      "loss": 5.2491,
      "mean_token_accuracy": 0.1957872211933136,
      "num_tokens": 21206300.0,
      "step": 9500
    },
    {
      "entropy": 5.516879224777222,
      "epoch": 1.0675015443364968,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004897174364758783,
      "loss": 5.1883,
      "mean_token_accuracy": 0.19922613501548767,
      "num_tokens": 21217137.0,
      "step": 9505
    },
    {
      "entropy": 5.482291221618652,
      "epoch": 1.0680631212444545,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004897054345641291,
      "loss": 5.1546,
      "mean_token_accuracy": 0.2049860805273056,
      "num_tokens": 21228399.0,
      "step": 9510
    },
    {
      "entropy": 5.500407743453979,
      "epoch": 1.068624698152412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004896934258160757,
      "loss": 5.2742,
      "mean_token_accuracy": 0.19596217423677445,
      "num_tokens": 21239396.0,
      "step": 9515
    },
    {
      "entropy": 5.498952722549438,
      "epoch": 1.0691862750603696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896814102321004,
      "loss": 5.1819,
      "mean_token_accuracy": 0.20035670697689056,
      "num_tokens": 21249966.0,
      "step": 9520
    },
    {
      "entropy": 5.473047208786011,
      "epoch": 1.069747851968327,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004896693878125857,
      "loss": 5.1818,
      "mean_token_accuracy": 0.20321815758943557,
      "num_tokens": 21260157.0,
      "step": 9525
    },
    {
      "entropy": 5.507161808013916,
      "epoch": 1.0703094288762847,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004896573585579144,
      "loss": 5.2423,
      "mean_token_accuracy": 0.20358157753944398,
      "num_tokens": 21270738.0,
      "step": 9530
    },
    {
      "entropy": 5.459889268875122,
      "epoch": 1.070871005784242,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004896453224684695,
      "loss": 5.1929,
      "mean_token_accuracy": 0.2016464203596115,
      "num_tokens": 21282469.0,
      "step": 9535
    },
    {
      "entropy": 5.420162153244019,
      "epoch": 1.0714325826921998,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004896332795446345,
      "loss": 5.1494,
      "mean_token_accuracy": 0.19594088941812515,
      "num_tokens": 21293474.0,
      "step": 9540
    },
    {
      "entropy": 5.5055560111999515,
      "epoch": 1.0719941596001572,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896212297867924,
      "loss": 5.2751,
      "mean_token_accuracy": 0.19317637383937836,
      "num_tokens": 21304904.0,
      "step": 9545
    },
    {
      "entropy": 5.5614612102508545,
      "epoch": 1.0725557365081149,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896091731953271,
      "loss": 5.2233,
      "mean_token_accuracy": 0.19620270878076554,
      "num_tokens": 21315407.0,
      "step": 9550
    },
    {
      "entropy": 5.440782308578491,
      "epoch": 1.0731173134160723,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895971097706227,
      "loss": 5.1531,
      "mean_token_accuracy": 0.20602274537086487,
      "num_tokens": 21326496.0,
      "step": 9555
    },
    {
      "entropy": 5.476405477523803,
      "epoch": 1.07367889032403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004895850395130629,
      "loss": 5.2445,
      "mean_token_accuracy": 0.19865497350692748,
      "num_tokens": 21339449.0,
      "step": 9560
    },
    {
      "entropy": 5.495727062225342,
      "epoch": 1.0742404672319874,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004895729624230323,
      "loss": 5.2984,
      "mean_token_accuracy": 0.19452223628759385,
      "num_tokens": 21350431.0,
      "step": 9565
    },
    {
      "entropy": 5.462311553955078,
      "epoch": 1.074802044139945,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004895608785009153,
      "loss": 5.185,
      "mean_token_accuracy": 0.20112013518810273,
      "num_tokens": 21361168.0,
      "step": 9570
    },
    {
      "entropy": 5.50471568107605,
      "epoch": 1.0753636210479025,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004895487877470969,
      "loss": 5.1944,
      "mean_token_accuracy": 0.20499631464481355,
      "num_tokens": 21372635.0,
      "step": 9575
    },
    {
      "entropy": 5.435432958602905,
      "epoch": 1.07592519795586,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895366901619617,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19774627536535264,
      "num_tokens": 21384051.0,
      "step": 9580
    },
    {
      "entropy": 5.450306129455567,
      "epoch": 1.0764867748638176,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004895245857458951,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19065044969320297,
      "num_tokens": 21397496.0,
      "step": 9585
    },
    {
      "entropy": 5.439748001098633,
      "epoch": 1.077048351771775,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004895124744992825,
      "loss": 5.2761,
      "mean_token_accuracy": 0.1959267020225525,
      "num_tokens": 21409349.0,
      "step": 9590
    },
    {
      "entropy": 5.500283527374267,
      "epoch": 1.0776099286797327,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004895003564225095,
      "loss": 5.2343,
      "mean_token_accuracy": 0.2013654440641403,
      "num_tokens": 21421843.0,
      "step": 9595
    },
    {
      "entropy": 5.4338454246521,
      "epoch": 1.0781715055876901,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004894882315159621,
      "loss": 5.1038,
      "mean_token_accuracy": 0.203974212706089,
      "num_tokens": 21433230.0,
      "step": 9600
    },
    {
      "entropy": 5.507578992843628,
      "epoch": 1.0787330824956478,
      "grad_norm": 0.9375,
      "learning_rate": 0.000489476099780026,
      "loss": 5.226,
      "mean_token_accuracy": 0.19867302775382994,
      "num_tokens": 21444295.0,
      "step": 9605
    },
    {
      "entropy": 5.505974197387696,
      "epoch": 1.0792946594036052,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004894639612150879,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1997839853167534,
      "num_tokens": 21455721.0,
      "step": 9610
    },
    {
      "entropy": 5.469446182250977,
      "epoch": 1.079856236311563,
      "grad_norm": 0.875,
      "learning_rate": 0.0004894518158215339,
      "loss": 5.2085,
      "mean_token_accuracy": 0.20309991687536239,
      "num_tokens": 21466514.0,
      "step": 9615
    },
    {
      "entropy": 5.479997968673706,
      "epoch": 1.0804178132195204,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000489439663599751,
      "loss": 5.2424,
      "mean_token_accuracy": 0.19875058829784392,
      "num_tokens": 21478782.0,
      "step": 9620
    },
    {
      "entropy": 5.5083568572998045,
      "epoch": 1.080979390127478,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000489427504550126,
      "loss": 5.207,
      "mean_token_accuracy": 0.20856829285621642,
      "num_tokens": 21490428.0,
      "step": 9625
    },
    {
      "entropy": 5.390064907073975,
      "epoch": 1.0815409670354355,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894153386730462,
      "loss": 5.0636,
      "mean_token_accuracy": 0.2056634843349457,
      "num_tokens": 21501530.0,
      "step": 9630
    },
    {
      "entropy": 5.457349157333374,
      "epoch": 1.0821025439433931,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004894031659688988,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20669750720262528,
      "num_tokens": 21512122.0,
      "step": 9635
    },
    {
      "entropy": 5.423903083801269,
      "epoch": 1.0826641208513506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893909864380714,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20614136904478073,
      "num_tokens": 21521896.0,
      "step": 9640
    },
    {
      "entropy": 5.446603393554687,
      "epoch": 1.0832256977593082,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489378800080952,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1994054138660431,
      "num_tokens": 21533584.0,
      "step": 9645
    },
    {
      "entropy": 5.423087167739868,
      "epoch": 1.0837872746672657,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004893666068979283,
      "loss": 5.2331,
      "mean_token_accuracy": 0.20098746865987777,
      "num_tokens": 21545432.0,
      "step": 9650
    },
    {
      "entropy": 5.487134695053101,
      "epoch": 1.0843488515752233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004893544068893887,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19565849304199218,
      "num_tokens": 21556547.0,
      "step": 9655
    },
    {
      "entropy": 5.498757028579712,
      "epoch": 1.0849104284831808,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004893422000557217,
      "loss": 5.2763,
      "mean_token_accuracy": 0.19918344020843506,
      "num_tokens": 21567354.0,
      "step": 9660
    },
    {
      "entropy": 5.472388076782226,
      "epoch": 1.0854720053911384,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004893299863973159,
      "loss": 5.1548,
      "mean_token_accuracy": 0.20034267902374267,
      "num_tokens": 21578694.0,
      "step": 9665
    },
    {
      "entropy": 5.425048875808716,
      "epoch": 1.0860335822990959,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004893177659145601,
      "loss": 5.112,
      "mean_token_accuracy": 0.21066864877939223,
      "num_tokens": 21589365.0,
      "step": 9670
    },
    {
      "entropy": 5.487691116333008,
      "epoch": 1.0865951592070533,
      "grad_norm": 1.0,
      "learning_rate": 0.0004893055386078438,
      "loss": 5.1736,
      "mean_token_accuracy": 0.19831692576408386,
      "num_tokens": 21600766.0,
      "step": 9675
    },
    {
      "entropy": 5.518604183197022,
      "epoch": 1.087156736115011,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004892933044775558,
      "loss": 5.3533,
      "mean_token_accuracy": 0.1903294190764427,
      "num_tokens": 21613969.0,
      "step": 9680
    },
    {
      "entropy": 5.443315315246582,
      "epoch": 1.0877183130229684,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004892810635240859,
      "loss": 5.1925,
      "mean_token_accuracy": 0.20269359052181243,
      "num_tokens": 21626302.0,
      "step": 9685
    },
    {
      "entropy": 5.517415285110474,
      "epoch": 1.088279889930926,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004892688157478238,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19961371272802353,
      "num_tokens": 21637058.0,
      "step": 9690
    },
    {
      "entropy": 5.529087495803833,
      "epoch": 1.0888414668388835,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004892565611491597,
      "loss": 5.2549,
      "mean_token_accuracy": 0.1948593810200691,
      "num_tokens": 21648729.0,
      "step": 9695
    },
    {
      "entropy": 5.3914531707763675,
      "epoch": 1.0894030437468412,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004892442997284834,
      "loss": 5.0555,
      "mean_token_accuracy": 0.21181618720293044,
      "num_tokens": 21659961.0,
      "step": 9700
    },
    {
      "entropy": 5.424667549133301,
      "epoch": 1.0899646206547986,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004892320314861855,
      "loss": 5.1851,
      "mean_token_accuracy": 0.20293384343385695,
      "num_tokens": 21670577.0,
      "step": 9705
    },
    {
      "entropy": 5.436867141723633,
      "epoch": 1.0905261975627563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892197564226568,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20987384170293807,
      "num_tokens": 21679882.0,
      "step": 9710
    },
    {
      "entropy": 5.49023928642273,
      "epoch": 1.0910877744707137,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004892074745382878,
      "loss": 5.2676,
      "mean_token_accuracy": 0.20245472937822342,
      "num_tokens": 21690595.0,
      "step": 9715
    },
    {
      "entropy": 5.464753103256226,
      "epoch": 1.0916493513786714,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004891951858334698,
      "loss": 5.136,
      "mean_token_accuracy": 0.2062764808535576,
      "num_tokens": 21701251.0,
      "step": 9720
    },
    {
      "entropy": 5.532423591613769,
      "epoch": 1.0922109282866288,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004891828903085939,
      "loss": 5.3047,
      "mean_token_accuracy": 0.19762960970401763,
      "num_tokens": 21713286.0,
      "step": 9725
    },
    {
      "entropy": 5.527106380462646,
      "epoch": 1.0927725051945865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891705879640518,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1951225608587265,
      "num_tokens": 21725508.0,
      "step": 9730
    },
    {
      "entropy": 5.4577264308929445,
      "epoch": 1.093334082102544,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004891582788002351,
      "loss": 5.157,
      "mean_token_accuracy": 0.200393471121788,
      "num_tokens": 21736467.0,
      "step": 9735
    },
    {
      "entropy": 5.542578458786011,
      "epoch": 1.0938956590105016,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004891459628175358,
      "loss": 5.2538,
      "mean_token_accuracy": 0.20124907940626144,
      "num_tokens": 21746114.0,
      "step": 9740
    },
    {
      "entropy": 5.439834117889404,
      "epoch": 1.094457235918459,
      "grad_norm": 0.84375,
      "learning_rate": 0.000489133640016346,
      "loss": 5.2539,
      "mean_token_accuracy": 0.20151838213205336,
      "num_tokens": 21757145.0,
      "step": 9745
    },
    {
      "entropy": 5.491384601593017,
      "epoch": 1.0950188128264167,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004891213103970579,
      "loss": 5.2419,
      "mean_token_accuracy": 0.1950704574584961,
      "num_tokens": 21768294.0,
      "step": 9750
    },
    {
      "entropy": 5.5173389434814455,
      "epoch": 1.095580389734374,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004891089739600643,
      "loss": 5.2482,
      "mean_token_accuracy": 0.198778235912323,
      "num_tokens": 21779667.0,
      "step": 9755
    },
    {
      "entropy": 5.372928142547607,
      "epoch": 1.0961419666423318,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890966307057581,
      "loss": 5.1024,
      "mean_token_accuracy": 0.20430002510547637,
      "num_tokens": 21789635.0,
      "step": 9760
    },
    {
      "entropy": 5.38895993232727,
      "epoch": 1.0967035435502892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000489084280634532,
      "loss": 5.0931,
      "mean_token_accuracy": 0.2070479452610016,
      "num_tokens": 21799462.0,
      "step": 9765
    },
    {
      "entropy": 5.384057474136353,
      "epoch": 1.0972651204582466,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004890719237467794,
      "loss": 5.1536,
      "mean_token_accuracy": 0.2038656949996948,
      "num_tokens": 21811520.0,
      "step": 9770
    },
    {
      "entropy": 5.5412544250488285,
      "epoch": 1.0978266973662043,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004890595600428937,
      "loss": 5.2188,
      "mean_token_accuracy": 0.2000924363732338,
      "num_tokens": 21821997.0,
      "step": 9775
    },
    {
      "entropy": 5.500632238388062,
      "epoch": 1.0983882742741617,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004890471895232684,
      "loss": 5.3001,
      "mean_token_accuracy": 0.19372042864561081,
      "num_tokens": 21833353.0,
      "step": 9780
    },
    {
      "entropy": 5.440681743621826,
      "epoch": 1.0989498511821194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004890348121882978,
      "loss": 5.1997,
      "mean_token_accuracy": 0.2040751725435257,
      "num_tokens": 21843691.0,
      "step": 9785
    },
    {
      "entropy": 5.427799654006958,
      "epoch": 1.0995114280900768,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004890224280383758,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20187733620405196,
      "num_tokens": 21853771.0,
      "step": 9790
    },
    {
      "entropy": 5.376798343658447,
      "epoch": 1.1000730049980345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004890100370738966,
      "loss": 5.0538,
      "mean_token_accuracy": 0.20886366069316864,
      "num_tokens": 21863424.0,
      "step": 9795
    },
    {
      "entropy": 5.479954528808594,
      "epoch": 1.100634581905992,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004889976392952548,
      "loss": 5.1764,
      "mean_token_accuracy": 0.20241716057062148,
      "num_tokens": 21875407.0,
      "step": 9800
    },
    {
      "entropy": 5.504074382781982,
      "epoch": 1.1011961588139496,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004889852347028452,
      "loss": 5.2538,
      "mean_token_accuracy": 0.19704989045858384,
      "num_tokens": 21887081.0,
      "step": 9805
    },
    {
      "entropy": 5.442853927612305,
      "epoch": 1.101757735721907,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889728232970626,
      "loss": 5.1446,
      "mean_token_accuracy": 0.20125831961631774,
      "num_tokens": 21897858.0,
      "step": 9810
    },
    {
      "entropy": 5.461687278747559,
      "epoch": 1.1023193126298647,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889604050783025,
      "loss": 5.1929,
      "mean_token_accuracy": 0.2005182236433029,
      "num_tokens": 21907985.0,
      "step": 9815
    },
    {
      "entropy": 5.402402591705322,
      "epoch": 1.1028808895378222,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048894798004696,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19613478630781173,
      "num_tokens": 21918726.0,
      "step": 9820
    },
    {
      "entropy": 5.437289237976074,
      "epoch": 1.1034424664457798,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889355482034309,
      "loss": 5.155,
      "mean_token_accuracy": 0.2014207124710083,
      "num_tokens": 21929027.0,
      "step": 9825
    },
    {
      "entropy": 5.447432661056519,
      "epoch": 1.1040040433537373,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004889231095481111,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20712201595306395,
      "num_tokens": 21938356.0,
      "step": 9830
    },
    {
      "entropy": 5.424928522109985,
      "epoch": 1.104565620261695,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004889106640813964,
      "loss": 5.145,
      "mean_token_accuracy": 0.19907539188861847,
      "num_tokens": 21947889.0,
      "step": 9835
    },
    {
      "entropy": 5.466288805007935,
      "epoch": 1.1051271971696524,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004888982118036831,
      "loss": 5.2122,
      "mean_token_accuracy": 0.20456399768590927,
      "num_tokens": 21958409.0,
      "step": 9840
    },
    {
      "entropy": 5.431293201446533,
      "epoch": 1.10568877407761,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004888857527153679,
      "loss": 5.2104,
      "mean_token_accuracy": 0.1990182027220726,
      "num_tokens": 21970120.0,
      "step": 9845
    },
    {
      "entropy": 5.476894378662109,
      "epoch": 1.1062503509855675,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004888732868168473,
      "loss": 5.175,
      "mean_token_accuracy": 0.20790013670921326,
      "num_tokens": 21981499.0,
      "step": 9850
    },
    {
      "entropy": 5.395249557495117,
      "epoch": 1.1068119278935251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004888608141085183,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20860708951950074,
      "num_tokens": 21991263.0,
      "step": 9855
    },
    {
      "entropy": 5.4008495807647705,
      "epoch": 1.1073735048014826,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888483345907781,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2011837899684906,
      "num_tokens": 22001582.0,
      "step": 9860
    },
    {
      "entropy": 5.442922353744507,
      "epoch": 1.10793508170944,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000488835848264024,
      "loss": 5.1606,
      "mean_token_accuracy": 0.2074039801955223,
      "num_tokens": 22012447.0,
      "step": 9865
    },
    {
      "entropy": 5.4711754322052,
      "epoch": 1.1084966586173977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004888233551286534,
      "loss": 5.2498,
      "mean_token_accuracy": 0.19839707762002945,
      "num_tokens": 22024865.0,
      "step": 9870
    },
    {
      "entropy": 5.407338285446167,
      "epoch": 1.109058235525355,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004888108551850644,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20068008899688722,
      "num_tokens": 22036298.0,
      "step": 9875
    },
    {
      "entropy": 5.506173849105835,
      "epoch": 1.1096198124333128,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887983484336548,
      "loss": 5.1779,
      "mean_token_accuracy": 0.20358308255672455,
      "num_tokens": 22047499.0,
      "step": 9880
    },
    {
      "entropy": 5.49197359085083,
      "epoch": 1.1101813893412702,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004887858348748229,
      "loss": 5.2004,
      "mean_token_accuracy": 0.204771189391613,
      "num_tokens": 22057612.0,
      "step": 9885
    },
    {
      "entropy": 5.463781356811523,
      "epoch": 1.1107429662492279,
      "grad_norm": 0.90625,
      "learning_rate": 0.000488773314508967,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19992247968912125,
      "num_tokens": 22068566.0,
      "step": 9890
    },
    {
      "entropy": 5.419338655471802,
      "epoch": 1.1113045431571853,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004887607873364859,
      "loss": 5.1273,
      "mean_token_accuracy": 0.20196017026901245,
      "num_tokens": 22079821.0,
      "step": 9895
    },
    {
      "entropy": 5.438948631286621,
      "epoch": 1.111866120065143,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004887482533577786,
      "loss": 5.1689,
      "mean_token_accuracy": 0.19802613258361818,
      "num_tokens": 22090143.0,
      "step": 9900
    },
    {
      "entropy": 5.403488731384277,
      "epoch": 1.1124276969731004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004887357125732438,
      "loss": 5.2256,
      "mean_token_accuracy": 0.20469283014535905,
      "num_tokens": 22101601.0,
      "step": 9905
    },
    {
      "entropy": 5.418672657012939,
      "epoch": 1.112989273881058,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488723164983281,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20835370272397996,
      "num_tokens": 22113102.0,
      "step": 9910
    },
    {
      "entropy": 5.528254222869873,
      "epoch": 1.1135508507890155,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004887106105882898,
      "loss": 5.2779,
      "mean_token_accuracy": 0.1999873161315918,
      "num_tokens": 22124650.0,
      "step": 9915
    },
    {
      "entropy": 5.521338939666748,
      "epoch": 1.1141124276969732,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004886980493886699,
      "loss": 5.2067,
      "mean_token_accuracy": 0.2058244228363037,
      "num_tokens": 22136637.0,
      "step": 9920
    },
    {
      "entropy": 5.4421162605285645,
      "epoch": 1.1146740046049306,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004886854813848212,
      "loss": 5.1248,
      "mean_token_accuracy": 0.20553045719861984,
      "num_tokens": 22149318.0,
      "step": 9925
    },
    {
      "entropy": 5.325225496292115,
      "epoch": 1.1152355815128883,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004886729065771439,
      "loss": 5.0283,
      "mean_token_accuracy": 0.20389310419559478,
      "num_tokens": 22159410.0,
      "step": 9930
    },
    {
      "entropy": 5.426838827133179,
      "epoch": 1.1157971584208457,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886603249660384,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20301332771778108,
      "num_tokens": 22170017.0,
      "step": 9935
    },
    {
      "entropy": 5.4066869735717775,
      "epoch": 1.1163587353288034,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886477365519052,
      "loss": 5.1446,
      "mean_token_accuracy": 0.20368731021881104,
      "num_tokens": 22179961.0,
      "step": 9940
    },
    {
      "entropy": 5.476810169219971,
      "epoch": 1.1169203122367608,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004886351413351453,
      "loss": 5.2501,
      "mean_token_accuracy": 0.1994064450263977,
      "num_tokens": 22191415.0,
      "step": 9945
    },
    {
      "entropy": 5.526557970046997,
      "epoch": 1.1174818891447185,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004886225393161596,
      "loss": 5.2173,
      "mean_token_accuracy": 0.21241141259670257,
      "num_tokens": 22205354.0,
      "step": 9950
    },
    {
      "entropy": 5.5682755470275875,
      "epoch": 1.118043466052676,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004886099304953495,
      "loss": 5.247,
      "mean_token_accuracy": 0.19583525955677034,
      "num_tokens": 22216989.0,
      "step": 9955
    },
    {
      "entropy": 5.50190839767456,
      "epoch": 1.1186050429606333,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004885973148731162,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19580172151327133,
      "num_tokens": 22228421.0,
      "step": 9960
    },
    {
      "entropy": 5.464150047302246,
      "epoch": 1.119166619868591,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004885846924498617,
      "loss": 5.121,
      "mean_token_accuracy": 0.20090800225734712,
      "num_tokens": 22240621.0,
      "step": 9965
    },
    {
      "entropy": 5.37536358833313,
      "epoch": 1.1197281967765484,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004885720632259876,
      "loss": 5.1199,
      "mean_token_accuracy": 0.20998639315366746,
      "num_tokens": 22250936.0,
      "step": 9970
    },
    {
      "entropy": 5.427014493942261,
      "epoch": 1.120289773684506,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004885594272018963,
      "loss": 5.2387,
      "mean_token_accuracy": 0.2014521688222885,
      "num_tokens": 22261766.0,
      "step": 9975
    },
    {
      "entropy": 5.487332344055176,
      "epoch": 1.1208513505924635,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048854678437799,
      "loss": 5.1579,
      "mean_token_accuracy": 0.20486941784620286,
      "num_tokens": 22271831.0,
      "step": 9980
    },
    {
      "entropy": 5.405507898330688,
      "epoch": 1.1214129275004212,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004885341347546713,
      "loss": 5.1642,
      "mean_token_accuracy": 0.20529917031526565,
      "num_tokens": 22283756.0,
      "step": 9985
    },
    {
      "entropy": 5.389522361755371,
      "epoch": 1.1219745044083786,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004885214783323428,
      "loss": 5.1341,
      "mean_token_accuracy": 0.19990193098783493,
      "num_tokens": 22294576.0,
      "step": 9990
    },
    {
      "entropy": 5.433452892303467,
      "epoch": 1.1225360813163363,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885088151114076,
      "loss": 5.194,
      "mean_token_accuracy": 0.20485766381025314,
      "num_tokens": 22305003.0,
      "step": 9995
    },
    {
      "entropy": 5.4245411396026615,
      "epoch": 1.1230976582242937,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004884961450922691,
      "loss": 5.1628,
      "mean_token_accuracy": 0.20570219159126282,
      "num_tokens": 22316505.0,
      "step": 10000
    },
    {
      "entropy": 5.469008493423462,
      "epoch": 1.1236592351322514,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004884834682753305,
      "loss": 5.2264,
      "mean_token_accuracy": 0.20076436698436737,
      "num_tokens": 22327853.0,
      "step": 10005
    },
    {
      "entropy": 5.483531427383423,
      "epoch": 1.1242208120402089,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004884707846609953,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20654843151569366,
      "num_tokens": 22337386.0,
      "step": 10010
    },
    {
      "entropy": 5.4929742336273195,
      "epoch": 1.1247823889481665,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004884580942496677,
      "loss": 5.318,
      "mean_token_accuracy": 0.1976382926106453,
      "num_tokens": 22348742.0,
      "step": 10015
    },
    {
      "entropy": 5.5046240329742435,
      "epoch": 1.125343965856124,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004884453970417515,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1997467577457428,
      "num_tokens": 22360083.0,
      "step": 10020
    },
    {
      "entropy": 5.460844087600708,
      "epoch": 1.1259055427640816,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488432693037651,
      "loss": 5.1693,
      "mean_token_accuracy": 0.19988480806350709,
      "num_tokens": 22370982.0,
      "step": 10025
    },
    {
      "entropy": 5.420691251754761,
      "epoch": 1.126467119672039,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000488419982237771,
      "loss": 5.1526,
      "mean_token_accuracy": 0.2097501888871193,
      "num_tokens": 22382383.0,
      "step": 10030
    },
    {
      "entropy": 5.4389097690582275,
      "epoch": 1.1270286965799967,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004884072646425157,
      "loss": 5.1665,
      "mean_token_accuracy": 0.19761568903923035,
      "num_tokens": 22393994.0,
      "step": 10035
    },
    {
      "entropy": 5.352759122848511,
      "epoch": 1.1275902734879542,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004883945402522907,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2109847366809845,
      "num_tokens": 22405354.0,
      "step": 10040
    },
    {
      "entropy": 5.398524761199951,
      "epoch": 1.1281518503959118,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004883818090675005,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20210367292165757,
      "num_tokens": 22416708.0,
      "step": 10045
    },
    {
      "entropy": 5.3918815612792965,
      "epoch": 1.1287134273038693,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004883690710885507,
      "loss": 5.1226,
      "mean_token_accuracy": 0.20373654812574388,
      "num_tokens": 22427724.0,
      "step": 10050
    },
    {
      "entropy": 5.479042482376099,
      "epoch": 1.1292750042118267,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004883563263158471,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19887548238039016,
      "num_tokens": 22437823.0,
      "step": 10055
    },
    {
      "entropy": 5.485242176055908,
      "epoch": 1.1298365811197844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883435747497952,
      "loss": 5.2367,
      "mean_token_accuracy": 0.2065964385867119,
      "num_tokens": 22450155.0,
      "step": 10060
    },
    {
      "entropy": 5.501348257064819,
      "epoch": 1.130398158027742,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883308163908012,
      "loss": 5.1666,
      "mean_token_accuracy": 0.19969362765550613,
      "num_tokens": 22460484.0,
      "step": 10065
    },
    {
      "entropy": 5.432768678665161,
      "epoch": 1.1309597349356995,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004883180512392714,
      "loss": 5.1926,
      "mean_token_accuracy": 0.20196976214647294,
      "num_tokens": 22472063.0,
      "step": 10070
    },
    {
      "entropy": 5.421315336227417,
      "epoch": 1.131521311843657,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004883052792956119,
      "loss": 5.1724,
      "mean_token_accuracy": 0.20803795158863067,
      "num_tokens": 22483669.0,
      "step": 10075
    },
    {
      "entropy": 5.5212242126464846,
      "epoch": 1.1320828887516146,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004882925005602296,
      "loss": 5.2428,
      "mean_token_accuracy": 0.19664522558450698,
      "num_tokens": 22495505.0,
      "step": 10080
    },
    {
      "entropy": 5.4770959377288815,
      "epoch": 1.132644465659572,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004882797150335314,
      "loss": 5.2166,
      "mean_token_accuracy": 0.198384989798069,
      "num_tokens": 22505260.0,
      "step": 10085
    },
    {
      "entropy": 5.46422848701477,
      "epoch": 1.1332060425675297,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048826692271592434,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19802864491939545,
      "num_tokens": 22516362.0,
      "step": 10090
    },
    {
      "entropy": 5.519613647460938,
      "epoch": 1.133767619475487,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004882541236078158,
      "loss": 5.2171,
      "mean_token_accuracy": 0.19751207828521727,
      "num_tokens": 22527760.0,
      "step": 10095
    },
    {
      "entropy": 5.597751760482788,
      "epoch": 1.1343291963834448,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048824131770961316,
      "loss": 5.314,
      "mean_token_accuracy": 0.192897330224514,
      "num_tokens": 22540601.0,
      "step": 10100
    },
    {
      "entropy": 5.436729574203492,
      "epoch": 1.1348907732914022,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048822850502172437,
      "loss": 5.2103,
      "mean_token_accuracy": 0.19936495274305344,
      "num_tokens": 22552469.0,
      "step": 10105
    },
    {
      "entropy": 5.3802038669586185,
      "epoch": 1.1354523501993599,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004882156855445571,
      "loss": 5.16,
      "mean_token_accuracy": 0.2056436613202095,
      "num_tokens": 22563552.0,
      "step": 10110
    },
    {
      "entropy": 5.454670095443726,
      "epoch": 1.1360139271073173,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004882028592785199,
      "loss": 5.1567,
      "mean_token_accuracy": 0.20039760023355485,
      "num_tokens": 22575619.0,
      "step": 10115
    },
    {
      "entropy": 5.393833065032959,
      "epoch": 1.136575504015275,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048819002622402095,
      "loss": 5.1408,
      "mean_token_accuracy": 0.20351850092411042,
      "num_tokens": 22585854.0,
      "step": 10120
    },
    {
      "entropy": 5.495423936843872,
      "epoch": 1.1371370809232324,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004881771863814688,
      "loss": 5.1452,
      "mean_token_accuracy": 0.21022045910358428,
      "num_tokens": 22595899.0,
      "step": 10125
    },
    {
      "entropy": 5.46891565322876,
      "epoch": 1.13769865783119,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048816433975127244,
      "loss": 5.1262,
      "mean_token_accuracy": 0.2041967362165451,
      "num_tokens": 22607517.0,
      "step": 10130
    },
    {
      "entropy": 5.447305297851562,
      "epoch": 1.1382602347391475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004881514863338407,
      "loss": 5.2343,
      "mean_token_accuracy": 0.201247338950634,
      "num_tokens": 22618531.0,
      "step": 10135
    },
    {
      "entropy": 5.369872903823852,
      "epoch": 1.1388218116471052,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048813862612958307,
      "loss": 5.0846,
      "mean_token_accuracy": 0.2056419864296913,
      "num_tokens": 22629193.0,
      "step": 10140
    },
    {
      "entropy": 5.504584550857544,
      "epoch": 1.1393833885550626,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004881257591389089,
      "loss": 5.2001,
      "mean_token_accuracy": 0.20484146624803543,
      "num_tokens": 22639794.0,
      "step": 10145
    },
    {
      "entropy": 5.500708436965942,
      "epoch": 1.13994496546302,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048811288536222794,
      "loss": 5.1459,
      "mean_token_accuracy": 0.2108313485980034,
      "num_tokens": 22650061.0,
      "step": 10150
    },
    {
      "entropy": 5.47721037864685,
      "epoch": 1.1405065423709777,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048810000479995,
      "loss": 5.2162,
      "mean_token_accuracy": 0.19785867333412172,
      "num_tokens": 22661399.0,
      "step": 10155
    },
    {
      "entropy": 5.5784751892089846,
      "epoch": 1.1410681192789354,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048808711745248524,
      "loss": 5.3067,
      "mean_token_accuracy": 0.19488762468099594,
      "num_tokens": 22672445.0,
      "step": 10160
    },
    {
      "entropy": 5.492064809799194,
      "epoch": 1.1416296961868928,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048807422332024404,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20528274178504943,
      "num_tokens": 22682545.0,
      "step": 10165
    },
    {
      "entropy": 5.460781526565552,
      "epoch": 1.1421912730948502,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004880613224036369,
      "loss": 5.1952,
      "mean_token_accuracy": 0.19861621856689454,
      "num_tokens": 22693885.0,
      "step": 10170
    },
    {
      "entropy": 5.429856681823731,
      "epoch": 1.142752850002808,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048804841470307466,
      "loss": 5.1185,
      "mean_token_accuracy": 0.2062612310051918,
      "num_tokens": 22703841.0,
      "step": 10175
    },
    {
      "entropy": 5.505462884902954,
      "epoch": 1.1433144269107653,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048803550021896824,
      "loss": 5.276,
      "mean_token_accuracy": 0.19903681725263594,
      "num_tokens": 22714142.0,
      "step": 10180
    },
    {
      "entropy": 5.498914051055908,
      "epoch": 1.143876003818723,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004880225789517289,
      "loss": 5.1526,
      "mean_token_accuracy": 0.1991946205496788,
      "num_tokens": 22725056.0,
      "step": 10185
    },
    {
      "entropy": 5.409213209152222,
      "epoch": 1.1444375807266804,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000488009650901768,
      "loss": 5.1839,
      "mean_token_accuracy": 0.2094482570886612,
      "num_tokens": 22735664.0,
      "step": 10190
    },
    {
      "entropy": 5.412260246276856,
      "epoch": 1.144999157634638,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048799671606949724,
      "loss": 5.1714,
      "mean_token_accuracy": 0.20313270390033722,
      "num_tokens": 22748071.0,
      "step": 10195
    },
    {
      "entropy": 5.5024231433868405,
      "epoch": 1.1455607345425955,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048798377445532837,
      "loss": 5.2221,
      "mean_token_accuracy": 0.20186836272478104,
      "num_tokens": 22759477.0,
      "step": 10200
    },
    {
      "entropy": 5.418139171600342,
      "epoch": 1.1461223114505532,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004879708260596736,
      "loss": 5.0972,
      "mean_token_accuracy": 0.2112067773938179,
      "num_tokens": 22769955.0,
      "step": 10205
    },
    {
      "entropy": 5.372469854354859,
      "epoch": 1.1466838883585107,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004879578708829451,
      "loss": 5.1874,
      "mean_token_accuracy": 0.20161296129226686,
      "num_tokens": 22781216.0,
      "step": 10210
    },
    {
      "entropy": 5.448965740203858,
      "epoch": 1.1472454652664683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004879449089255553,
      "loss": 5.1296,
      "mean_token_accuracy": 0.20125143676996232,
      "num_tokens": 22792581.0,
      "step": 10215
    },
    {
      "entropy": 5.421170234680176,
      "epoch": 1.1478070421744258,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004879319401879171,
      "loss": 5.0927,
      "mean_token_accuracy": 0.205929633975029,
      "num_tokens": 22803305.0,
      "step": 10220
    },
    {
      "entropy": 5.489952230453492,
      "epoch": 1.1483686190823834,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048791896467044335,
      "loss": 5.205,
      "mean_token_accuracy": 0.20840319842100144,
      "num_tokens": 22815558.0,
      "step": 10225
    },
    {
      "entropy": 5.416747665405273,
      "epoch": 1.1489301959903409,
      "grad_norm": 1.0,
      "learning_rate": 0.00048790598237354716,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20226485431194305,
      "num_tokens": 22825428.0,
      "step": 10230
    },
    {
      "entropy": 5.494300413131714,
      "epoch": 1.1494917728982985,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004878929932976419,
      "loss": 5.1723,
      "mean_token_accuracy": 0.2065751373767853,
      "num_tokens": 22836294.0,
      "step": 10235
    },
    {
      "entropy": 5.503645753860473,
      "epoch": 1.150053349806256,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004878799974431412,
      "loss": 5.2625,
      "mean_token_accuracy": 0.20460931211709976,
      "num_tokens": 22847835.0,
      "step": 10240
    },
    {
      "entropy": 5.4766439437866214,
      "epoch": 1.1506149267142134,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004878669948104588,
      "loss": 5.2098,
      "mean_token_accuracy": 0.20376038998365403,
      "num_tokens": 22859726.0,
      "step": 10245
    },
    {
      "entropy": 5.381437492370606,
      "epoch": 1.151176503622171,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048785398540000867,
      "loss": 5.1593,
      "mean_token_accuracy": 0.2043628767132759,
      "num_tokens": 22871250.0,
      "step": 10250
    },
    {
      "entropy": 5.486570644378662,
      "epoch": 1.1517380805301287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048784096921220514,
      "loss": 5.2188,
      "mean_token_accuracy": 0.19860162287950517,
      "num_tokens": 22882034.0,
      "step": 10255
    },
    {
      "entropy": 5.564865493774414,
      "epoch": 1.1522996574380862,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004878279462474627,
      "loss": 5.3111,
      "mean_token_accuracy": 0.19359066486358642,
      "num_tokens": 22894606.0,
      "step": 10260
    },
    {
      "entropy": 5.424198007583618,
      "epoch": 1.1528612343460436,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048781491650619574,
      "loss": 5.1371,
      "mean_token_accuracy": 0.20448300540447234,
      "num_tokens": 22905246.0,
      "step": 10265
    },
    {
      "entropy": 5.418663549423218,
      "epoch": 1.1534228112540013,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048780187998881944,
      "loss": 5.1001,
      "mean_token_accuracy": 0.19922555088996888,
      "num_tokens": 22916202.0,
      "step": 10270
    },
    {
      "entropy": 5.474135875701904,
      "epoch": 1.1539843881619587,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004877888366957486,
      "loss": 5.0901,
      "mean_token_accuracy": 0.20679617822170257,
      "num_tokens": 22926451.0,
      "step": 10275
    },
    {
      "entropy": 5.475333261489868,
      "epoch": 1.1545459650699164,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004877757866273988,
      "loss": 5.1615,
      "mean_token_accuracy": 0.20335488319396972,
      "num_tokens": 22936618.0,
      "step": 10280
    },
    {
      "entropy": 5.454792308807373,
      "epoch": 1.1551075419778738,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004877627297841853,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1988489806652069,
      "num_tokens": 22948278.0,
      "step": 10285
    },
    {
      "entropy": 5.392830848693848,
      "epoch": 1.1556691188858315,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004877496661665241,
      "loss": 5.1031,
      "mean_token_accuracy": 0.2060924395918846,
      "num_tokens": 22959928.0,
      "step": 10290
    },
    {
      "entropy": 5.403790426254273,
      "epoch": 1.156230695793789,
      "grad_norm": 0.875,
      "learning_rate": 0.000487736595774831,
      "loss": 5.1643,
      "mean_token_accuracy": 0.19947277158498763,
      "num_tokens": 22971050.0,
      "step": 10295
    },
    {
      "entropy": 5.35574369430542,
      "epoch": 1.1567922727017466,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048772351860952215,
      "loss": 5.0878,
      "mean_token_accuracy": 0.2012922391295433,
      "num_tokens": 22981471.0,
      "step": 10300
    },
    {
      "entropy": 5.515440988540649,
      "epoch": 1.157353849609704,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004877104346710139,
      "loss": 5.3299,
      "mean_token_accuracy": 0.2002103105187416,
      "num_tokens": 22992526.0,
      "step": 10305
    },
    {
      "entropy": 5.466501665115357,
      "epoch": 1.1579154265176617,
      "grad_norm": 0.890625,
      "learning_rate": 0.000487697343959723,
      "loss": 5.1834,
      "mean_token_accuracy": 0.2042958766222,
      "num_tokens": 23003408.0,
      "step": 10310
    },
    {
      "entropy": 5.4243903160095215,
      "epoch": 1.158477003425619,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004876842464760661,
      "loss": 5.0855,
      "mean_token_accuracy": 0.206155401468277,
      "num_tokens": 23013406.0,
      "step": 10315
    },
    {
      "entropy": 5.3888379573822025,
      "epoch": 1.1590385803335768,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048767114222046036,
      "loss": 5.1601,
      "mean_token_accuracy": 0.20603561401367188,
      "num_tokens": 23024960.0,
      "step": 10320
    },
    {
      "entropy": 5.507277536392212,
      "epoch": 1.1596001572415342,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000487658031193323,
      "loss": 5.1695,
      "mean_token_accuracy": 0.20135279297828673,
      "num_tokens": 23037514.0,
      "step": 10325
    },
    {
      "entropy": 5.502745771408081,
      "epoch": 1.1601617341494919,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004876449133950714,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19563647657632827,
      "num_tokens": 23048460.0,
      "step": 10330
    },
    {
      "entropy": 5.487990093231201,
      "epoch": 1.1607233110574493,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004876317888261233,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20132656991481782,
      "num_tokens": 23058866.0,
      "step": 10335
    },
    {
      "entropy": 5.396505355834961,
      "epoch": 1.1612848879654067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048761865748689657,
      "loss": 5.1956,
      "mean_token_accuracy": 0.1960831567645073,
      "num_tokens": 23069097.0,
      "step": 10340
    },
    {
      "entropy": 5.519156169891358,
      "epoch": 1.1618464648733644,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048760551937780927,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20661058872938157,
      "num_tokens": 23079553.0,
      "step": 10345
    },
    {
      "entropy": 5.427708721160888,
      "epoch": 1.162408041781322,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048759237449927986,
      "loss": 5.164,
      "mean_token_accuracy": 0.2044447660446167,
      "num_tokens": 23090554.0,
      "step": 10350
    },
    {
      "entropy": 5.380775690078735,
      "epoch": 1.1629696186892795,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048757922285172674,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20737981349229812,
      "num_tokens": 23101740.0,
      "step": 10355
    },
    {
      "entropy": 5.528803014755249,
      "epoch": 1.163531195597237,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048756606443556874,
      "loss": 5.1714,
      "mean_token_accuracy": 0.2024669960141182,
      "num_tokens": 23112842.0,
      "step": 10360
    },
    {
      "entropy": 5.425769233703614,
      "epoch": 1.1640927725051946,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048755289925122474,
      "loss": 5.1718,
      "mean_token_accuracy": 0.20082434713840486,
      "num_tokens": 23123822.0,
      "step": 10365
    },
    {
      "entropy": 5.48276653289795,
      "epoch": 1.164654349413152,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004875397272991141,
      "loss": 5.1842,
      "mean_token_accuracy": 0.19915650337934493,
      "num_tokens": 23134045.0,
      "step": 10370
    },
    {
      "entropy": 5.461555242538452,
      "epoch": 1.1652159263211097,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000487526548579656,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1933575764298439,
      "num_tokens": 23144200.0,
      "step": 10375
    },
    {
      "entropy": 5.4488602638244625,
      "epoch": 1.1657775032290671,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048751336309327026,
      "loss": 5.1307,
      "mean_token_accuracy": 0.2122505247592926,
      "num_tokens": 23155671.0,
      "step": 10380
    },
    {
      "entropy": 5.553076314926147,
      "epoch": 1.1663390801370248,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004875001708403766,
      "loss": 5.2824,
      "mean_token_accuracy": 0.202136991918087,
      "num_tokens": 23166644.0,
      "step": 10385
    },
    {
      "entropy": 5.429000329971314,
      "epoch": 1.1669006570449822,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048748697182139503,
      "loss": 5.1709,
      "mean_token_accuracy": 0.19815527349710466,
      "num_tokens": 23177074.0,
      "step": 10390
    },
    {
      "entropy": 5.527188014984131,
      "epoch": 1.16746223395294,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004874737660367459,
      "loss": 5.2801,
      "mean_token_accuracy": 0.19702013581991196,
      "num_tokens": 23188724.0,
      "step": 10395
    },
    {
      "entropy": 5.453551006317139,
      "epoch": 1.1680238108608973,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048746055348684967,
      "loss": 5.2035,
      "mean_token_accuracy": 0.19932918548583983,
      "num_tokens": 23198459.0,
      "step": 10400
    },
    {
      "entropy": 5.491748809814453,
      "epoch": 1.168585387768855,
      "grad_norm": 0.890625,
      "learning_rate": 0.000487447334172127,
      "loss": 5.2002,
      "mean_token_accuracy": 0.20066653937101364,
      "num_tokens": 23208991.0,
      "step": 10405
    },
    {
      "entropy": 5.4116840839385985,
      "epoch": 1.1691469646768125,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048743410809299877,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20399949699640274,
      "num_tokens": 23219309.0,
      "step": 10410
    },
    {
      "entropy": 5.481403589248657,
      "epoch": 1.1697085415847701,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004874208752498862,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19908248484134675,
      "num_tokens": 23230551.0,
      "step": 10415
    },
    {
      "entropy": 5.541622686386108,
      "epoch": 1.1702701184927276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004874076356432106,
      "loss": 5.299,
      "mean_token_accuracy": 0.19814835637807846,
      "num_tokens": 23242799.0,
      "step": 10420
    },
    {
      "entropy": 5.526991558074951,
      "epoch": 1.1708316954006852,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048739438927339346,
      "loss": 5.2995,
      "mean_token_accuracy": 0.19496002346277236,
      "num_tokens": 23255112.0,
      "step": 10425
    },
    {
      "entropy": 5.430134677886963,
      "epoch": 1.1713932723086427,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048738113614085656,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20294931679964065,
      "num_tokens": 23264800.0,
      "step": 10430
    },
    {
      "entropy": 5.430749320983887,
      "epoch": 1.1719548492166,
      "grad_norm": 0.859375,
      "learning_rate": 0.000487367876246022,
      "loss": 5.1214,
      "mean_token_accuracy": 0.2002952590584755,
      "num_tokens": 23276672.0,
      "step": 10435
    },
    {
      "entropy": 5.4335911750793455,
      "epoch": 1.1725164261245578,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004873546095893118,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20360282361507415,
      "num_tokens": 23288282.0,
      "step": 10440
    },
    {
      "entropy": 5.475942707061767,
      "epoch": 1.1730780030325154,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004873413361711485,
      "loss": 5.1875,
      "mean_token_accuracy": 0.19931413233280182,
      "num_tokens": 23299078.0,
      "step": 10445
    },
    {
      "entropy": 5.4522277355194095,
      "epoch": 1.1736395799404729,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048732805599195465,
      "loss": 5.1803,
      "mean_token_accuracy": 0.20039283484220505,
      "num_tokens": 23310036.0,
      "step": 10450
    },
    {
      "entropy": 5.439061450958252,
      "epoch": 1.1742011568484303,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048731476905215326,
      "loss": 5.1034,
      "mean_token_accuracy": 0.21084504574537277,
      "num_tokens": 23321029.0,
      "step": 10455
    },
    {
      "entropy": 5.468477344512939,
      "epoch": 1.174762733756388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004873014753521673,
      "loss": 5.2109,
      "mean_token_accuracy": 0.19901111871004104,
      "num_tokens": 23331584.0,
      "step": 10460
    },
    {
      "entropy": 5.5725603103637695,
      "epoch": 1.1753243106643454,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004872881748924199,
      "loss": 5.274,
      "mean_token_accuracy": 0.1877749890089035,
      "num_tokens": 23343171.0,
      "step": 10465
    },
    {
      "entropy": 5.559176826477051,
      "epoch": 1.175885887572303,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004872748676733347,
      "loss": 5.1924,
      "mean_token_accuracy": 0.20036645382642745,
      "num_tokens": 23353047.0,
      "step": 10470
    },
    {
      "entropy": 5.48087739944458,
      "epoch": 1.1764474644802605,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048726155369533547,
      "loss": 5.1956,
      "mean_token_accuracy": 0.19847763925790787,
      "num_tokens": 23364447.0,
      "step": 10475
    },
    {
      "entropy": 5.4802289485931395,
      "epoch": 1.1770090413882182,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487248232958846,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1980299934744835,
      "num_tokens": 23375616.0,
      "step": 10480
    },
    {
      "entropy": 5.5473105907440186,
      "epoch": 1.1775706182961756,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004872349054642904,
      "loss": 5.3356,
      "mean_token_accuracy": 0.19207555204629898,
      "num_tokens": 23388176.0,
      "step": 10485
    },
    {
      "entropy": 5.543276929855347,
      "epoch": 1.1781321952041333,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004872215712120932,
      "loss": 5.2672,
      "mean_token_accuracy": 0.2016679286956787,
      "num_tokens": 23400563.0,
      "step": 10490
    },
    {
      "entropy": 5.5292072772979735,
      "epoch": 1.1786937721120907,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004872082302026788,
      "loss": 5.2022,
      "mean_token_accuracy": 0.19479056894779206,
      "num_tokens": 23410740.0,
      "step": 10495
    },
    {
      "entropy": 5.482979726791382,
      "epoch": 1.1792553490200484,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004871948824364721,
      "loss": 5.2787,
      "mean_token_accuracy": 0.19374168217182158,
      "num_tokens": 23423086.0,
      "step": 10500
    },
    {
      "entropy": 5.52589602470398,
      "epoch": 1.1798169259280058,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000487181527913898,
      "loss": 5.2759,
      "mean_token_accuracy": 0.20016781687736512,
      "num_tokens": 23435869.0,
      "step": 10505
    },
    {
      "entropy": 5.4544442653656,
      "epoch": 1.1803785028359635,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048716816663538177,
      "loss": 5.1453,
      "mean_token_accuracy": 0.21112471967935562,
      "num_tokens": 23446635.0,
      "step": 10510
    },
    {
      "entropy": 5.41698145866394,
      "epoch": 1.180940079743921,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004871547986013488,
      "loss": 5.1737,
      "mean_token_accuracy": 0.20363107323646545,
      "num_tokens": 23458214.0,
      "step": 10515
    },
    {
      "entropy": 5.498543357849121,
      "epoch": 1.1815016566518786,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004871414238122248,
      "loss": 5.1998,
      "mean_token_accuracy": 0.20429729670286179,
      "num_tokens": 23469383.0,
      "step": 10520
    },
    {
      "entropy": 5.4363365650177,
      "epoch": 1.182063233559836,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004871280422684355,
      "loss": 5.1624,
      "mean_token_accuracy": 0.20632849037647247,
      "num_tokens": 23480841.0,
      "step": 10525
    },
    {
      "entropy": 5.401475667953491,
      "epoch": 1.1826248104677934,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048711465397040726,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20598978996276857,
      "num_tokens": 23490796.0,
      "step": 10530
    },
    {
      "entropy": 5.431342697143554,
      "epoch": 1.183186387375751,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048710125891856604,
      "loss": 5.1337,
      "mean_token_accuracy": 0.2041391983628273,
      "num_tokens": 23501759.0,
      "step": 10535
    },
    {
      "entropy": 5.466823291778565,
      "epoch": 1.1837479642837088,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048708785711333845,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20600284337997438,
      "num_tokens": 23512209.0,
      "step": 10540
    },
    {
      "entropy": 5.526998853683471,
      "epoch": 1.1843095411916662,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048707444855515124,
      "loss": 5.2572,
      "mean_token_accuracy": 0.20000497847795487,
      "num_tokens": 23524316.0,
      "step": 10545
    },
    {
      "entropy": 5.462821006774902,
      "epoch": 1.1848711180996236,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004870610332444314,
      "loss": 5.2265,
      "mean_token_accuracy": 0.20672667920589446,
      "num_tokens": 23536199.0,
      "step": 10550
    },
    {
      "entropy": 5.4836501121521,
      "epoch": 1.1854326950075813,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004870476111816059,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19707887768745422,
      "num_tokens": 23548034.0,
      "step": 10555
    },
    {
      "entropy": 5.517248106002808,
      "epoch": 1.1859942719155387,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048703418236710234,
      "loss": 5.188,
      "mean_token_accuracy": 0.20197130590677262,
      "num_tokens": 23559280.0,
      "step": 10560
    },
    {
      "entropy": 5.46593918800354,
      "epoch": 1.1865558488234964,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004870207468013481,
      "loss": 5.1591,
      "mean_token_accuracy": 0.20195926278829573,
      "num_tokens": 23569163.0,
      "step": 10565
    },
    {
      "entropy": 5.484252405166626,
      "epoch": 1.1871174257314538,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004870073044847711,
      "loss": 5.2334,
      "mean_token_accuracy": 0.2005589172244072,
      "num_tokens": 23580526.0,
      "step": 10570
    },
    {
      "entropy": 5.418462133407592,
      "epoch": 1.1876790026394115,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004869938554177993,
      "loss": 5.1224,
      "mean_token_accuracy": 0.20061395168304444,
      "num_tokens": 23590552.0,
      "step": 10575
    },
    {
      "entropy": 5.359176158905029,
      "epoch": 1.188240579547369,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048698039960086084,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21109791994094848,
      "num_tokens": 23601889.0,
      "step": 10580
    },
    {
      "entropy": 5.500014162063598,
      "epoch": 1.1888021564553266,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004869669370343843,
      "loss": 5.2873,
      "mean_token_accuracy": 0.20560032576322557,
      "num_tokens": 23613833.0,
      "step": 10585
    },
    {
      "entropy": 5.528821897506714,
      "epoch": 1.189363733363284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004869534677187983,
      "loss": 5.2405,
      "mean_token_accuracy": 0.2039456471800804,
      "num_tokens": 23625404.0,
      "step": 10590
    },
    {
      "entropy": 5.4273459911346436,
      "epoch": 1.1899253102712417,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004869399916545317,
      "loss": 5.1079,
      "mean_token_accuracy": 0.21329526454210282,
      "num_tokens": 23635670.0,
      "step": 10595
    },
    {
      "entropy": 5.422181940078735,
      "epoch": 1.1904868871791991,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004869265088420135,
      "loss": 5.1136,
      "mean_token_accuracy": 0.20950076431035997,
      "num_tokens": 23646893.0,
      "step": 10600
    },
    {
      "entropy": 5.382355546951294,
      "epoch": 1.1910484640871568,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004869130192816731,
      "loss": 5.2188,
      "mean_token_accuracy": 0.20114482641220094,
      "num_tokens": 23657733.0,
      "step": 10605
    },
    {
      "entropy": 5.361873626708984,
      "epoch": 1.1916100409951143,
      "grad_norm": 0.875,
      "learning_rate": 0.00048689952297394,
      "loss": 5.068,
      "mean_token_accuracy": 0.2130013331770897,
      "num_tokens": 23668100.0,
      "step": 10610
    },
    {
      "entropy": 5.543952322006225,
      "epoch": 1.192171617903072,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004868860199192438,
      "loss": 5.2399,
      "mean_token_accuracy": 0.19444207102060318,
      "num_tokens": 23678295.0,
      "step": 10615
    },
    {
      "entropy": 5.437207508087158,
      "epoch": 1.1927331948110294,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004868725101180146,
      "loss": 5.0724,
      "mean_token_accuracy": 0.20954830944538116,
      "num_tokens": 23689310.0,
      "step": 10620
    },
    {
      "entropy": 5.394340705871582,
      "epoch": 1.1932947717189868,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004868589935706826,
      "loss": 5.2243,
      "mean_token_accuracy": 0.19757401943206787,
      "num_tokens": 23699475.0,
      "step": 10625
    },
    {
      "entropy": 5.569411849975586,
      "epoch": 1.1938563486269445,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048684547027767793,
      "loss": 5.2217,
      "mean_token_accuracy": 0.2028184324502945,
      "num_tokens": 23710186.0,
      "step": 10630
    },
    {
      "entropy": 5.379452133178711,
      "epoch": 1.1944179255349021,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004868319402394314,
      "loss": 5.041,
      "mean_token_accuracy": 0.20852069705724716,
      "num_tokens": 23720449.0,
      "step": 10635
    },
    {
      "entropy": 5.522710561752319,
      "epoch": 1.1949795024428596,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048681840345637375,
      "loss": 5.2949,
      "mean_token_accuracy": 0.19844478964805604,
      "num_tokens": 23732295.0,
      "step": 10640
    },
    {
      "entropy": 5.436313629150391,
      "epoch": 1.195541079350817,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048680485992893597,
      "loss": 5.1985,
      "mean_token_accuracy": 0.20026141703128814,
      "num_tokens": 23743616.0,
      "step": 10645
    },
    {
      "entropy": 5.547460079193115,
      "epoch": 1.1961026562587747,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048679130965754926,
      "loss": 5.2263,
      "mean_token_accuracy": 0.1993878036737442,
      "num_tokens": 23755288.0,
      "step": 10650
    },
    {
      "entropy": 5.4789165496826175,
      "epoch": 1.196664233166732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004867777526426451,
      "loss": 5.1004,
      "mean_token_accuracy": 0.2067498356103897,
      "num_tokens": 23765916.0,
      "step": 10655
    },
    {
      "entropy": 5.32796459197998,
      "epoch": 1.1972258100746898,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048676418888465525,
      "loss": 5.0536,
      "mean_token_accuracy": 0.20764026194810867,
      "num_tokens": 23778891.0,
      "step": 10660
    },
    {
      "entropy": 5.469976377487183,
      "epoch": 1.1977873869826472,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004867506183840114,
      "loss": 5.2073,
      "mean_token_accuracy": 0.2060576245188713,
      "num_tokens": 23790402.0,
      "step": 10665
    },
    {
      "entropy": 5.490409898757934,
      "epoch": 1.1983489638906049,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004867370411411459,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20888323932886124,
      "num_tokens": 23801928.0,
      "step": 10670
    },
    {
      "entropy": 5.4994309425354,
      "epoch": 1.1989105407985623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048672345715649066,
      "loss": 5.2279,
      "mean_token_accuracy": 0.19480618238449096,
      "num_tokens": 23812302.0,
      "step": 10675
    },
    {
      "entropy": 5.4917809009552006,
      "epoch": 1.19947211770652,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004867098664304785,
      "loss": 5.1897,
      "mean_token_accuracy": 0.21261890828609467,
      "num_tokens": 23824033.0,
      "step": 10680
    },
    {
      "entropy": 5.448282527923584,
      "epoch": 1.2000336946144774,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004866962689635422,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20728238821029663,
      "num_tokens": 23834907.0,
      "step": 10685
    },
    {
      "entropy": 5.517509269714355,
      "epoch": 1.200595271522435,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048668266475611447,
      "loss": 5.3365,
      "mean_token_accuracy": 0.1988241881132126,
      "num_tokens": 23846418.0,
      "step": 10690
    },
    {
      "entropy": 5.59075403213501,
      "epoch": 1.2011568484303925,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004866690538086286,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19889292865991592,
      "num_tokens": 23858423.0,
      "step": 10695
    },
    {
      "entropy": 5.479468297958374,
      "epoch": 1.2017184253383502,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048665543612151803,
      "loss": 5.1608,
      "mean_token_accuracy": 0.20260724276304246,
      "num_tokens": 23868340.0,
      "step": 10700
    },
    {
      "entropy": 5.385057020187378,
      "epoch": 1.2022800022463076,
      "grad_norm": 1.0,
      "learning_rate": 0.00048664181169521625,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20785311758518218,
      "num_tokens": 23878752.0,
      "step": 10705
    },
    {
      "entropy": 5.496807289123535,
      "epoch": 1.2028415791542653,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004866281805301571,
      "loss": 5.1426,
      "mean_token_accuracy": 0.21058329939842224,
      "num_tokens": 23889374.0,
      "step": 10710
    },
    {
      "entropy": 5.563439416885376,
      "epoch": 1.2034031560622227,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004866145426267746,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1995503157377243,
      "num_tokens": 23901651.0,
      "step": 10715
    },
    {
      "entropy": 5.381554174423218,
      "epoch": 1.2039647329701804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000486600897985503,
      "loss": 5.1467,
      "mean_token_accuracy": 0.20176431089639663,
      "num_tokens": 23911818.0,
      "step": 10720
    },
    {
      "entropy": 5.4619310855865475,
      "epoch": 1.2045263098781378,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048658724660677674,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20634490698575975,
      "num_tokens": 23923840.0,
      "step": 10725
    },
    {
      "entropy": 5.409549140930176,
      "epoch": 1.2050878867860955,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004865735884910305,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2146542862057686,
      "num_tokens": 23935406.0,
      "step": 10730
    },
    {
      "entropy": 5.452820348739624,
      "epoch": 1.205649463694053,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004865599236386991,
      "loss": 5.1717,
      "mean_token_accuracy": 0.20672981441020966,
      "num_tokens": 23946350.0,
      "step": 10735
    },
    {
      "entropy": 5.372952604293824,
      "epoch": 1.2062110406020103,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004865462520502177,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20743412524461746,
      "num_tokens": 23956315.0,
      "step": 10740
    },
    {
      "entropy": 5.4654529094696045,
      "epoch": 1.206772617509968,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004865325737260216,
      "loss": 5.0573,
      "mean_token_accuracy": 0.20352966636419295,
      "num_tokens": 23966649.0,
      "step": 10745
    },
    {
      "entropy": 5.573237705230713,
      "epoch": 1.2073341944179254,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048651888866654624,
      "loss": 5.3107,
      "mean_token_accuracy": 0.1902357503771782,
      "num_tokens": 23977674.0,
      "step": 10750
    },
    {
      "entropy": 5.4229268550872805,
      "epoch": 1.207895771325883,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004865051968722275,
      "loss": 5.1958,
      "mean_token_accuracy": 0.20081054270267487,
      "num_tokens": 23987985.0,
      "step": 10755
    },
    {
      "entropy": 5.516906642913819,
      "epoch": 1.2084573482338405,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048649149834350126,
      "loss": 5.3134,
      "mean_token_accuracy": 0.1920422911643982,
      "num_tokens": 23998730.0,
      "step": 10760
    },
    {
      "entropy": 5.43920488357544,
      "epoch": 1.2090189251417982,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004864777930808037,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1964148297905922,
      "num_tokens": 24008743.0,
      "step": 10765
    },
    {
      "entropy": 5.443512105941773,
      "epoch": 1.2095805020497556,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004864640810845711,
      "loss": 5.1677,
      "mean_token_accuracy": 0.2087377980351448,
      "num_tokens": 24019805.0,
      "step": 10770
    },
    {
      "entropy": 5.486453628540039,
      "epoch": 1.2101420789577133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004864503623552402,
      "loss": 5.1634,
      "mean_token_accuracy": 0.20224840193986893,
      "num_tokens": 24031437.0,
      "step": 10775
    },
    {
      "entropy": 5.548153305053711,
      "epoch": 1.2107036558656707,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004864366368932478,
      "loss": 5.2921,
      "mean_token_accuracy": 0.1932049050927162,
      "num_tokens": 24043922.0,
      "step": 10780
    },
    {
      "entropy": 5.472371435165405,
      "epoch": 1.2112652327736284,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048642290469903083,
      "loss": 5.1144,
      "mean_token_accuracy": 0.2126319393515587,
      "num_tokens": 24054755.0,
      "step": 10785
    },
    {
      "entropy": 5.415264511108399,
      "epoch": 1.2118268096815858,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048640916577302664,
      "loss": 5.1249,
      "mean_token_accuracy": 0.19344184547662735,
      "num_tokens": 24067347.0,
      "step": 10790
    },
    {
      "entropy": 5.47366189956665,
      "epoch": 1.2123883865895435,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048639542011567243,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20277249217033386,
      "num_tokens": 24078938.0,
      "step": 10795
    },
    {
      "entropy": 5.45726752281189,
      "epoch": 1.212949963497501,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004863816677274062,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19966355562210084,
      "num_tokens": 24090275.0,
      "step": 10800
    },
    {
      "entropy": 5.412062549591065,
      "epoch": 1.2135115404054586,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048636790860866576,
      "loss": 5.1919,
      "mean_token_accuracy": 0.20711246728897095,
      "num_tokens": 24101518.0,
      "step": 10805
    },
    {
      "entropy": 5.506921720504761,
      "epoch": 1.214073117313416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004863541427598889,
      "loss": 5.295,
      "mean_token_accuracy": 0.19296726137399672,
      "num_tokens": 24113320.0,
      "step": 10810
    },
    {
      "entropy": 5.426114940643311,
      "epoch": 1.2146346942213737,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004863403701815144,
      "loss": 5.1204,
      "mean_token_accuracy": 0.20659309476614,
      "num_tokens": 24124842.0,
      "step": 10815
    },
    {
      "entropy": 5.5163712978363035,
      "epoch": 1.2151962711293312,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004863265908739805,
      "loss": 5.2611,
      "mean_token_accuracy": 0.20111954510211943,
      "num_tokens": 24136391.0,
      "step": 10820
    },
    {
      "entropy": 5.548802280426026,
      "epoch": 1.2157578480372888,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004863128048377259,
      "loss": 5.2166,
      "mean_token_accuracy": 0.20157045125961304,
      "num_tokens": 24147194.0,
      "step": 10825
    },
    {
      "entropy": 5.518996667861939,
      "epoch": 1.2163194249452463,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004862990120731896,
      "loss": 5.2752,
      "mean_token_accuracy": 0.19721537530422212,
      "num_tokens": 24158170.0,
      "step": 10830
    },
    {
      "entropy": 5.4756945133209225,
      "epoch": 1.2168810018532037,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048628521258081096,
      "loss": 5.1292,
      "mean_token_accuracy": 0.20673321634531022,
      "num_tokens": 24168671.0,
      "step": 10835
    },
    {
      "entropy": 5.464004802703857,
      "epoch": 1.2174425787611614,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048627140636102906,
      "loss": 5.1714,
      "mean_token_accuracy": 0.19875272810459138,
      "num_tokens": 24179898.0,
      "step": 10840
    },
    {
      "entropy": 5.399494647979736,
      "epoch": 1.2180041556691188,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004862575934142837,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20442479401826857,
      "num_tokens": 24190622.0,
      "step": 10845
    },
    {
      "entropy": 5.607158136367798,
      "epoch": 1.2185657325770765,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004862437737410147,
      "loss": 5.3196,
      "mean_token_accuracy": 0.19658107310533524,
      "num_tokens": 24202526.0,
      "step": 10850
    },
    {
      "entropy": 5.4285320281982425,
      "epoch": 1.219127309485034,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004862299473416619,
      "loss": 5.1456,
      "mean_token_accuracy": 0.20412747710943221,
      "num_tokens": 24213186.0,
      "step": 10855
    },
    {
      "entropy": 5.414621591567993,
      "epoch": 1.2196888863929916,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048621611421666573,
      "loss": 5.1718,
      "mean_token_accuracy": 0.20253748297691346,
      "num_tokens": 24225490.0,
      "step": 10860
    },
    {
      "entropy": 5.472359466552734,
      "epoch": 1.220250463300949,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004862022743664665,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20252657383680345,
      "num_tokens": 24236653.0,
      "step": 10865
    },
    {
      "entropy": 5.406805372238159,
      "epoch": 1.2208120402089067,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048618842779150494,
      "loss": 5.1188,
      "mean_token_accuracy": 0.21016918867826462,
      "num_tokens": 24247843.0,
      "step": 10870
    },
    {
      "entropy": 5.416406679153442,
      "epoch": 1.221373617116864,
      "grad_norm": 0.875,
      "learning_rate": 0.00048617457449222205,
      "loss": 5.1972,
      "mean_token_accuracy": 0.20435844659805297,
      "num_tokens": 24258332.0,
      "step": 10875
    },
    {
      "entropy": 5.469857454299927,
      "epoch": 1.2219351940248218,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004861607144690587,
      "loss": 5.1766,
      "mean_token_accuracy": 0.20377371162176133,
      "num_tokens": 24268685.0,
      "step": 10880
    },
    {
      "entropy": 5.480464601516724,
      "epoch": 1.2224967709327792,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004861468477224563,
      "loss": 5.2216,
      "mean_token_accuracy": 0.1970214933156967,
      "num_tokens": 24280660.0,
      "step": 10885
    },
    {
      "entropy": 5.505575656890869,
      "epoch": 1.2230583478407369,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048613297425285635,
      "loss": 5.1466,
      "mean_token_accuracy": 0.2031639412045479,
      "num_tokens": 24291649.0,
      "step": 10890
    },
    {
      "entropy": 5.464556360244751,
      "epoch": 1.2236199247486943,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004861190940607007,
      "loss": 5.1683,
      "mean_token_accuracy": 0.201264625787735,
      "num_tokens": 24303108.0,
      "step": 10895
    },
    {
      "entropy": 5.432361030578614,
      "epoch": 1.224181501656652,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004861052071464312,
      "loss": 5.1958,
      "mean_token_accuracy": 0.19480210691690444,
      "num_tokens": 24315939.0,
      "step": 10900
    },
    {
      "entropy": 5.437289237976074,
      "epoch": 1.2247430785646094,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048609131351048994,
      "loss": 5.172,
      "mean_token_accuracy": 0.20418543219566346,
      "num_tokens": 24326916.0,
      "step": 10905
    },
    {
      "entropy": 5.464480352401734,
      "epoch": 1.225304655472567,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004860774131533195,
      "loss": 5.2156,
      "mean_token_accuracy": 0.2050207659602165,
      "num_tokens": 24337790.0,
      "step": 10910
    },
    {
      "entropy": 5.4242981433868405,
      "epoch": 1.2258662323805245,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048606350607536235,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21056256145238877,
      "num_tokens": 24347930.0,
      "step": 10915
    },
    {
      "entropy": 5.456901550292969,
      "epoch": 1.2264278092884822,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004860495922770612,
      "loss": 5.1384,
      "mean_token_accuracy": 0.20334243923425674,
      "num_tokens": 24359213.0,
      "step": 10920
    },
    {
      "entropy": 5.4033571720123295,
      "epoch": 1.2269893861964396,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048603567175885926,
      "loss": 5.0931,
      "mean_token_accuracy": 0.20321608781814576,
      "num_tokens": 24371000.0,
      "step": 10925
    },
    {
      "entropy": 5.375303888320923,
      "epoch": 1.227550963104397,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004860217445211997,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20741177648305892,
      "num_tokens": 24381301.0,
      "step": 10930
    },
    {
      "entropy": 5.5339268207550045,
      "epoch": 1.2281125400123547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004860078105645259,
      "loss": 5.2733,
      "mean_token_accuracy": 0.191807122528553,
      "num_tokens": 24391808.0,
      "step": 10935
    },
    {
      "entropy": 5.547538089752197,
      "epoch": 1.2286741169203121,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004859938698892816,
      "loss": 5.4052,
      "mean_token_accuracy": 0.19819806516170502,
      "num_tokens": 24404220.0,
      "step": 10940
    },
    {
      "entropy": 5.517003870010376,
      "epoch": 1.2292356938282698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048597992249591065,
      "loss": 5.1978,
      "mean_token_accuracy": 0.203224977850914,
      "num_tokens": 24414906.0,
      "step": 10945
    },
    {
      "entropy": 5.440996313095093,
      "epoch": 1.2297972707362272,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048596596838485704,
      "loss": 5.1872,
      "mean_token_accuracy": 0.19764235615730286,
      "num_tokens": 24426469.0,
      "step": 10950
    },
    {
      "entropy": 5.34379768371582,
      "epoch": 1.230358847644185,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004859520075565654,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20072033256292343,
      "num_tokens": 24437683.0,
      "step": 10955
    },
    {
      "entropy": 5.355475902557373,
      "epoch": 1.2309204245521423,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048593804001147986,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20771514028310775,
      "num_tokens": 24449038.0,
      "step": 10960
    },
    {
      "entropy": 5.527215957641602,
      "epoch": 1.2314820014601,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004859240657500453,
      "loss": 5.1871,
      "mean_token_accuracy": 0.20605241358280182,
      "num_tokens": 24460228.0,
      "step": 10965
    },
    {
      "entropy": 5.4582466125488285,
      "epoch": 1.2320435783680574,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048591008477270675,
      "loss": 5.2425,
      "mean_token_accuracy": 0.20028189420700074,
      "num_tokens": 24471222.0,
      "step": 10970
    },
    {
      "entropy": 5.398322248458863,
      "epoch": 1.232605155276015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048589609707990925,
      "loss": 5.1741,
      "mean_token_accuracy": 0.19482695907354355,
      "num_tokens": 24482127.0,
      "step": 10975
    },
    {
      "entropy": 5.483496284484863,
      "epoch": 1.2331667321839725,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048588210267209823,
      "loss": 5.1482,
      "mean_token_accuracy": 0.1991759181022644,
      "num_tokens": 24492963.0,
      "step": 10980
    },
    {
      "entropy": 5.550104093551636,
      "epoch": 1.2337283090919302,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048586810154971935,
      "loss": 5.1802,
      "mean_token_accuracy": 0.2066708415746689,
      "num_tokens": 24502885.0,
      "step": 10985
    },
    {
      "entropy": 5.5322870254516605,
      "epoch": 1.2342898859998876,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048585409371321826,
      "loss": 5.2664,
      "mean_token_accuracy": 0.20242838710546493,
      "num_tokens": 24514990.0,
      "step": 10990
    },
    {
      "entropy": 5.49631929397583,
      "epoch": 1.2348514629078453,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048584007916304106,
      "loss": 5.1829,
      "mean_token_accuracy": 0.19860262125730516,
      "num_tokens": 24525555.0,
      "step": 10995
    },
    {
      "entropy": 5.542101716995239,
      "epoch": 1.2354130398158027,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048582605789963395,
      "loss": 5.229,
      "mean_token_accuracy": 0.19425695687532424,
      "num_tokens": 24536350.0,
      "step": 11000
    },
    {
      "entropy": 5.568210029602051,
      "epoch": 1.2359746167237604,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004858120299234434,
      "loss": 5.2355,
      "mean_token_accuracy": 0.19715877771377563,
      "num_tokens": 24547756.0,
      "step": 11005
    },
    {
      "entropy": 5.382920217514038,
      "epoch": 1.2365361936317179,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00048579799523491607,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20450157076120376,
      "num_tokens": 24559897.0,
      "step": 11010
    },
    {
      "entropy": 5.426002359390258,
      "epoch": 1.2370977705396755,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004857839538344988,
      "loss": 5.1456,
      "mean_token_accuracy": 0.20201408118009567,
      "num_tokens": 24570168.0,
      "step": 11015
    },
    {
      "entropy": 5.467796277999878,
      "epoch": 1.237659347447633,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048576990572263875,
      "loss": 5.0934,
      "mean_token_accuracy": 0.2145301192998886,
      "num_tokens": 24580918.0,
      "step": 11020
    },
    {
      "entropy": 5.4157970428466795,
      "epoch": 1.2382209243555904,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004857558508997831,
      "loss": 5.1342,
      "mean_token_accuracy": 0.1992151290178299,
      "num_tokens": 24592378.0,
      "step": 11025
    },
    {
      "entropy": 5.55806074142456,
      "epoch": 1.238782501263548,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004857417893663794,
      "loss": 5.2255,
      "mean_token_accuracy": 0.2024653360247612,
      "num_tokens": 24603666.0,
      "step": 11030
    },
    {
      "entropy": 5.452266359329224,
      "epoch": 1.2393440781715055,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004857277211228754,
      "loss": 5.189,
      "mean_token_accuracy": 0.20117369592189788,
      "num_tokens": 24615299.0,
      "step": 11035
    },
    {
      "entropy": 5.435538482666016,
      "epoch": 1.2399056550794632,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004857136461697191,
      "loss": 5.1443,
      "mean_token_accuracy": 0.2065530389547348,
      "num_tokens": 24626138.0,
      "step": 11040
    },
    {
      "entropy": 5.3945800304412845,
      "epoch": 1.2404672319874206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048569956450735853,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20395744144916533,
      "num_tokens": 24636720.0,
      "step": 11045
    },
    {
      "entropy": 5.486225175857544,
      "epoch": 1.2410288088953783,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048568547613624206,
      "loss": 5.1512,
      "mean_token_accuracy": 0.19981513917446136,
      "num_tokens": 24648147.0,
      "step": 11050
    },
    {
      "entropy": 5.487588167190552,
      "epoch": 1.2415903858033357,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048567138105681833,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20424194931983947,
      "num_tokens": 24659825.0,
      "step": 11055
    },
    {
      "entropy": 5.433658409118652,
      "epoch": 1.2421519627112934,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048565727926953617,
      "loss": 5.1537,
      "mean_token_accuracy": 0.20468665510416031,
      "num_tokens": 24671583.0,
      "step": 11060
    },
    {
      "entropy": 5.405447626113892,
      "epoch": 1.2427135396192508,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004856431707748445,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2054637387394905,
      "num_tokens": 24682034.0,
      "step": 11065
    },
    {
      "entropy": 5.4570588111877445,
      "epoch": 1.2432751165272085,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004856290555731926,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20111593455076218,
      "num_tokens": 24693624.0,
      "step": 11070
    },
    {
      "entropy": 5.522160196304322,
      "epoch": 1.243836693435166,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004856149336650299,
      "loss": 5.1827,
      "mean_token_accuracy": 0.20202843695878983,
      "num_tokens": 24705340.0,
      "step": 11075
    },
    {
      "entropy": 5.522665119171142,
      "epoch": 1.2443982703431236,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004856008050508059,
      "loss": 5.2435,
      "mean_token_accuracy": 0.198203207552433,
      "num_tokens": 24717115.0,
      "step": 11080
    },
    {
      "entropy": 5.455497360229492,
      "epoch": 1.244959847251081,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004855866697309707,
      "loss": 5.0737,
      "mean_token_accuracy": 0.20833736211061477,
      "num_tokens": 24727768.0,
      "step": 11085
    },
    {
      "entropy": 5.496109104156494,
      "epoch": 1.2455214241590387,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004855725277059742,
      "loss": 5.2545,
      "mean_token_accuracy": 0.1944260910153389,
      "num_tokens": 24740054.0,
      "step": 11090
    },
    {
      "entropy": 5.526373100280762,
      "epoch": 1.246083001066996,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004855583789762668,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20010430216789246,
      "num_tokens": 24750658.0,
      "step": 11095
    },
    {
      "entropy": 5.564990377426147,
      "epoch": 1.2466445779749538,
      "grad_norm": 0.984375,
      "learning_rate": 0.000485544223542299,
      "loss": 5.3215,
      "mean_token_accuracy": 0.19970221519470216,
      "num_tokens": 24762730.0,
      "step": 11100
    },
    {
      "entropy": 5.395041131973267,
      "epoch": 1.2472061548829112,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004855300614045214,
      "loss": 5.0793,
      "mean_token_accuracy": 0.2079702153801918,
      "num_tokens": 24773865.0,
      "step": 11105
    },
    {
      "entropy": 5.4118711948394775,
      "epoch": 1.2477677317908689,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048551589256338497,
      "loss": 5.1863,
      "mean_token_accuracy": 0.20108642876148225,
      "num_tokens": 24783030.0,
      "step": 11110
    },
    {
      "entropy": 5.443980979919433,
      "epoch": 1.2483293086988263,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048550171701934087,
      "loss": 5.103,
      "mean_token_accuracy": 0.20783449411392213,
      "num_tokens": 24793177.0,
      "step": 11115
    },
    {
      "entropy": 5.536599254608154,
      "epoch": 1.2488908856067837,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048548753477284053,
      "loss": 5.2105,
      "mean_token_accuracy": 0.20435337275266646,
      "num_tokens": 24804796.0,
      "step": 11120
    },
    {
      "entropy": 5.532714796066284,
      "epoch": 1.2494524625147414,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004854733458243354,
      "loss": 5.2097,
      "mean_token_accuracy": 0.20136349648237228,
      "num_tokens": 24816806.0,
      "step": 11125
    },
    {
      "entropy": 5.456060934066772,
      "epoch": 1.250014039422699,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004854591501742773,
      "loss": 5.161,
      "mean_token_accuracy": 0.20636742562055588,
      "num_tokens": 24826888.0,
      "step": 11130
    },
    {
      "entropy": 5.468593978881836,
      "epoch": 1.2505756163306565,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048544494782311825,
      "loss": 5.1188,
      "mean_token_accuracy": 0.20324206501245498,
      "num_tokens": 24837731.0,
      "step": 11135
    },
    {
      "entropy": 5.426183652877808,
      "epoch": 1.251137193238614,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048543073877131037,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20506934374570845,
      "num_tokens": 24847900.0,
      "step": 11140
    },
    {
      "entropy": 5.436147308349609,
      "epoch": 1.2516987701465716,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048541652301930627,
      "loss": 5.2296,
      "mean_token_accuracy": 0.19481280148029329,
      "num_tokens": 24858878.0,
      "step": 11145
    },
    {
      "entropy": 5.443544912338257,
      "epoch": 1.252260347054529,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004854023005675585,
      "loss": 5.1642,
      "mean_token_accuracy": 0.20288995802402496,
      "num_tokens": 24870051.0,
      "step": 11150
    },
    {
      "entropy": 5.446250152587891,
      "epoch": 1.2528219239624867,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004853880714165197,
      "loss": 5.2615,
      "mean_token_accuracy": 0.1973303362727165,
      "num_tokens": 24881671.0,
      "step": 11155
    },
    {
      "entropy": 5.491633129119873,
      "epoch": 1.2533835008704441,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048537383556664315,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20208802074193954,
      "num_tokens": 24892944.0,
      "step": 11160
    },
    {
      "entropy": 5.46076545715332,
      "epoch": 1.2539450777784018,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004853595930183822,
      "loss": 5.1902,
      "mean_token_accuracy": 0.20062874257564545,
      "num_tokens": 24903346.0,
      "step": 11165
    },
    {
      "entropy": 5.431779241561889,
      "epoch": 1.2545066546863592,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004853453437721901,
      "loss": 5.1845,
      "mean_token_accuracy": 0.20402949750423433,
      "num_tokens": 24913904.0,
      "step": 11170
    },
    {
      "entropy": 5.443333768844605,
      "epoch": 1.255068231594317,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004853310878285208,
      "loss": 5.1325,
      "mean_token_accuracy": 0.20814678817987442,
      "num_tokens": 24924635.0,
      "step": 11175
    },
    {
      "entropy": 5.53477258682251,
      "epoch": 1.2556298085022743,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048531682518782797,
      "loss": 5.3233,
      "mean_token_accuracy": 0.19479757696390151,
      "num_tokens": 24935466.0,
      "step": 11180
    },
    {
      "entropy": 5.540110635757446,
      "epoch": 1.256191385410232,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048530255585056594,
      "loss": 5.2816,
      "mean_token_accuracy": 0.19894088357686995,
      "num_tokens": 24948348.0,
      "step": 11185
    },
    {
      "entropy": 5.453719186782837,
      "epoch": 1.2567529623181894,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004852882798171889,
      "loss": 5.1108,
      "mean_token_accuracy": 0.2054199367761612,
      "num_tokens": 24959242.0,
      "step": 11190
    },
    {
      "entropy": 5.401065111160278,
      "epoch": 1.2573145392261469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004852739970881515,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20700909495353698,
      "num_tokens": 24969855.0,
      "step": 11195
    },
    {
      "entropy": 5.419329643249512,
      "epoch": 1.2578761161341045,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048525970766390847,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20535863935947418,
      "num_tokens": 24980378.0,
      "step": 11200
    },
    {
      "entropy": 5.498160409927368,
      "epoch": 1.2584376930420622,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004852454115449148,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19778600484132766,
      "num_tokens": 24990646.0,
      "step": 11205
    },
    {
      "entropy": 5.592801570892334,
      "epoch": 1.2589992699500197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004852311087316257,
      "loss": 5.2931,
      "mean_token_accuracy": 0.19426003843545914,
      "num_tokens": 25002644.0,
      "step": 11210
    },
    {
      "entropy": 5.5401918411254885,
      "epoch": 1.259560846857977,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004852167992244966,
      "loss": 5.199,
      "mean_token_accuracy": 0.20457497835159302,
      "num_tokens": 25012635.0,
      "step": 11215
    },
    {
      "entropy": 5.463432455062867,
      "epoch": 1.2601224237659348,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000485202483023983,
      "loss": 5.314,
      "mean_token_accuracy": 0.19231504648923875,
      "num_tokens": 25024289.0,
      "step": 11220
    },
    {
      "entropy": 5.45431547164917,
      "epoch": 1.2606840006738924,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048518816013054083,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20552338659763336,
      "num_tokens": 25035316.0,
      "step": 11225
    },
    {
      "entropy": 5.462117767333984,
      "epoch": 1.2612455775818499,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048517383054462617,
      "loss": 5.1318,
      "mean_token_accuracy": 0.20711493492126465,
      "num_tokens": 25046011.0,
      "step": 11230
    },
    {
      "entropy": 5.504916143417359,
      "epoch": 1.2618071544898073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004851594942666952,
      "loss": 5.1261,
      "mean_token_accuracy": 0.20258721709251404,
      "num_tokens": 25057402.0,
      "step": 11235
    },
    {
      "entropy": 5.396705627441406,
      "epoch": 1.262368731397765,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048514515129720445,
      "loss": 5.143,
      "mean_token_accuracy": 0.20553721934556962,
      "num_tokens": 25068287.0,
      "step": 11240
    },
    {
      "entropy": 5.369332790374756,
      "epoch": 1.2629303083057224,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004851308016366105,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20623514652252198,
      "num_tokens": 25079534.0,
      "step": 11245
    },
    {
      "entropy": 5.372199773788452,
      "epoch": 1.26349188521368,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048511644528537043,
      "loss": 5.0711,
      "mean_token_accuracy": 0.205873604118824,
      "num_tokens": 25090825.0,
      "step": 11250
    },
    {
      "entropy": 5.48745584487915,
      "epoch": 1.2640534621216375,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004851020822439412,
      "loss": 5.179,
      "mean_token_accuracy": 0.20095258951187134,
      "num_tokens": 25101859.0,
      "step": 11255
    },
    {
      "entropy": 5.397821140289307,
      "epoch": 1.2646150390295952,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048508771251278015,
      "loss": 5.1107,
      "mean_token_accuracy": 0.2097848579287529,
      "num_tokens": 25111416.0,
      "step": 11260
    },
    {
      "entropy": 5.455532360076904,
      "epoch": 1.2651766159375526,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850733360923449,
      "loss": 5.2124,
      "mean_token_accuracy": 0.19858232140541077,
      "num_tokens": 25122129.0,
      "step": 11265
    },
    {
      "entropy": 5.469708251953125,
      "epoch": 1.2657381928455103,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004850589529830931,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20311749577522278,
      "num_tokens": 25132429.0,
      "step": 11270
    },
    {
      "entropy": 5.446085739135742,
      "epoch": 1.2662997697534677,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048504456318548285,
      "loss": 5.2332,
      "mean_token_accuracy": 0.20629845261573793,
      "num_tokens": 25144242.0,
      "step": 11275
    },
    {
      "entropy": 5.529765605926514,
      "epoch": 1.2668613466614254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004850301666999722,
      "loss": 5.2235,
      "mean_token_accuracy": 0.20506616681814194,
      "num_tokens": 25155332.0,
      "step": 11280
    },
    {
      "entropy": 5.51335186958313,
      "epoch": 1.2674229235693828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850157635270196,
      "loss": 5.1317,
      "mean_token_accuracy": 0.2070830434560776,
      "num_tokens": 25166024.0,
      "step": 11285
    },
    {
      "entropy": 5.408167886734009,
      "epoch": 1.2679845004773402,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004850013536670836,
      "loss": 5.0906,
      "mean_token_accuracy": 0.2088567242026329,
      "num_tokens": 25176697.0,
      "step": 11290
    },
    {
      "entropy": 5.465492391586304,
      "epoch": 1.268546077385298,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004849869371206231,
      "loss": 5.1892,
      "mean_token_accuracy": 0.20691042691469191,
      "num_tokens": 25187421.0,
      "step": 11295
    },
    {
      "entropy": 5.4739899158477785,
      "epoch": 1.2691076542932556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048497251388809697,
      "loss": 5.1947,
      "mean_token_accuracy": 0.20538371354341506,
      "num_tokens": 25198851.0,
      "step": 11300
    },
    {
      "entropy": 5.529881715774536,
      "epoch": 1.269669231201213,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004849580839699647,
      "loss": 5.2498,
      "mean_token_accuracy": 0.19940355271100998,
      "num_tokens": 25209984.0,
      "step": 11305
    },
    {
      "entropy": 5.468442010879516,
      "epoch": 1.2702308081091704,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048494364736668554,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20520361363887787,
      "num_tokens": 25220672.0,
      "step": 11310
    },
    {
      "entropy": 5.479077625274658,
      "epoch": 1.270792385017128,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004849292040787192,
      "loss": 5.1742,
      "mean_token_accuracy": 0.20183171182870865,
      "num_tokens": 25231906.0,
      "step": 11315
    },
    {
      "entropy": 5.472109699249268,
      "epoch": 1.2713539619250858,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004849147541065256,
      "loss": 5.1973,
      "mean_token_accuracy": 0.19674336910247803,
      "num_tokens": 25242902.0,
      "step": 11320
    },
    {
      "entropy": 5.454363536834717,
      "epoch": 1.2719155388330432,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004849002974505649,
      "loss": 5.251,
      "mean_token_accuracy": 0.19797829538583755,
      "num_tokens": 25254363.0,
      "step": 11325
    },
    {
      "entropy": 5.445118999481201,
      "epoch": 1.2724771157410006,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00048488583411129723,
      "loss": 5.1761,
      "mean_token_accuracy": 0.2032540336251259,
      "num_tokens": 25267108.0,
      "step": 11330
    },
    {
      "entropy": 5.450331544876098,
      "epoch": 1.2730386926489583,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004848713640891832,
      "loss": 5.1342,
      "mean_token_accuracy": 0.20622156262397767,
      "num_tokens": 25278583.0,
      "step": 11335
    },
    {
      "entropy": 5.533629417419434,
      "epoch": 1.2736002695569157,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004848568873846835,
      "loss": 5.2091,
      "mean_token_accuracy": 0.20306390523910522,
      "num_tokens": 25289813.0,
      "step": 11340
    },
    {
      "entropy": 5.403175354003906,
      "epoch": 1.2741618464648734,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048484240399825925,
      "loss": 5.125,
      "mean_token_accuracy": 0.19999108612537383,
      "num_tokens": 25301003.0,
      "step": 11345
    },
    {
      "entropy": 5.450924444198608,
      "epoch": 1.2747234233728308,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048482791393037134,
      "loss": 5.1203,
      "mean_token_accuracy": 0.20888183116912842,
      "num_tokens": 25312535.0,
      "step": 11350
    },
    {
      "entropy": 5.439462423324585,
      "epoch": 1.2752850002807885,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048481341718148137,
      "loss": 5.1494,
      "mean_token_accuracy": 0.20261918902397155,
      "num_tokens": 25323333.0,
      "step": 11355
    },
    {
      "entropy": 5.508570623397827,
      "epoch": 1.275846577188746,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048479891375205074,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19969874769449233,
      "num_tokens": 25334865.0,
      "step": 11360
    },
    {
      "entropy": 5.513240337371826,
      "epoch": 1.2764081540967036,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048478440364254136,
      "loss": 5.2319,
      "mean_token_accuracy": 0.20283337384462358,
      "num_tokens": 25346672.0,
      "step": 11365
    },
    {
      "entropy": 5.468957614898682,
      "epoch": 1.276969731004661,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004847698868534152,
      "loss": 5.2316,
      "mean_token_accuracy": 0.2065620318055153,
      "num_tokens": 25358702.0,
      "step": 11370
    },
    {
      "entropy": 5.472881698608399,
      "epoch": 1.2775313079126187,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004847553633851345,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20115307420492173,
      "num_tokens": 25369813.0,
      "step": 11375
    },
    {
      "entropy": 5.530847454071045,
      "epoch": 1.2780928848205761,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004847408332381616,
      "loss": 5.1907,
      "mean_token_accuracy": 0.20128828138113022,
      "num_tokens": 25381042.0,
      "step": 11380
    },
    {
      "entropy": 5.4648497104644775,
      "epoch": 1.2786544617285336,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048472629641295933,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19565146565437316,
      "num_tokens": 25391286.0,
      "step": 11385
    },
    {
      "entropy": 5.441607522964477,
      "epoch": 1.2792160386364912,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004847117529099904,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20442129224538802,
      "num_tokens": 25402335.0,
      "step": 11390
    },
    {
      "entropy": 5.449425268173218,
      "epoch": 1.279777615544449,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048469720272971797,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21114486455917358,
      "num_tokens": 25413484.0,
      "step": 11395
    },
    {
      "entropy": 5.5259312152862545,
      "epoch": 1.2803391924524063,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004846826458726052,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19536657482385636,
      "num_tokens": 25424276.0,
      "step": 11400
    },
    {
      "entropy": 5.353202772140503,
      "epoch": 1.2809007693603638,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048466808233911567,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20694508105516435,
      "num_tokens": 25435178.0,
      "step": 11405
    },
    {
      "entropy": 5.352679920196533,
      "epoch": 1.2814623462683215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048465351212971313,
      "loss": 5.0281,
      "mean_token_accuracy": 0.2078810840845108,
      "num_tokens": 25445792.0,
      "step": 11410
    },
    {
      "entropy": 5.429979658126831,
      "epoch": 1.2820239231762791,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004846389352448614,
      "loss": 5.1231,
      "mean_token_accuracy": 0.20382663756608962,
      "num_tokens": 25456598.0,
      "step": 11415
    },
    {
      "entropy": 5.446813011169434,
      "epoch": 1.2825855000842366,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004846243516850247,
      "loss": 5.1952,
      "mean_token_accuracy": 0.2079135850071907,
      "num_tokens": 25468311.0,
      "step": 11420
    },
    {
      "entropy": 5.5507570743560795,
      "epoch": 1.283147076992194,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004846097614506673,
      "loss": 5.2211,
      "mean_token_accuracy": 0.19903699308633804,
      "num_tokens": 25480680.0,
      "step": 11425
    },
    {
      "entropy": 5.561702013015747,
      "epoch": 1.2837086539001517,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048459516454225384,
      "loss": 5.1965,
      "mean_token_accuracy": 0.20443698316812514,
      "num_tokens": 25492255.0,
      "step": 11430
    },
    {
      "entropy": 5.449211072921753,
      "epoch": 1.284270230808109,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000484580560960249,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19804598689079284,
      "num_tokens": 25504530.0,
      "step": 11435
    },
    {
      "entropy": 5.413377380371093,
      "epoch": 1.2848318077160668,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004845659507051178,
      "loss": 5.1894,
      "mean_token_accuracy": 0.20682209730148315,
      "num_tokens": 25515273.0,
      "step": 11440
    },
    {
      "entropy": 5.530810308456421,
      "epoch": 1.2853933846240242,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004845513337773255,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19805059731006622,
      "num_tokens": 25526050.0,
      "step": 11445
    },
    {
      "entropy": 5.519785165786743,
      "epoch": 1.2859549615319819,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048453671017733755,
      "loss": 5.1893,
      "mean_token_accuracy": 0.1991667255759239,
      "num_tokens": 25538372.0,
      "step": 11450
    },
    {
      "entropy": 5.563655519485474,
      "epoch": 1.2865165384399393,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004845220799056194,
      "loss": 5.2899,
      "mean_token_accuracy": 0.2067184716463089,
      "num_tokens": 25549636.0,
      "step": 11455
    },
    {
      "entropy": 5.508194637298584,
      "epoch": 1.287078115347897,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048450744296263685,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19643103182315827,
      "num_tokens": 25560850.0,
      "step": 11460
    },
    {
      "entropy": 5.457922792434692,
      "epoch": 1.2876396922558544,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048449279934885625,
      "loss": 5.1498,
      "mean_token_accuracy": 0.2051679864525795,
      "num_tokens": 25572433.0,
      "step": 11465
    },
    {
      "entropy": 5.559819889068604,
      "epoch": 1.288201269163812,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004844781490647435,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20135812610387802,
      "num_tokens": 25582372.0,
      "step": 11470
    },
    {
      "entropy": 5.409285068511963,
      "epoch": 1.2887628460717695,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004844634921107654,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21044907718896866,
      "num_tokens": 25592570.0,
      "step": 11475
    },
    {
      "entropy": 5.465415954589844,
      "epoch": 1.289324422979727,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048444882848738835,
      "loss": 5.1894,
      "mean_token_accuracy": 0.19669198393821716,
      "num_tokens": 25603573.0,
      "step": 11480
    },
    {
      "entropy": 5.538363456726074,
      "epoch": 1.2898859998876846,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048443415819507944,
      "loss": 5.2092,
      "mean_token_accuracy": 0.2019707441329956,
      "num_tokens": 25614233.0,
      "step": 11485
    },
    {
      "entropy": 5.475957727432251,
      "epoch": 1.2904475767956423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004844194812343057,
      "loss": 5.1754,
      "mean_token_accuracy": 0.19788192361593246,
      "num_tokens": 25625077.0,
      "step": 11490
    },
    {
      "entropy": 5.425382995605469,
      "epoch": 1.2910091537035997,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004844047976055344,
      "loss": 5.2113,
      "mean_token_accuracy": 0.20058429688215257,
      "num_tokens": 25636693.0,
      "step": 11495
    },
    {
      "entropy": 5.38034839630127,
      "epoch": 1.2915707306115571,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048439010730923313,
      "loss": 5.0493,
      "mean_token_accuracy": 0.203751640021801,
      "num_tokens": 25647644.0,
      "step": 11500
    },
    {
      "entropy": 5.410211277008057,
      "epoch": 1.2921323075195148,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048437541034586964,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21092393845319748,
      "num_tokens": 25659373.0,
      "step": 11505
    },
    {
      "entropy": 5.485157871246338,
      "epoch": 1.2926938844274725,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048436070671591195,
      "loss": 5.2177,
      "mean_token_accuracy": 0.20779784470796586,
      "num_tokens": 25670479.0,
      "step": 11510
    },
    {
      "entropy": 5.474063301086426,
      "epoch": 1.29325546133543,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048434599641982806,
      "loss": 5.2252,
      "mean_token_accuracy": 0.2070049911737442,
      "num_tokens": 25682748.0,
      "step": 11515
    },
    {
      "entropy": 5.407731723785401,
      "epoch": 1.2938170382433873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048433127945808645,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21043873429298401,
      "num_tokens": 25693170.0,
      "step": 11520
    },
    {
      "entropy": 5.4454591274261475,
      "epoch": 1.294378615151345,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004843165558311558,
      "loss": 5.1931,
      "mean_token_accuracy": 0.2012602850794792,
      "num_tokens": 25706009.0,
      "step": 11525
    },
    {
      "entropy": 5.533812046051025,
      "epoch": 1.2949401920593024,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004843018255395048,
      "loss": 5.2137,
      "mean_token_accuracy": 0.2025035187602043,
      "num_tokens": 25716036.0,
      "step": 11530
    },
    {
      "entropy": 5.439382791519165,
      "epoch": 1.29550176896726,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048428708858360244,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20981121212244033,
      "num_tokens": 25726516.0,
      "step": 11535
    },
    {
      "entropy": 5.455747413635254,
      "epoch": 1.2960633458752175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048427234496391816,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1975255012512207,
      "num_tokens": 25736340.0,
      "step": 11540
    },
    {
      "entropy": 5.373382949829102,
      "epoch": 1.2966249227831752,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048425759468092113,
      "loss": 5.1584,
      "mean_token_accuracy": 0.19684485644102095,
      "num_tokens": 25746439.0,
      "step": 11545
    },
    {
      "entropy": 5.475325679779052,
      "epoch": 1.2971864996911326,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004842428377350811,
      "loss": 5.0736,
      "mean_token_accuracy": 0.2052624061703682,
      "num_tokens": 25757065.0,
      "step": 11550
    },
    {
      "entropy": 5.4440085887908936,
      "epoch": 1.2977480765990903,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048422807412686796,
      "loss": 5.1114,
      "mean_token_accuracy": 0.2017744794487953,
      "num_tokens": 25768021.0,
      "step": 11555
    },
    {
      "entropy": 5.467814111709595,
      "epoch": 1.2983096535070477,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004842133038567519,
      "loss": 5.2243,
      "mean_token_accuracy": 0.20331018716096877,
      "num_tokens": 25780197.0,
      "step": 11560
    },
    {
      "entropy": 5.463663816452026,
      "epoch": 1.2988712304150054,
      "grad_norm": 0.984375,
      "learning_rate": 0.000484198526925203,
      "loss": 5.114,
      "mean_token_accuracy": 0.2099816843867302,
      "num_tokens": 25790536.0,
      "step": 11565
    },
    {
      "entropy": 5.439821195602417,
      "epoch": 1.2994328073229628,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004841837433326919,
      "loss": 5.1183,
      "mean_token_accuracy": 0.20663629919290544,
      "num_tokens": 25802397.0,
      "step": 11570
    },
    {
      "entropy": 5.4992047309875485,
      "epoch": 1.2999943842309205,
      "grad_norm": 1.0,
      "learning_rate": 0.0004841689530796893,
      "loss": 5.2305,
      "mean_token_accuracy": 0.20374708175659179,
      "num_tokens": 25813961.0,
      "step": 11575
    },
    {
      "entropy": 5.498616552352905,
      "epoch": 1.300555961138878,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004841541561666661,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19572438150644303,
      "num_tokens": 25825610.0,
      "step": 11580
    },
    {
      "entropy": 5.44111852645874,
      "epoch": 1.3011175380468356,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048413935259409343,
      "loss": 5.0623,
      "mean_token_accuracy": 0.2041834697127342,
      "num_tokens": 25836076.0,
      "step": 11585
    },
    {
      "entropy": 5.455546188354492,
      "epoch": 1.301679114954793,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048412454236244267,
      "loss": 5.1961,
      "mean_token_accuracy": 0.20554250329732895,
      "num_tokens": 25847341.0,
      "step": 11590
    },
    {
      "entropy": 5.3581174373626705,
      "epoch": 1.3022406918627505,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048410972547218545,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20857871621847152,
      "num_tokens": 25856912.0,
      "step": 11595
    },
    {
      "entropy": 5.463742589950561,
      "epoch": 1.3028022687707081,
      "grad_norm": 1.0,
      "learning_rate": 0.00048409490192379337,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20344580411911012,
      "num_tokens": 25868370.0,
      "step": 11600
    },
    {
      "entropy": 5.422854089736939,
      "epoch": 1.3033638456786658,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004840800717177386,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19839558452367784,
      "num_tokens": 25879104.0,
      "step": 11605
    },
    {
      "entropy": 5.443877029418945,
      "epoch": 1.3039254225866233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004840652348544931,
      "loss": 5.1832,
      "mean_token_accuracy": 0.1958763599395752,
      "num_tokens": 25889744.0,
      "step": 11610
    },
    {
      "entropy": 5.575684261322022,
      "epoch": 1.3044869994945807,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004840503913345296,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1950350135564804,
      "num_tokens": 25901615.0,
      "step": 11615
    },
    {
      "entropy": 5.508506917953492,
      "epoch": 1.3050485764025384,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004840355411583205,
      "loss": 5.1775,
      "mean_token_accuracy": 0.19562424421310426,
      "num_tokens": 25911418.0,
      "step": 11620
    },
    {
      "entropy": 5.5307636737823485,
      "epoch": 1.3056101533104958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048402068432633864,
      "loss": 5.251,
      "mean_token_accuracy": 0.20618563294410705,
      "num_tokens": 25922141.0,
      "step": 11625
    },
    {
      "entropy": 5.491657733917236,
      "epoch": 1.3061717302184535,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004840058208390572,
      "loss": 5.2633,
      "mean_token_accuracy": 0.19767802059650422,
      "num_tokens": 25933897.0,
      "step": 11630
    },
    {
      "entropy": 5.506998348236084,
      "epoch": 1.306733307126411,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004839909506969494,
      "loss": 5.1973,
      "mean_token_accuracy": 0.20301807224750518,
      "num_tokens": 25943960.0,
      "step": 11635
    },
    {
      "entropy": 5.594176864624023,
      "epoch": 1.3072948840343686,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004839760739004886,
      "loss": 5.2126,
      "mean_token_accuracy": 0.19951733499765395,
      "num_tokens": 25955104.0,
      "step": 11640
    },
    {
      "entropy": 5.459704113006592,
      "epoch": 1.307856460942326,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048396119045014853,
      "loss": 5.1978,
      "mean_token_accuracy": 0.2035454422235489,
      "num_tokens": 25965180.0,
      "step": 11645
    },
    {
      "entropy": 5.434148502349854,
      "epoch": 1.3084180378502837,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004839463003464032,
      "loss": 5.1775,
      "mean_token_accuracy": 0.2046498641371727,
      "num_tokens": 25976600.0,
      "step": 11650
    },
    {
      "entropy": 5.451714944839478,
      "epoch": 1.308979614758241,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004839314035897265,
      "loss": 5.1563,
      "mean_token_accuracy": 0.19908519089221954,
      "num_tokens": 25987509.0,
      "step": 11655
    },
    {
      "entropy": 5.469056272506714,
      "epoch": 1.3095411916661988,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000483916500180593,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20938700437545776,
      "num_tokens": 25997962.0,
      "step": 11660
    },
    {
      "entropy": 5.459711647033691,
      "epoch": 1.3101027685741562,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048390159011947706,
      "loss": 5.1428,
      "mean_token_accuracy": 0.20587421357631683,
      "num_tokens": 26008932.0,
      "step": 11665
    },
    {
      "entropy": 5.430537462234497,
      "epoch": 1.3106643454821139,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004838866734068535,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1993084579706192,
      "num_tokens": 26020513.0,
      "step": 11670
    },
    {
      "entropy": 5.495750808715821,
      "epoch": 1.3112259223900713,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048387175004319723,
      "loss": 5.2143,
      "mean_token_accuracy": 0.20191456824541093,
      "num_tokens": 26032526.0,
      "step": 11675
    },
    {
      "entropy": 5.485930252075195,
      "epoch": 1.311787499298029,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004838568200289834,
      "loss": 5.1493,
      "mean_token_accuracy": 0.21214838773012162,
      "num_tokens": 26044216.0,
      "step": 11680
    },
    {
      "entropy": 5.437861728668213,
      "epoch": 1.3123490762059864,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004838418833646875,
      "loss": 5.0629,
      "mean_token_accuracy": 0.20374813079833984,
      "num_tokens": 26054012.0,
      "step": 11685
    },
    {
      "entropy": 5.460941791534424,
      "epoch": 1.3129106531139438,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048382694005078485,
      "loss": 5.2113,
      "mean_token_accuracy": 0.20273168832063676,
      "num_tokens": 26065365.0,
      "step": 11690
    },
    {
      "entropy": 5.4863046169281,
      "epoch": 1.3134722300219015,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048381199008775165,
      "loss": 5.2089,
      "mean_token_accuracy": 0.19587379097938537,
      "num_tokens": 26075964.0,
      "step": 11695
    },
    {
      "entropy": 5.436554956436157,
      "epoch": 1.3140338069298592,
      "grad_norm": 0.875,
      "learning_rate": 0.0004837970334760635,
      "loss": 5.1805,
      "mean_token_accuracy": 0.20143571645021438,
      "num_tokens": 26087572.0,
      "step": 11700
    },
    {
      "entropy": 5.450084638595581,
      "epoch": 1.3145953838378166,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000483782070216197,
      "loss": 5.0736,
      "mean_token_accuracy": 0.20971333086490632,
      "num_tokens": 26098612.0,
      "step": 11705
    },
    {
      "entropy": 5.492118835449219,
      "epoch": 1.315156960745774,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004837671003086283,
      "loss": 5.2371,
      "mean_token_accuracy": 0.20154134929180145,
      "num_tokens": 26109491.0,
      "step": 11710
    },
    {
      "entropy": 5.451224231719971,
      "epoch": 1.3157185376537317,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004837521237538342,
      "loss": 5.1378,
      "mean_token_accuracy": 0.2013961061835289,
      "num_tokens": 26120544.0,
      "step": 11715
    },
    {
      "entropy": 5.533825063705445,
      "epoch": 1.3162801145616891,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004837371405522915,
      "loss": 5.218,
      "mean_token_accuracy": 0.19877072423696518,
      "num_tokens": 26131549.0,
      "step": 11720
    },
    {
      "entropy": 5.461523771286011,
      "epoch": 1.3168416914696468,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048372215070447734,
      "loss": 5.2072,
      "mean_token_accuracy": 0.20401265770196914,
      "num_tokens": 26142871.0,
      "step": 11725
    },
    {
      "entropy": 5.5595122337341305,
      "epoch": 1.3174032683776042,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048370715421086884,
      "loss": 5.3066,
      "mean_token_accuracy": 0.20428167879581452,
      "num_tokens": 26155293.0,
      "step": 11730
    },
    {
      "entropy": 5.5505678176879885,
      "epoch": 1.317964845285562,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004836921510719438,
      "loss": 5.223,
      "mean_token_accuracy": 0.2081404894590378,
      "num_tokens": 26167069.0,
      "step": 11735
    },
    {
      "entropy": 5.487321853637695,
      "epoch": 1.3185264221935193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004836771412881795,
      "loss": 5.2136,
      "mean_token_accuracy": 0.19338219165802,
      "num_tokens": 26179142.0,
      "step": 11740
    },
    {
      "entropy": 5.482232666015625,
      "epoch": 1.319087999101477,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004836621248600542,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20442728847265243,
      "num_tokens": 26190059.0,
      "step": 11745
    },
    {
      "entropy": 5.506552076339721,
      "epoch": 1.3196495760094344,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004836471017880459,
      "loss": 5.1684,
      "mean_token_accuracy": 0.20677920579910278,
      "num_tokens": 26199859.0,
      "step": 11750
    },
    {
      "entropy": 5.50367841720581,
      "epoch": 1.320211152917392,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004836320720726329,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1983455240726471,
      "num_tokens": 26211920.0,
      "step": 11755
    },
    {
      "entropy": 5.465070009231567,
      "epoch": 1.3207727298253495,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004836170357142939,
      "loss": 5.1618,
      "mean_token_accuracy": 0.20120519548654556,
      "num_tokens": 26222875.0,
      "step": 11760
    },
    {
      "entropy": 5.475478029251098,
      "epoch": 1.3213343067333072,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004836019927135076,
      "loss": 5.2014,
      "mean_token_accuracy": 0.20030269920825958,
      "num_tokens": 26233549.0,
      "step": 11765
    },
    {
      "entropy": 5.4200211524963375,
      "epoch": 1.3218958836412646,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004835869430707529,
      "loss": 5.0754,
      "mean_token_accuracy": 0.21423650830984114,
      "num_tokens": 26243886.0,
      "step": 11770
    },
    {
      "entropy": 5.416882038116455,
      "epoch": 1.3224574605492223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048357188678650905,
      "loss": 5.1065,
      "mean_token_accuracy": 0.20935882329940797,
      "num_tokens": 26254475.0,
      "step": 11775
    },
    {
      "entropy": 5.542265462875366,
      "epoch": 1.3230190374571797,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004835568238612553,
      "loss": 5.3059,
      "mean_token_accuracy": 0.1964566320180893,
      "num_tokens": 26265412.0,
      "step": 11780
    },
    {
      "entropy": 5.55201096534729,
      "epoch": 1.3235806143651372,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048354175429547155,
      "loss": 5.2227,
      "mean_token_accuracy": 0.2014821782708168,
      "num_tokens": 26276942.0,
      "step": 11785
    },
    {
      "entropy": 5.533633899688721,
      "epoch": 1.3241421912730948,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004835266780896374,
      "loss": 5.1803,
      "mean_token_accuracy": 0.2036018192768097,
      "num_tokens": 26288009.0,
      "step": 11790
    },
    {
      "entropy": 5.531124925613403,
      "epoch": 1.3247037681810525,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004835115952442329,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1918231353163719,
      "num_tokens": 26301583.0,
      "step": 11795
    },
    {
      "entropy": 5.5593818664550785,
      "epoch": 1.32526534508901,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048349650575973843,
      "loss": 5.2334,
      "mean_token_accuracy": 0.20014299750328063,
      "num_tokens": 26312922.0,
      "step": 11800
    },
    {
      "entropy": 5.514520692825317,
      "epoch": 1.3258269219969674,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048348140963663436,
      "loss": 5.1904,
      "mean_token_accuracy": 0.20839452892541885,
      "num_tokens": 26324108.0,
      "step": 11805
    },
    {
      "entropy": 5.502158784866333,
      "epoch": 1.326388498904925,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004834663068754012,
      "loss": 5.1988,
      "mean_token_accuracy": 0.20654161423444747,
      "num_tokens": 26335669.0,
      "step": 11810
    },
    {
      "entropy": 5.386802387237549,
      "epoch": 1.3269500758128827,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004834511974765202,
      "loss": 5.1043,
      "mean_token_accuracy": 0.20874847322702408,
      "num_tokens": 26346065.0,
      "step": 11815
    },
    {
      "entropy": 5.475204086303711,
      "epoch": 1.3275116527208402,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048343608144047194,
      "loss": 5.0872,
      "mean_token_accuracy": 0.20399933755397798,
      "num_tokens": 26357567.0,
      "step": 11820
    },
    {
      "entropy": 5.473665332794189,
      "epoch": 1.3280732296287976,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048342095876773825,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20347937047481537,
      "num_tokens": 26367805.0,
      "step": 11825
    },
    {
      "entropy": 5.417301416397095,
      "epoch": 1.3286348065367553,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004834058294588003,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2077891618013382,
      "num_tokens": 26379343.0,
      "step": 11830
    },
    {
      "entropy": 5.500841093063355,
      "epoch": 1.3291963834447127,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00048339069351413993,
      "loss": 5.2879,
      "mean_token_accuracy": 0.19741900116205216,
      "num_tokens": 26392039.0,
      "step": 11835
    },
    {
      "entropy": 5.3539954662323,
      "epoch": 1.3297579603526704,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000483375550934239,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2151583731174469,
      "num_tokens": 26403324.0,
      "step": 11840
    },
    {
      "entropy": 5.414818334579468,
      "epoch": 1.3303195372606278,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004833604017195798,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20471826940774918,
      "num_tokens": 26413435.0,
      "step": 11845
    },
    {
      "entropy": 5.462999868392944,
      "epoch": 1.3308811141685855,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004833452458706445,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19662725329399108,
      "num_tokens": 26424254.0,
      "step": 11850
    },
    {
      "entropy": 5.52613377571106,
      "epoch": 1.331442691076543,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004833300833879159,
      "loss": 5.1488,
      "mean_token_accuracy": 0.2063932716846466,
      "num_tokens": 26435658.0,
      "step": 11855
    },
    {
      "entropy": 5.5593963146209715,
      "epoch": 1.3320042679845006,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004833149142718765,
      "loss": 5.2046,
      "mean_token_accuracy": 0.2018562972545624,
      "num_tokens": 26447174.0,
      "step": 11860
    },
    {
      "entropy": 5.541952466964721,
      "epoch": 1.332565844892458,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004832997385230096,
      "loss": 5.2961,
      "mean_token_accuracy": 0.1981729656457901,
      "num_tokens": 26459303.0,
      "step": 11865
    },
    {
      "entropy": 5.471814203262329,
      "epoch": 1.3331274218004157,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004832845561417981,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1970518112182617,
      "num_tokens": 26469981.0,
      "step": 11870
    },
    {
      "entropy": 5.430123233795166,
      "epoch": 1.333688998708373,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004832693671287256,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20668212324380875,
      "num_tokens": 26481176.0,
      "step": 11875
    },
    {
      "entropy": 5.432497978210449,
      "epoch": 1.3342505756163305,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004832541714842757,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20593966394662858,
      "num_tokens": 26492795.0,
      "step": 11880
    },
    {
      "entropy": 5.516925382614136,
      "epoch": 1.3348121525242882,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004832389692089321,
      "loss": 5.2748,
      "mean_token_accuracy": 0.1994077369570732,
      "num_tokens": 26504731.0,
      "step": 11885
    },
    {
      "entropy": 5.546652984619141,
      "epoch": 1.3353737294322459,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048322376030317905,
      "loss": 5.1481,
      "mean_token_accuracy": 0.20360081940889357,
      "num_tokens": 26516816.0,
      "step": 11890
    },
    {
      "entropy": 5.492306327819824,
      "epoch": 1.3359353063402033,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004832085447675007,
      "loss": 5.2653,
      "mean_token_accuracy": 0.20224468559026718,
      "num_tokens": 26528861.0,
      "step": 11895
    },
    {
      "entropy": 5.547552871704101,
      "epoch": 1.3364968832481607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048319332260238153,
      "loss": 5.22,
      "mean_token_accuracy": 0.20258211195468903,
      "num_tokens": 26538816.0,
      "step": 11900
    },
    {
      "entropy": 5.58981499671936,
      "epoch": 1.3370584601561184,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048317809380830624,
      "loss": 5.2639,
      "mean_token_accuracy": 0.1990650162100792,
      "num_tokens": 26548959.0,
      "step": 11905
    },
    {
      "entropy": 5.390769195556641,
      "epoch": 1.337620037064076,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004831628583857597,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20601357370615006,
      "num_tokens": 26559189.0,
      "step": 11910
    },
    {
      "entropy": 5.428310394287109,
      "epoch": 1.3381816139720335,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048314761633522696,
      "loss": 5.083,
      "mean_token_accuracy": 0.21442463546991347,
      "num_tokens": 26570066.0,
      "step": 11915
    },
    {
      "entropy": 5.528180932998657,
      "epoch": 1.338743190879991,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004831323676571934,
      "loss": 5.264,
      "mean_token_accuracy": 0.20499378442764282,
      "num_tokens": 26581489.0,
      "step": 11920
    },
    {
      "entropy": 5.541092777252198,
      "epoch": 1.3393047677879486,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004831171123521446,
      "loss": 5.1896,
      "mean_token_accuracy": 0.19802570939064026,
      "num_tokens": 26592051.0,
      "step": 11925
    },
    {
      "entropy": 5.507169008255005,
      "epoch": 1.339866344695906,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004831018504205662,
      "loss": 5.1857,
      "mean_token_accuracy": 0.20604109019041061,
      "num_tokens": 26602296.0,
      "step": 11930
    },
    {
      "entropy": 5.451931715011597,
      "epoch": 1.3404279216038637,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048308658186294414,
      "loss": 5.1866,
      "mean_token_accuracy": 0.19702045023441314,
      "num_tokens": 26613837.0,
      "step": 11935
    },
    {
      "entropy": 5.490351581573487,
      "epoch": 1.3409894985118211,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004830713066797646,
      "loss": 5.2221,
      "mean_token_accuracy": 0.19948205053806306,
      "num_tokens": 26624898.0,
      "step": 11940
    },
    {
      "entropy": 5.468191623687744,
      "epoch": 1.3415510754197788,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048305602487151396,
      "loss": 5.2077,
      "mean_token_accuracy": 0.20306269824504852,
      "num_tokens": 26636449.0,
      "step": 11945
    },
    {
      "entropy": 5.6156463623046875,
      "epoch": 1.3421126523277362,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004830407364386789,
      "loss": 5.2994,
      "mean_token_accuracy": 0.19780701100826265,
      "num_tokens": 26649192.0,
      "step": 11950
    },
    {
      "entropy": 5.46129789352417,
      "epoch": 1.342674229235694,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000483025441381746,
      "loss": 5.1128,
      "mean_token_accuracy": 0.2069813773036003,
      "num_tokens": 26660005.0,
      "step": 11955
    },
    {
      "entropy": 5.548576545715332,
      "epoch": 1.3432358061436513,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048301013970120235,
      "loss": 5.2574,
      "mean_token_accuracy": 0.19758400022983552,
      "num_tokens": 26672165.0,
      "step": 11960
    },
    {
      "entropy": 5.449719429016113,
      "epoch": 1.343797383051609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048299483139753517,
      "loss": 5.1989,
      "mean_token_accuracy": 0.20835707783699037,
      "num_tokens": 26683135.0,
      "step": 11965
    },
    {
      "entropy": 5.4656572341918945,
      "epoch": 1.3443589599595664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004829795164712319,
      "loss": 5.0699,
      "mean_token_accuracy": 0.21054854840040207,
      "num_tokens": 26693724.0,
      "step": 11970
    },
    {
      "entropy": 5.51731243133545,
      "epoch": 1.3449205368675239,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048296419492278023,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19961208552122117,
      "num_tokens": 26705880.0,
      "step": 11975
    },
    {
      "entropy": 5.526037311553955,
      "epoch": 1.3454821137754815,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004829488667526679,
      "loss": 5.2702,
      "mean_token_accuracy": 0.20107422024011612,
      "num_tokens": 26716978.0,
      "step": 11980
    },
    {
      "entropy": 5.580895280838012,
      "epoch": 1.3460436906834392,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048293353196138294,
      "loss": 5.1931,
      "mean_token_accuracy": 0.21192722767591476,
      "num_tokens": 26728380.0,
      "step": 11985
    },
    {
      "entropy": 5.41425085067749,
      "epoch": 1.3466052675913966,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048291819054941375,
      "loss": 5.1529,
      "mean_token_accuracy": 0.2044837862253189,
      "num_tokens": 26739034.0,
      "step": 11990
    },
    {
      "entropy": 5.317131519317627,
      "epoch": 1.347166844499354,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004829028425172487,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2115240603685379,
      "num_tokens": 26750123.0,
      "step": 11995
    },
    {
      "entropy": 5.422116661071778,
      "epoch": 1.3477284214073117,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048288748786537656,
      "loss": 5.1378,
      "mean_token_accuracy": 0.21104429811239242,
      "num_tokens": 26760411.0,
      "step": 12000
    },
    {
      "epoch": 1.3477284214073117,
      "eval_entropy": 5.247241792464828,
      "eval_loss": 5.248316764831543,
      "eval_mean_token_accuracy": 0.20806338551973355,
      "eval_num_tokens": 26760411.0,
      "eval_runtime": 18.4095,
      "eval_samples_per_second": 1684.621,
      "eval_steps_per_second": 210.598,
      "step": 12000
    },
    {
      "entropy": 5.4755456924438475,
      "epoch": 1.3482899983152694,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004828721265942861,
      "loss": 5.2038,
      "mean_token_accuracy": 0.2055389776825905,
      "num_tokens": 26770821.0,
      "step": 12005
    },
    {
      "entropy": 5.50641131401062,
      "epoch": 1.3488515752232269,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048285675870446653,
      "loss": 5.1779,
      "mean_token_accuracy": 0.20557235628366471,
      "num_tokens": 26782099.0,
      "step": 12010
    },
    {
      "entropy": 5.571593618392944,
      "epoch": 1.3494131521311843,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048284138419640724,
      "loss": 5.2992,
      "mean_token_accuracy": 0.20141823291778566,
      "num_tokens": 26794056.0,
      "step": 12015
    },
    {
      "entropy": 5.455738687515259,
      "epoch": 1.349974729039142,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004828260030705976,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20524767637252808,
      "num_tokens": 26804501.0,
      "step": 12020
    },
    {
      "entropy": 5.5493810176849365,
      "epoch": 1.3505363059470994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048281061532752734,
      "loss": 5.203,
      "mean_token_accuracy": 0.2008921205997467,
      "num_tokens": 26814996.0,
      "step": 12025
    },
    {
      "entropy": 5.494820547103882,
      "epoch": 1.351097882855057,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004827952209676866,
      "loss": 5.1394,
      "mean_token_accuracy": 0.21072678714990617,
      "num_tokens": 26825838.0,
      "step": 12030
    },
    {
      "entropy": 5.4804669380187985,
      "epoch": 1.3516594597630145,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048277981999156534,
      "loss": 5.1606,
      "mean_token_accuracy": 0.2049419865012169,
      "num_tokens": 26836320.0,
      "step": 12035
    },
    {
      "entropy": 5.483975124359131,
      "epoch": 1.3522210366709722,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048276441239965414,
      "loss": 5.1393,
      "mean_token_accuracy": 0.20834504812955856,
      "num_tokens": 26846485.0,
      "step": 12040
    },
    {
      "entropy": 5.588258266448975,
      "epoch": 1.3527826135789296,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004827489981924434,
      "loss": 5.2574,
      "mean_token_accuracy": 0.19489187747240067,
      "num_tokens": 26858554.0,
      "step": 12045
    },
    {
      "entropy": 5.501278686523437,
      "epoch": 1.3533441904868873,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048273357737042393,
      "loss": 5.1941,
      "mean_token_accuracy": 0.20413232147693633,
      "num_tokens": 26869538.0,
      "step": 12050
    },
    {
      "entropy": 5.51456847190857,
      "epoch": 1.3539057673948447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048271814993408694,
      "loss": 5.2619,
      "mean_token_accuracy": 0.195559298992157,
      "num_tokens": 26880284.0,
      "step": 12055
    },
    {
      "entropy": 5.510677671432495,
      "epoch": 1.3544673443028024,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004827027158839233,
      "loss": 5.2273,
      "mean_token_accuracy": 0.19601806849241257,
      "num_tokens": 26893180.0,
      "step": 12060
    },
    {
      "entropy": 5.453428459167481,
      "epoch": 1.3550289212107598,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004826872752204248,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20847864747047423,
      "num_tokens": 26904213.0,
      "step": 12065
    },
    {
      "entropy": 5.478480052947998,
      "epoch": 1.3555904981187172,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048267182794408276,
      "loss": 5.1921,
      "mean_token_accuracy": 0.20832038074731826,
      "num_tokens": 26915566.0,
      "step": 12070
    },
    {
      "entropy": 5.387891435623169,
      "epoch": 1.356152075026675,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004826563740553893,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20773425847291946,
      "num_tokens": 26926253.0,
      "step": 12075
    },
    {
      "entropy": 5.4375075340271,
      "epoch": 1.3567136519346326,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004826409135548363,
      "loss": 5.1337,
      "mean_token_accuracy": 0.2113666296005249,
      "num_tokens": 26936723.0,
      "step": 12080
    },
    {
      "entropy": 5.441137218475342,
      "epoch": 1.35727522884259,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004826254464429161,
      "loss": 5.126,
      "mean_token_accuracy": 0.20685262978076935,
      "num_tokens": 26947551.0,
      "step": 12085
    },
    {
      "entropy": 5.473425817489624,
      "epoch": 1.3578368057505474,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048260997272012114,
      "loss": 5.2377,
      "mean_token_accuracy": 0.2061762347817421,
      "num_tokens": 26959969.0,
      "step": 12090
    },
    {
      "entropy": 5.567711544036865,
      "epoch": 1.358398382658505,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004825944923869441,
      "loss": 5.315,
      "mean_token_accuracy": 0.20031530112028123,
      "num_tokens": 26972511.0,
      "step": 12095
    },
    {
      "entropy": 5.546752691268921,
      "epoch": 1.3589599595664628,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048257900544387795,
      "loss": 5.2878,
      "mean_token_accuracy": 0.1961741790175438,
      "num_tokens": 26983827.0,
      "step": 12100
    },
    {
      "entropy": 5.5921002388000485,
      "epoch": 1.3595215364744202,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048256351189141574,
      "loss": 5.234,
      "mean_token_accuracy": 0.19696773439645768,
      "num_tokens": 26994668.0,
      "step": 12105
    },
    {
      "entropy": 5.437639904022217,
      "epoch": 1.3600831133823776,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004825480117300508,
      "loss": 5.1133,
      "mean_token_accuracy": 0.2130228951573372,
      "num_tokens": 27006313.0,
      "step": 12110
    },
    {
      "entropy": 5.476600599288941,
      "epoch": 1.3606446902903353,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048253250496027675,
      "loss": 5.1635,
      "mean_token_accuracy": 0.20102080702781677,
      "num_tokens": 27017628.0,
      "step": 12115
    },
    {
      "entropy": 5.485951232910156,
      "epoch": 1.3612062671982927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048251699158258714,
      "loss": 5.0766,
      "mean_token_accuracy": 0.2036074712872505,
      "num_tokens": 27028303.0,
      "step": 12120
    },
    {
      "entropy": 5.554651212692261,
      "epoch": 1.3617678441062504,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004825014715974761,
      "loss": 5.2313,
      "mean_token_accuracy": 0.19830369651317598,
      "num_tokens": 27039307.0,
      "step": 12125
    },
    {
      "entropy": 5.4430098056793215,
      "epoch": 1.3623294210142078,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048248594500543767,
      "loss": 5.128,
      "mean_token_accuracy": 0.20548534244298935,
      "num_tokens": 27049271.0,
      "step": 12130
    },
    {
      "entropy": 5.33948917388916,
      "epoch": 1.3628909979221655,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048247041180696633,
      "loss": 5.037,
      "mean_token_accuracy": 0.2095663070678711,
      "num_tokens": 27059510.0,
      "step": 12135
    },
    {
      "entropy": 5.5913416862487795,
      "epoch": 1.363452574830123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004824548720025566,
      "loss": 5.3602,
      "mean_token_accuracy": 0.19199225753545762,
      "num_tokens": 27071685.0,
      "step": 12140
    },
    {
      "entropy": 5.4333202838897705,
      "epoch": 1.3640141517380806,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004824393255927033,
      "loss": 5.0785,
      "mean_token_accuracy": 0.19921932369470596,
      "num_tokens": 27081957.0,
      "step": 12145
    },
    {
      "entropy": 5.423873090744019,
      "epoch": 1.364575728646038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048242377257790145,
      "loss": 5.1124,
      "mean_token_accuracy": 0.20965125858783723,
      "num_tokens": 27091609.0,
      "step": 12150
    },
    {
      "entropy": 5.493868350982666,
      "epoch": 1.3651373055539957,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048240821295864614,
      "loss": 5.2821,
      "mean_token_accuracy": 0.1964243918657303,
      "num_tokens": 27103889.0,
      "step": 12155
    },
    {
      "entropy": 5.56975040435791,
      "epoch": 1.3656988824619531,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000482392646735433,
      "loss": 5.218,
      "mean_token_accuracy": 0.20331528335809707,
      "num_tokens": 27114186.0,
      "step": 12160
    },
    {
      "entropy": 5.397453260421753,
      "epoch": 1.3662604593699106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004823770739087574,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20726438611745834,
      "num_tokens": 27124677.0,
      "step": 12165
    },
    {
      "entropy": 5.566719818115234,
      "epoch": 1.3668220362778682,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004823614944791154,
      "loss": 5.2474,
      "mean_token_accuracy": 0.20322871655225755,
      "num_tokens": 27134873.0,
      "step": 12170
    },
    {
      "entropy": 5.488037204742431,
      "epoch": 1.367383613185826,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004823459084470031,
      "loss": 5.2432,
      "mean_token_accuracy": 0.20298130363225936,
      "num_tokens": 27146107.0,
      "step": 12175
    },
    {
      "entropy": 5.446479988098145,
      "epoch": 1.3679451900937833,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004823303158129165,
      "loss": 5.0958,
      "mean_token_accuracy": 0.2135227158665657,
      "num_tokens": 27158821.0,
      "step": 12180
    },
    {
      "entropy": 5.428579330444336,
      "epoch": 1.3685067670017408,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004823147165773523,
      "loss": 5.1231,
      "mean_token_accuracy": 0.20875139087438582,
      "num_tokens": 27169513.0,
      "step": 12185
    },
    {
      "entropy": 5.406365203857422,
      "epoch": 1.3690683439096984,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048229911074080715,
      "loss": 5.1285,
      "mean_token_accuracy": 0.20999689251184464,
      "num_tokens": 27180059.0,
      "step": 12190
    },
    {
      "entropy": 5.4545081615447994,
      "epoch": 1.369629920817656,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048228349830377787,
      "loss": 5.1922,
      "mean_token_accuracy": 0.20368644893169402,
      "num_tokens": 27190738.0,
      "step": 12195
    },
    {
      "entropy": 5.462658882141113,
      "epoch": 1.3701914977256135,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048226787926676165,
      "loss": 5.0913,
      "mean_token_accuracy": 0.20748201161623,
      "num_tokens": 27200664.0,
      "step": 12200
    },
    {
      "entropy": 5.45639934539795,
      "epoch": 1.370753074633571,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048225225363025576,
      "loss": 5.1573,
      "mean_token_accuracy": 0.20425347983837128,
      "num_tokens": 27211011.0,
      "step": 12205
    },
    {
      "entropy": 5.495973205566406,
      "epoch": 1.3713146515415287,
      "grad_norm": 1.0,
      "learning_rate": 0.00048223662139475773,
      "loss": 5.2185,
      "mean_token_accuracy": 0.20922895520925522,
      "num_tokens": 27221888.0,
      "step": 12210
    },
    {
      "entropy": 5.507156419754028,
      "epoch": 1.371876228449486,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048222098256076523,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19892387837171555,
      "num_tokens": 27234671.0,
      "step": 12215
    },
    {
      "entropy": 5.399673938751221,
      "epoch": 1.3724378053574438,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004822053371287763,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21189790219068527,
      "num_tokens": 27245283.0,
      "step": 12220
    },
    {
      "entropy": 5.458009386062622,
      "epoch": 1.3729993822654012,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004821896850992891,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20680995285511017,
      "num_tokens": 27255056.0,
      "step": 12225
    },
    {
      "entropy": 5.5271751403808596,
      "epoch": 1.3735609591733589,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000482174026472802,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19984131008386613,
      "num_tokens": 27266893.0,
      "step": 12230
    },
    {
      "entropy": 5.452425527572632,
      "epoch": 1.3741225360813163,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004821583612498135,
      "loss": 5.2406,
      "mean_token_accuracy": 0.20188457816839217,
      "num_tokens": 27279504.0,
      "step": 12235
    },
    {
      "entropy": 5.572341346740723,
      "epoch": 1.374684112989274,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048214268943082246,
      "loss": 5.2713,
      "mean_token_accuracy": 0.20340170711278915,
      "num_tokens": 27290525.0,
      "step": 12240
    },
    {
      "entropy": 5.548085880279541,
      "epoch": 1.3752456898972314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821270110163278,
      "loss": 5.1554,
      "mean_token_accuracy": 0.2059785917401314,
      "num_tokens": 27301404.0,
      "step": 12245
    },
    {
      "entropy": 5.460307836532593,
      "epoch": 1.375807266805189,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004821113260068288,
      "loss": 5.0776,
      "mean_token_accuracy": 0.2048070177435875,
      "num_tokens": 27312635.0,
      "step": 12250
    },
    {
      "entropy": 5.3788275718688965,
      "epoch": 1.3763688437131465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004820956344028249,
      "loss": 5.1251,
      "mean_token_accuracy": 0.20740974843502044,
      "num_tokens": 27322485.0,
      "step": 12255
    },
    {
      "entropy": 5.5183103561401365,
      "epoch": 1.376930420621104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048207993620481566,
      "loss": 5.2027,
      "mean_token_accuracy": 0.2012050524353981,
      "num_tokens": 27332568.0,
      "step": 12260
    },
    {
      "entropy": 5.5855084419250485,
      "epoch": 1.3774919975290616,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004820642314133009,
      "loss": 5.2455,
      "mean_token_accuracy": 0.1964574620127678,
      "num_tokens": 27343058.0,
      "step": 12265
    },
    {
      "entropy": 5.468304395675659,
      "epoch": 1.3780535744370193,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048204852002878073,
      "loss": 5.1143,
      "mean_token_accuracy": 0.2115562677383423,
      "num_tokens": 27355316.0,
      "step": 12270
    },
    {
      "entropy": 5.427171087265014,
      "epoch": 1.3786151513449767,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048203280205175544,
      "loss": 5.0642,
      "mean_token_accuracy": 0.21111166030168532,
      "num_tokens": 27366098.0,
      "step": 12275
    },
    {
      "entropy": 5.408334541320801,
      "epoch": 1.3791767282529341,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004820170774827253,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21402163952589034,
      "num_tokens": 27376923.0,
      "step": 12280
    },
    {
      "entropy": 5.456176233291626,
      "epoch": 1.3797383051608918,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004820013463221912,
      "loss": 5.2352,
      "mean_token_accuracy": 0.20305778980255126,
      "num_tokens": 27386789.0,
      "step": 12285
    },
    {
      "entropy": 5.467389297485352,
      "epoch": 1.3802998820688495,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000481985608570654,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20468004941940307,
      "num_tokens": 27399241.0,
      "step": 12290
    },
    {
      "entropy": 5.4002477645874025,
      "epoch": 1.380861458976807,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004819698642286147,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2179410293698311,
      "num_tokens": 27409123.0,
      "step": 12295
    },
    {
      "entropy": 5.382656002044678,
      "epoch": 1.3814230358847643,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048195411329657464,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20253762304782869,
      "num_tokens": 27419920.0,
      "step": 12300
    },
    {
      "entropy": 5.3557709693908695,
      "epoch": 1.381984612792722,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004819383557750353,
      "loss": 5.0009,
      "mean_token_accuracy": 0.2055969163775444,
      "num_tokens": 27429552.0,
      "step": 12305
    },
    {
      "entropy": 5.489142560958863,
      "epoch": 1.3825461897006794,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048192259166449845,
      "loss": 5.2036,
      "mean_token_accuracy": 0.2049853578209877,
      "num_tokens": 27440404.0,
      "step": 12310
    },
    {
      "entropy": 5.488819217681884,
      "epoch": 1.383107766608637,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048190682096546606,
      "loss": 5.2402,
      "mean_token_accuracy": 0.2003431051969528,
      "num_tokens": 27452231.0,
      "step": 12315
    },
    {
      "entropy": 5.475946092605591,
      "epoch": 1.3836693435165945,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048189104367844034,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20725544095039367,
      "num_tokens": 27463658.0,
      "step": 12320
    },
    {
      "entropy": 5.466873121261597,
      "epoch": 1.3842309204245522,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048187525980392343,
      "loss": 5.1843,
      "mean_token_accuracy": 0.20116572976112365,
      "num_tokens": 27475403.0,
      "step": 12325
    },
    {
      "entropy": 5.524809885025024,
      "epoch": 1.3847924973325096,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000481859469342418,
      "loss": 5.2646,
      "mean_token_accuracy": 0.19959893822669983,
      "num_tokens": 27487056.0,
      "step": 12330
    },
    {
      "entropy": 5.4767341136932375,
      "epoch": 1.3853540742404673,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004818436722944269,
      "loss": 5.1191,
      "mean_token_accuracy": 0.21175259202718735,
      "num_tokens": 27498659.0,
      "step": 12335
    },
    {
      "entropy": 5.587176513671875,
      "epoch": 1.3859156511484247,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004818278686604531,
      "loss": 5.2691,
      "mean_token_accuracy": 0.19665899276733398,
      "num_tokens": 27510577.0,
      "step": 12340
    },
    {
      "entropy": 5.446998548507691,
      "epoch": 1.3864772280563824,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004818120584409996,
      "loss": 5.1659,
      "mean_token_accuracy": 0.20675605088472365,
      "num_tokens": 27521441.0,
      "step": 12345
    },
    {
      "entropy": 5.467502641677856,
      "epoch": 1.3870388049643398,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004817962416365701,
      "loss": 5.1011,
      "mean_token_accuracy": 0.21017972379922867,
      "num_tokens": 27533443.0,
      "step": 12350
    },
    {
      "entropy": 5.5100866794586185,
      "epoch": 1.3876003818722973,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048178041824766796,
      "loss": 5.1485,
      "mean_token_accuracy": 0.2124373808503151,
      "num_tokens": 27543872.0,
      "step": 12355
    },
    {
      "entropy": 5.563232278823852,
      "epoch": 1.388161958780255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004817645882747971,
      "loss": 5.3531,
      "mean_token_accuracy": 0.19551461786031724,
      "num_tokens": 27556556.0,
      "step": 12360
    },
    {
      "entropy": 5.503177165985107,
      "epoch": 1.3887235356882126,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004817487517184616,
      "loss": 5.2227,
      "mean_token_accuracy": 0.19677228033542632,
      "num_tokens": 27570739.0,
      "step": 12365
    },
    {
      "entropy": 5.522423505783081,
      "epoch": 1.38928511259617,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004817329085791657,
      "loss": 5.2735,
      "mean_token_accuracy": 0.19688376784324646,
      "num_tokens": 27582438.0,
      "step": 12370
    },
    {
      "entropy": 5.53391580581665,
      "epoch": 1.3898466895041275,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004817170588574138,
      "loss": 5.1738,
      "mean_token_accuracy": 0.20429896414279938,
      "num_tokens": 27593854.0,
      "step": 12375
    },
    {
      "entropy": 5.539191246032715,
      "epoch": 1.3904082664120851,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004817012025537105,
      "loss": 5.1355,
      "mean_token_accuracy": 0.20545539259910583,
      "num_tokens": 27604459.0,
      "step": 12380
    },
    {
      "entropy": 5.548080682754517,
      "epoch": 1.3909698433200428,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004816853396685608,
      "loss": 5.354,
      "mean_token_accuracy": 0.1883169785141945,
      "num_tokens": 27617901.0,
      "step": 12385
    },
    {
      "entropy": 5.520606946945191,
      "epoch": 1.3915314202280002,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048166947020246973,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1959784671664238,
      "num_tokens": 27628574.0,
      "step": 12390
    },
    {
      "entropy": 5.478893184661866,
      "epoch": 1.3920929971359577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048165359415594245,
      "loss": 5.1407,
      "mean_token_accuracy": 0.20215887874364852,
      "num_tokens": 27639195.0,
      "step": 12395
    },
    {
      "entropy": 5.389469146728516,
      "epoch": 1.3926545740439153,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048163771152948473,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21549398005008696,
      "num_tokens": 27649898.0,
      "step": 12400
    },
    {
      "entropy": 5.482058763504028,
      "epoch": 1.3932161509518728,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000481621822323602,
      "loss": 5.1915,
      "mean_token_accuracy": 0.20455410331487656,
      "num_tokens": 27663617.0,
      "step": 12405
    },
    {
      "entropy": 5.41757116317749,
      "epoch": 1.3937777278598305,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004816059265388004,
      "loss": 5.0674,
      "mean_token_accuracy": 0.21241903454065322,
      "num_tokens": 27674823.0,
      "step": 12410
    },
    {
      "entropy": 5.471855497360229,
      "epoch": 1.394339304767788,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048159002417558596,
      "loss": 5.1816,
      "mean_token_accuracy": 0.20129034668207169,
      "num_tokens": 27686483.0,
      "step": 12415
    },
    {
      "entropy": 5.458209657669068,
      "epoch": 1.3949008816757456,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004815741152344651,
      "loss": 5.1231,
      "mean_token_accuracy": 0.2107679948210716,
      "num_tokens": 27697081.0,
      "step": 12420
    },
    {
      "entropy": 5.486216926574707,
      "epoch": 1.395462458583703,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004815581997159442,
      "loss": 5.189,
      "mean_token_accuracy": 0.20550450682640076,
      "num_tokens": 27706710.0,
      "step": 12425
    },
    {
      "entropy": 5.487536764144897,
      "epoch": 1.3960240354916607,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048154227762053003,
      "loss": 5.2288,
      "mean_token_accuracy": 0.19406247586011888,
      "num_tokens": 27718213.0,
      "step": 12430
    },
    {
      "entropy": 5.411784744262695,
      "epoch": 1.396585612399618,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004815263489487298,
      "loss": 5.1581,
      "mean_token_accuracy": 0.1991487443447113,
      "num_tokens": 27729064.0,
      "step": 12435
    },
    {
      "entropy": 5.541603755950928,
      "epoch": 1.3971471893075758,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004815104137010504,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20811409056186675,
      "num_tokens": 27740121.0,
      "step": 12440
    },
    {
      "entropy": 5.483473110198974,
      "epoch": 1.3977087662155332,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004814944718779993,
      "loss": 5.2108,
      "mean_token_accuracy": 0.20785385519266128,
      "num_tokens": 27751597.0,
      "step": 12445
    },
    {
      "entropy": 5.432009029388428,
      "epoch": 1.3982703431234906,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004814785234800843,
      "loss": 5.0965,
      "mean_token_accuracy": 0.2107295125722885,
      "num_tokens": 27761539.0,
      "step": 12450
    },
    {
      "entropy": 5.459167385101319,
      "epoch": 1.3988319200314483,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048146256850781287,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20399469435214995,
      "num_tokens": 27772904.0,
      "step": 12455
    },
    {
      "entropy": 5.525432062149048,
      "epoch": 1.399393496939406,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048144660696169327,
      "loss": 5.113,
      "mean_token_accuracy": 0.2093617334961891,
      "num_tokens": 27783706.0,
      "step": 12460
    },
    {
      "entropy": 5.44769253730774,
      "epoch": 1.3999550738473634,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048143063884223365,
      "loss": 5.1218,
      "mean_token_accuracy": 0.19786850064992906,
      "num_tokens": 27794503.0,
      "step": 12465
    },
    {
      "entropy": 5.435678100585937,
      "epoch": 1.4005166507553208,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004814146641499424,
      "loss": 5.1398,
      "mean_token_accuracy": 0.2010348618030548,
      "num_tokens": 27805952.0,
      "step": 12470
    },
    {
      "entropy": 5.498334836959839,
      "epoch": 1.4010782276632785,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048139868288532816,
      "loss": 5.1355,
      "mean_token_accuracy": 0.20386661291122438,
      "num_tokens": 27817530.0,
      "step": 12475
    },
    {
      "entropy": 5.498781347274781,
      "epoch": 1.4016398045712362,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004813826950488998,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19765318930149078,
      "num_tokens": 27829932.0,
      "step": 12480
    },
    {
      "entropy": 5.459283256530762,
      "epoch": 1.4022013814791936,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048136670064116645,
      "loss": 5.1853,
      "mean_token_accuracy": 0.20807783007621766,
      "num_tokens": 27840993.0,
      "step": 12485
    },
    {
      "entropy": 5.47070198059082,
      "epoch": 1.402762958387151,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004813506996626373,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20531768202781678,
      "num_tokens": 27852727.0,
      "step": 12490
    },
    {
      "entropy": 5.497899770736694,
      "epoch": 1.4033245352951087,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004813346921138218,
      "loss": 5.2867,
      "mean_token_accuracy": 0.20217883586883545,
      "num_tokens": 27865183.0,
      "step": 12495
    },
    {
      "entropy": 5.426887464523316,
      "epoch": 1.4038861122030661,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004813186779952297,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21312264502048492,
      "num_tokens": 27877266.0,
      "step": 12500
    },
    {
      "entropy": 5.531505441665649,
      "epoch": 1.4044476891110238,
      "grad_norm": 0.875,
      "learning_rate": 0.0004813026573073708,
      "loss": 5.1708,
      "mean_token_accuracy": 0.19764145314693451,
      "num_tokens": 27888863.0,
      "step": 12505
    },
    {
      "entropy": 5.466286563873291,
      "epoch": 1.4050092660189812,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004812866300507554,
      "loss": 5.1078,
      "mean_token_accuracy": 0.210232974588871,
      "num_tokens": 27900443.0,
      "step": 12510
    },
    {
      "entropy": 5.452420806884765,
      "epoch": 1.405570842926939,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004812705962258936,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20642200261354446,
      "num_tokens": 27911241.0,
      "step": 12515
    },
    {
      "entropy": 5.465728139877319,
      "epoch": 1.4061324198348963,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004812545558332961,
      "loss": 5.1254,
      "mean_token_accuracy": 0.20888977199792863,
      "num_tokens": 27922171.0,
      "step": 12520
    },
    {
      "entropy": 5.416845989227295,
      "epoch": 1.406693996742854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048123850887347346,
      "loss": 5.1005,
      "mean_token_accuracy": 0.207905550301075,
      "num_tokens": 27933627.0,
      "step": 12525
    },
    {
      "entropy": 5.353041362762451,
      "epoch": 1.4072555736508114,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048122245534693676,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2077394977211952,
      "num_tokens": 27943585.0,
      "step": 12530
    },
    {
      "entropy": 5.471788835525513,
      "epoch": 1.407817150558769,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004812063952541971,
      "loss": 5.1497,
      "mean_token_accuracy": 0.20525009483098983,
      "num_tokens": 27955554.0,
      "step": 12535
    },
    {
      "entropy": 5.496187496185303,
      "epoch": 1.4083787274667265,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048119032859576577,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21123435199260712,
      "num_tokens": 27966150.0,
      "step": 12540
    },
    {
      "entropy": 5.414362001419067,
      "epoch": 1.408940304374684,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004811742553721545,
      "loss": 5.0894,
      "mean_token_accuracy": 0.21142808049917222,
      "num_tokens": 27977436.0,
      "step": 12545
    },
    {
      "entropy": 5.485429048538208,
      "epoch": 1.4095018812826416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048115817558387494,
      "loss": 5.2557,
      "mean_token_accuracy": 0.19952340871095658,
      "num_tokens": 27988325.0,
      "step": 12550
    },
    {
      "entropy": 5.522013711929321,
      "epoch": 1.4100634581905993,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004811420892314391,
      "loss": 5.1072,
      "mean_token_accuracy": 0.20461919158697128,
      "num_tokens": 27999760.0,
      "step": 12555
    },
    {
      "entropy": 5.543334102630615,
      "epoch": 1.4106250350985567,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004811259963153592,
      "loss": 5.2293,
      "mean_token_accuracy": 0.1947714552283287,
      "num_tokens": 28011333.0,
      "step": 12560
    },
    {
      "entropy": 5.43923282623291,
      "epoch": 1.4111866120065142,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004811098968361476,
      "loss": 5.1946,
      "mean_token_accuracy": 0.20771200060844422,
      "num_tokens": 28023464.0,
      "step": 12565
    },
    {
      "entropy": 5.446501684188843,
      "epoch": 1.4117481889144718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000481093790794317,
      "loss": 5.1465,
      "mean_token_accuracy": 0.19933257550001143,
      "num_tokens": 28034205.0,
      "step": 12570
    },
    {
      "entropy": 5.528965091705322,
      "epoch": 1.4123097658224295,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004810776781903801,
      "loss": 5.1368,
      "mean_token_accuracy": 0.2043679401278496,
      "num_tokens": 28045846.0,
      "step": 12575
    },
    {
      "entropy": 5.518973684310913,
      "epoch": 1.412871342730387,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048106155902485003,
      "loss": 5.1513,
      "mean_token_accuracy": 0.2087111920118332,
      "num_tokens": 28057365.0,
      "step": 12580
    },
    {
      "entropy": 5.48867712020874,
      "epoch": 1.4134329196383444,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004810454332982401,
      "loss": 5.2396,
      "mean_token_accuracy": 0.1994670510292053,
      "num_tokens": 28069098.0,
      "step": 12585
    },
    {
      "entropy": 5.474959516525269,
      "epoch": 1.413994496546302,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004810293010110635,
      "loss": 5.1757,
      "mean_token_accuracy": 0.20220031291246415,
      "num_tokens": 28080008.0,
      "step": 12590
    },
    {
      "entropy": 5.5330305099487305,
      "epoch": 1.4145560734542595,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004810131621638342,
      "loss": 5.1869,
      "mean_token_accuracy": 0.2046856924891472,
      "num_tokens": 28091657.0,
      "step": 12595
    },
    {
      "entropy": 5.415086889266968,
      "epoch": 1.4151176503622171,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048099701675706583,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20730396509170532,
      "num_tokens": 28102586.0,
      "step": 12600
    },
    {
      "entropy": 5.468436861038208,
      "epoch": 1.4156792272701746,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004809808647912725,
      "loss": 5.1733,
      "mean_token_accuracy": 0.21040523648262024,
      "num_tokens": 28113066.0,
      "step": 12605
    },
    {
      "entropy": 5.466883802413941,
      "epoch": 1.4162408041781323,
      "grad_norm": 0.875,
      "learning_rate": 0.0004809647062669686,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20165132582187653,
      "num_tokens": 28124814.0,
      "step": 12610
    },
    {
      "entropy": 5.55237193107605,
      "epoch": 1.4168023810860897,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048094854118466856,
      "loss": 5.2247,
      "mean_token_accuracy": 0.19796105176210405,
      "num_tokens": 28138009.0,
      "step": 12615
    },
    {
      "entropy": 5.414374876022339,
      "epoch": 1.4173639579940474,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048093236954488715,
      "loss": 5.1088,
      "mean_token_accuracy": 0.2057243913412094,
      "num_tokens": 28149222.0,
      "step": 12620
    },
    {
      "entropy": 5.437690210342407,
      "epoch": 1.4179255349020048,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004809161913481392,
      "loss": 5.1611,
      "mean_token_accuracy": 0.20204149335622787,
      "num_tokens": 28161323.0,
      "step": 12625
    },
    {
      "entropy": 5.462181043624878,
      "epoch": 1.4184871118099625,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048090000659493985,
      "loss": 5.1617,
      "mean_token_accuracy": 0.20193156003952026,
      "num_tokens": 28173242.0,
      "step": 12630
    },
    {
      "entropy": 5.488607358932495,
      "epoch": 1.41904868871792,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048088381528580445,
      "loss": 5.113,
      "mean_token_accuracy": 0.20966652333736419,
      "num_tokens": 28183927.0,
      "step": 12635
    },
    {
      "entropy": 5.492193078994751,
      "epoch": 1.4196102656258773,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048086761742124846,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20702141970396043,
      "num_tokens": 28194375.0,
      "step": 12640
    },
    {
      "entropy": 5.47846622467041,
      "epoch": 1.420171842533835,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048085141300178774,
      "loss": 5.1851,
      "mean_token_accuracy": 0.2080398365855217,
      "num_tokens": 28206492.0,
      "step": 12645
    },
    {
      "entropy": 5.4760016918182375,
      "epoch": 1.4207334194417927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048083520202793815,
      "loss": 5.2059,
      "mean_token_accuracy": 0.20113577097654342,
      "num_tokens": 28217795.0,
      "step": 12650
    },
    {
      "entropy": 5.4678747177124025,
      "epoch": 1.42129499634975,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004808189845002159,
      "loss": 5.143,
      "mean_token_accuracy": 0.2044993370771408,
      "num_tokens": 28228796.0,
      "step": 12655
    },
    {
      "entropy": 5.497642183303833,
      "epoch": 1.4218565732577075,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004808027604191374,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20616001188755034,
      "num_tokens": 28238873.0,
      "step": 12660
    },
    {
      "entropy": 5.410908126831055,
      "epoch": 1.4224181501656652,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048078652978521914,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21906768679618835,
      "num_tokens": 28249832.0,
      "step": 12665
    },
    {
      "entropy": 5.463288307189941,
      "epoch": 1.4229797270736229,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048077029259897797,
      "loss": 5.1876,
      "mean_token_accuracy": 0.20350343734025955,
      "num_tokens": 28261532.0,
      "step": 12670
    },
    {
      "entropy": 5.576680374145508,
      "epoch": 1.4235413039815803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048075404886093094,
      "loss": 5.1971,
      "mean_token_accuracy": 0.20575254112482072,
      "num_tokens": 28272008.0,
      "step": 12675
    },
    {
      "entropy": 5.491054105758667,
      "epoch": 1.4241028808895377,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048073779857159504,
      "loss": 5.1723,
      "mean_token_accuracy": 0.20502566695213317,
      "num_tokens": 28283239.0,
      "step": 12680
    },
    {
      "entropy": 5.516871070861816,
      "epoch": 1.4246644577974954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000480721541731488,
      "loss": 5.2116,
      "mean_token_accuracy": 0.20059361159801484,
      "num_tokens": 28293597.0,
      "step": 12685
    },
    {
      "entropy": 5.484355735778808,
      "epoch": 1.4252260347054528,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048070527834112727,
      "loss": 5.1397,
      "mean_token_accuracy": 0.20740580260753633,
      "num_tokens": 28305124.0,
      "step": 12690
    },
    {
      "entropy": 5.408352613449097,
      "epoch": 1.4257876116134105,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004806890084010306,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20809488296508788,
      "num_tokens": 28315875.0,
      "step": 12695
    },
    {
      "entropy": 5.433827304840088,
      "epoch": 1.426349188521368,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048067273191171615,
      "loss": 5.1292,
      "mean_token_accuracy": 0.2041878119111061,
      "num_tokens": 28327705.0,
      "step": 12700
    },
    {
      "entropy": 5.496900415420532,
      "epoch": 1.4269107654293256,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004806564488737022,
      "loss": 5.0969,
      "mean_token_accuracy": 0.21106772422790526,
      "num_tokens": 28338300.0,
      "step": 12705
    },
    {
      "entropy": 5.4769346714019775,
      "epoch": 1.427472342337283,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004806401592875071,
      "loss": 5.0657,
      "mean_token_accuracy": 0.21399219185113907,
      "num_tokens": 28348485.0,
      "step": 12710
    },
    {
      "entropy": 5.451786661148072,
      "epoch": 1.4280339192452407,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048062386315364964,
      "loss": 5.0883,
      "mean_token_accuracy": 0.21437736302614213,
      "num_tokens": 28359118.0,
      "step": 12715
    },
    {
      "entropy": 5.427522087097168,
      "epoch": 1.4285954961531981,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004806075604726485,
      "loss": 5.1344,
      "mean_token_accuracy": 0.21084635108709335,
      "num_tokens": 28369608.0,
      "step": 12720
    },
    {
      "entropy": 5.51730170249939,
      "epoch": 1.4291570730611558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048059125124502304,
      "loss": 5.12,
      "mean_token_accuracy": 0.21088398694992067,
      "num_tokens": 28379709.0,
      "step": 12725
    },
    {
      "entropy": 5.464109277725219,
      "epoch": 1.4297186499691132,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004805749354712924,
      "loss": 5.1561,
      "mean_token_accuracy": 0.2023326948285103,
      "num_tokens": 28390758.0,
      "step": 12730
    },
    {
      "entropy": 5.66201171875,
      "epoch": 1.4302802268770707,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000480558613151976,
      "loss": 5.395,
      "mean_token_accuracy": 0.1917519152164459,
      "num_tokens": 28403485.0,
      "step": 12735
    },
    {
      "entropy": 5.537450981140137,
      "epoch": 1.4308418037850283,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004805422842875936,
      "loss": 5.2086,
      "mean_token_accuracy": 0.20338305830955505,
      "num_tokens": 28413627.0,
      "step": 12740
    },
    {
      "entropy": 5.484331035614014,
      "epoch": 1.431403380692986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048052594887866524,
      "loss": 5.0954,
      "mean_token_accuracy": 0.20481521636247635,
      "num_tokens": 28424202.0,
      "step": 12745
    },
    {
      "entropy": 5.365954446792602,
      "epoch": 1.4319649576009434,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048050960692571086,
      "loss": 5.069,
      "mean_token_accuracy": 0.20588970184326172,
      "num_tokens": 28435136.0,
      "step": 12750
    },
    {
      "entropy": 5.434122133255005,
      "epoch": 1.4325265345089009,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000480493258429251,
      "loss": 5.1009,
      "mean_token_accuracy": 0.2046312212944031,
      "num_tokens": 28444915.0,
      "step": 12755
    },
    {
      "entropy": 5.488479280471802,
      "epoch": 1.4330881114168585,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000480476903389806,
      "loss": 5.1681,
      "mean_token_accuracy": 0.19834898114204408,
      "num_tokens": 28458295.0,
      "step": 12760
    },
    {
      "entropy": 5.519573593139649,
      "epoch": 1.4336496883248162,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004804605418078968,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20759073197841643,
      "num_tokens": 28469587.0,
      "step": 12765
    },
    {
      "entropy": 5.556804275512695,
      "epoch": 1.4342112652327736,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048044417368404426,
      "loss": 5.1818,
      "mean_token_accuracy": 0.20200895965099336,
      "num_tokens": 28481498.0,
      "step": 12770
    },
    {
      "entropy": 5.546443700790405,
      "epoch": 1.434772842140731,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048042779901876946,
      "loss": 5.3138,
      "mean_token_accuracy": 0.19586526304483415,
      "num_tokens": 28492626.0,
      "step": 12775
    },
    {
      "entropy": 5.516772222518921,
      "epoch": 1.4353344190486887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004804114178125939,
      "loss": 5.1823,
      "mean_token_accuracy": 0.20486706048250197,
      "num_tokens": 28503882.0,
      "step": 12780
    },
    {
      "entropy": 5.508041143417358,
      "epoch": 1.4358959959566462,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004803950300660391,
      "loss": 5.1114,
      "mean_token_accuracy": 0.21730809062719345,
      "num_tokens": 28514786.0,
      "step": 12785
    },
    {
      "entropy": 5.482204437255859,
      "epoch": 1.4364575728646038,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004803786357796269,
      "loss": 5.1372,
      "mean_token_accuracy": 0.2062377005815506,
      "num_tokens": 28526045.0,
      "step": 12790
    },
    {
      "entropy": 5.4685100555419925,
      "epoch": 1.4370191497725613,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048036223495387934,
      "loss": 5.15,
      "mean_token_accuracy": 0.20564915239810944,
      "num_tokens": 28536773.0,
      "step": 12795
    },
    {
      "entropy": 5.4386231899261475,
      "epoch": 1.437580726680519,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004803458275893185,
      "loss": 5.1064,
      "mean_token_accuracy": 0.20556536465883254,
      "num_tokens": 28548099.0,
      "step": 12800
    },
    {
      "entropy": 5.511852025985718,
      "epoch": 1.4381423035884764,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004803294136864669,
      "loss": 5.1536,
      "mean_token_accuracy": 0.2053316429257393,
      "num_tokens": 28559372.0,
      "step": 12805
    },
    {
      "entropy": 5.4550145149230955,
      "epoch": 1.438703880496434,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048031299324584716,
      "loss": 5.1648,
      "mean_token_accuracy": 0.20393643379211426,
      "num_tokens": 28570106.0,
      "step": 12810
    },
    {
      "entropy": 5.442716646194458,
      "epoch": 1.4392654574043915,
      "grad_norm": 0.890625,
      "learning_rate": 0.000480296566267982,
      "loss": 5.1756,
      "mean_token_accuracy": 0.2047491565346718,
      "num_tokens": 28582417.0,
      "step": 12815
    },
    {
      "entropy": 5.5688739776611325,
      "epoch": 1.4398270343123492,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004802801327533947,
      "loss": 5.299,
      "mean_token_accuracy": 0.20020985305309297,
      "num_tokens": 28594341.0,
      "step": 12820
    },
    {
      "entropy": 5.508680820465088,
      "epoch": 1.4403886112203066,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048026369270260814,
      "loss": 5.174,
      "mean_token_accuracy": 0.20464228689670563,
      "num_tokens": 28605580.0,
      "step": 12825
    },
    {
      "entropy": 5.5399620056152346,
      "epoch": 1.440950188128264,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004802472461161462,
      "loss": 5.2073,
      "mean_token_accuracy": 0.19983220994472503,
      "num_tokens": 28617779.0,
      "step": 12830
    },
    {
      "entropy": 5.519721126556396,
      "epoch": 1.4415117650362217,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004802307929945322,
      "loss": 5.2028,
      "mean_token_accuracy": 0.1919047325849533,
      "num_tokens": 28629714.0,
      "step": 12835
    },
    {
      "entropy": 5.516920661926269,
      "epoch": 1.4420733419441794,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048021433333829015,
      "loss": 5.205,
      "mean_token_accuracy": 0.19859661012887955,
      "num_tokens": 28641381.0,
      "step": 12840
    },
    {
      "entropy": 5.493727493286133,
      "epoch": 1.4426349188521368,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048019786714794424,
      "loss": 5.1894,
      "mean_token_accuracy": 0.2069296732544899,
      "num_tokens": 28653650.0,
      "step": 12845
    },
    {
      "entropy": 5.473764705657959,
      "epoch": 1.4431964957600942,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004801813944240186,
      "loss": 5.1122,
      "mean_token_accuracy": 0.20553876757621764,
      "num_tokens": 28665252.0,
      "step": 12850
    },
    {
      "entropy": 5.449726390838623,
      "epoch": 1.443758072668052,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004801649151670378,
      "loss": 5.165,
      "mean_token_accuracy": 0.20538520365953444,
      "num_tokens": 28677072.0,
      "step": 12855
    },
    {
      "entropy": 5.424026155471802,
      "epoch": 1.4443196495760096,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004801484293775265,
      "loss": 5.0941,
      "mean_token_accuracy": 0.20484882593154907,
      "num_tokens": 28688767.0,
      "step": 12860
    },
    {
      "entropy": 5.549541664123535,
      "epoch": 1.444881226483967,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004801319370560097,
      "loss": 5.3319,
      "mean_token_accuracy": 0.20201486349105835,
      "num_tokens": 28699773.0,
      "step": 12865
    },
    {
      "entropy": 5.550434923171997,
      "epoch": 1.4454428033919244,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004801154382030124,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20406392514705657,
      "num_tokens": 28710971.0,
      "step": 12870
    },
    {
      "entropy": 5.477548456192016,
      "epoch": 1.446004380299882,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048009893281906004,
      "loss": 5.0843,
      "mean_token_accuracy": 0.20875932425260543,
      "num_tokens": 28721544.0,
      "step": 12875
    },
    {
      "entropy": 5.538469457626343,
      "epoch": 1.4465659572078395,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048008242090467803,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19580515772104262,
      "num_tokens": 28733790.0,
      "step": 12880
    },
    {
      "entropy": 5.577961540222168,
      "epoch": 1.4471275341157972,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004800659024603923,
      "loss": 5.2004,
      "mean_token_accuracy": 0.2073975071310997,
      "num_tokens": 28745218.0,
      "step": 12885
    },
    {
      "entropy": 5.562515020370483,
      "epoch": 1.4476891110237546,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048004937748672864,
      "loss": 5.2528,
      "mean_token_accuracy": 0.20050956010818483,
      "num_tokens": 28758061.0,
      "step": 12890
    },
    {
      "entropy": 5.422244215011597,
      "epoch": 1.4482506879317123,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048003284598421327,
      "loss": 5.0686,
      "mean_token_accuracy": 0.21108701676130295,
      "num_tokens": 28770037.0,
      "step": 12895
    },
    {
      "entropy": 5.511967134475708,
      "epoch": 1.4488122648396697,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004800163079533726,
      "loss": 5.223,
      "mean_token_accuracy": 0.1976005792617798,
      "num_tokens": 28781052.0,
      "step": 12900
    },
    {
      "entropy": 5.401247119903564,
      "epoch": 1.4493738417476274,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004799997633947332,
      "loss": 5.1005,
      "mean_token_accuracy": 0.19951864182949067,
      "num_tokens": 28791225.0,
      "step": 12905
    },
    {
      "entropy": 5.509648370742798,
      "epoch": 1.4499354186555848,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047998321230882174,
      "loss": 5.1317,
      "mean_token_accuracy": 0.20951839983463288,
      "num_tokens": 28803022.0,
      "step": 12910
    },
    {
      "entropy": 5.489766597747803,
      "epoch": 1.4504969955635425,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047996665469616536,
      "loss": 5.1047,
      "mean_token_accuracy": 0.2046004757285118,
      "num_tokens": 28813789.0,
      "step": 12915
    },
    {
      "entropy": 5.551026296615601,
      "epoch": 1.4510585724715,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047995009055729114,
      "loss": 5.1522,
      "mean_token_accuracy": 0.20666670054197311,
      "num_tokens": 28825015.0,
      "step": 12920
    },
    {
      "entropy": 5.596477079391479,
      "epoch": 1.4516201493794574,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004799335198927266,
      "loss": 5.2803,
      "mean_token_accuracy": 0.20229267179965973,
      "num_tokens": 28835852.0,
      "step": 12925
    },
    {
      "entropy": 5.4911417961120605,
      "epoch": 1.452181726287415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004799169427029993,
      "loss": 5.1899,
      "mean_token_accuracy": 0.20372844338417054,
      "num_tokens": 28846684.0,
      "step": 12930
    },
    {
      "entropy": 5.495855808258057,
      "epoch": 1.4527433031953727,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047990035898863706,
      "loss": 5.1913,
      "mean_token_accuracy": 0.2003163680434227,
      "num_tokens": 28857402.0,
      "step": 12935
    },
    {
      "entropy": 5.434234714508056,
      "epoch": 1.4533048801033301,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004798837687501679,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2175763502717018,
      "num_tokens": 28867866.0,
      "step": 12940
    },
    {
      "entropy": 5.445178747177124,
      "epoch": 1.4538664570112876,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004798671719881201,
      "loss": 5.1411,
      "mean_token_accuracy": 0.20677025020122528,
      "num_tokens": 28878102.0,
      "step": 12945
    },
    {
      "entropy": 5.588583946228027,
      "epoch": 1.4544280339192452,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004798505687030221,
      "loss": 5.242,
      "mean_token_accuracy": 0.20081905126571656,
      "num_tokens": 28888981.0,
      "step": 12950
    },
    {
      "entropy": 5.442296981811523,
      "epoch": 1.454989610827203,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004798339588954025,
      "loss": 5.1065,
      "mean_token_accuracy": 0.2057306468486786,
      "num_tokens": 28899733.0,
      "step": 12955
    },
    {
      "entropy": 5.418082857131958,
      "epoch": 1.4555511877351603,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004798173425657903,
      "loss": 5.0946,
      "mean_token_accuracy": 0.20657345950603484,
      "num_tokens": 28910701.0,
      "step": 12960
    },
    {
      "entropy": 5.532093238830567,
      "epoch": 1.4561127646431178,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004798007197147144,
      "loss": 5.1761,
      "mean_token_accuracy": 0.20050276815891266,
      "num_tokens": 28922627.0,
      "step": 12965
    },
    {
      "entropy": 5.520030546188354,
      "epoch": 1.4566743415510754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004797840903427042,
      "loss": 5.22,
      "mean_token_accuracy": 0.20199471116065978,
      "num_tokens": 28933292.0,
      "step": 12970
    },
    {
      "entropy": 5.394319534301758,
      "epoch": 1.4572359184590329,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047976745445028906,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21009878814220428,
      "num_tokens": 28943898.0,
      "step": 12975
    },
    {
      "entropy": 5.460964202880859,
      "epoch": 1.4577974953669905,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047975081203799883,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20172940641641618,
      "num_tokens": 28955248.0,
      "step": 12980
    },
    {
      "entropy": 5.477160310745239,
      "epoch": 1.458359072274948,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047973416310636327,
      "loss": 5.0597,
      "mean_token_accuracy": 0.212305711209774,
      "num_tokens": 28966905.0,
      "step": 12985
    },
    {
      "entropy": 5.506209850311279,
      "epoch": 1.4589206491829056,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004797175076559126,
      "loss": 5.181,
      "mean_token_accuracy": 0.210268734395504,
      "num_tokens": 28978515.0,
      "step": 12990
    },
    {
      "entropy": 5.438519716262817,
      "epoch": 1.459482226090863,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047970084568717703,
      "loss": 5.165,
      "mean_token_accuracy": 0.19923397600650788,
      "num_tokens": 28990185.0,
      "step": 12995
    },
    {
      "entropy": 5.452360057830811,
      "epoch": 1.4600438029988208,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004796841772006872,
      "loss": 5.1906,
      "mean_token_accuracy": 0.20202868133783342,
      "num_tokens": 29000779.0,
      "step": 13000
    },
    {
      "entropy": 5.549176788330078,
      "epoch": 1.4606053799067782,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047966750219697376,
      "loss": 5.2552,
      "mean_token_accuracy": 0.20228233486413955,
      "num_tokens": 29012066.0,
      "step": 13005
    },
    {
      "entropy": 5.6248729705810545,
      "epoch": 1.4611669568147359,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004796508206765677,
      "loss": 5.2966,
      "mean_token_accuracy": 0.19542959481477737,
      "num_tokens": 29023416.0,
      "step": 13010
    },
    {
      "entropy": 5.503637075424194,
      "epoch": 1.4617285337226933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047963413264000005,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20476240664720535,
      "num_tokens": 29034469.0,
      "step": 13015
    },
    {
      "entropy": 5.453005456924439,
      "epoch": 1.4622901106306507,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004796174380878023,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20023581087589265,
      "num_tokens": 29045306.0,
      "step": 13020
    },
    {
      "entropy": 5.4390764236450195,
      "epoch": 1.4628516875386084,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000479600737020506,
      "loss": 5.0963,
      "mean_token_accuracy": 0.2090238317847252,
      "num_tokens": 29057114.0,
      "step": 13025
    },
    {
      "entropy": 5.4582654476165775,
      "epoch": 1.463413264446566,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004795840294386429,
      "loss": 5.0866,
      "mean_token_accuracy": 0.19895012229681014,
      "num_tokens": 29067250.0,
      "step": 13030
    },
    {
      "entropy": 5.445709133148194,
      "epoch": 1.4639748413545235,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047956731534274483,
      "loss": 5.1126,
      "mean_token_accuracy": 0.21051748394966124,
      "num_tokens": 29078582.0,
      "step": 13035
    },
    {
      "entropy": 5.4906774997711185,
      "epoch": 1.464536418262481,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047955059473334425,
      "loss": 5.1486,
      "mean_token_accuracy": 0.21105076223611832,
      "num_tokens": 29090113.0,
      "step": 13040
    },
    {
      "entropy": 5.427413702011108,
      "epoch": 1.4650979951704386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004795338676109733,
      "loss": 5.1464,
      "mean_token_accuracy": 0.2052920252084732,
      "num_tokens": 29101117.0,
      "step": 13045
    },
    {
      "entropy": 5.534099531173706,
      "epoch": 1.4656595720783963,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004795171339761647,
      "loss": 5.1587,
      "mean_token_accuracy": 0.20264400094747542,
      "num_tokens": 29112569.0,
      "step": 13050
    },
    {
      "entropy": 5.5404692649841305,
      "epoch": 1.4662211489863537,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047950039382945126,
      "loss": 5.2492,
      "mean_token_accuracy": 0.2044188603758812,
      "num_tokens": 29124462.0,
      "step": 13055
    },
    {
      "entropy": 5.4422135829925535,
      "epoch": 1.4667827258943111,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047948364717136587,
      "loss": 5.1189,
      "mean_token_accuracy": 0.2060721069574356,
      "num_tokens": 29135401.0,
      "step": 13060
    },
    {
      "entropy": 5.480085706710815,
      "epoch": 1.4673443028022688,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047946689400244196,
      "loss": 5.1336,
      "mean_token_accuracy": 0.2062096118927002,
      "num_tokens": 29146601.0,
      "step": 13065
    },
    {
      "entropy": 5.481954145431518,
      "epoch": 1.4679058797102262,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047945013432321284,
      "loss": 5.1265,
      "mean_token_accuracy": 0.2071948990225792,
      "num_tokens": 29157299.0,
      "step": 13070
    },
    {
      "entropy": 5.567254304885864,
      "epoch": 1.468467456618184,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047943336813421205,
      "loss": 5.2233,
      "mean_token_accuracy": 0.1984789177775383,
      "num_tokens": 29169101.0,
      "step": 13075
    },
    {
      "entropy": 5.378282737731934,
      "epoch": 1.4690290335261413,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004794165954359736,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21008046567440034,
      "num_tokens": 29181286.0,
      "step": 13080
    },
    {
      "entropy": 5.495277786254883,
      "epoch": 1.469590610434099,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004793998162290315,
      "loss": 5.226,
      "mean_token_accuracy": 0.201323439180851,
      "num_tokens": 29191181.0,
      "step": 13085
    },
    {
      "entropy": 5.417777824401855,
      "epoch": 1.4701521873420564,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004793830305139199,
      "loss": 5.0796,
      "mean_token_accuracy": 0.20867829918861389,
      "num_tokens": 29202426.0,
      "step": 13090
    },
    {
      "entropy": 5.3995689868927,
      "epoch": 1.470713764250014,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004793662382911734,
      "loss": 5.0877,
      "mean_token_accuracy": 0.2065662920475006,
      "num_tokens": 29212817.0,
      "step": 13095
    },
    {
      "entropy": 5.461744594573974,
      "epoch": 1.4712753411579715,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004793494395613265,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20269558131694793,
      "num_tokens": 29224644.0,
      "step": 13100
    },
    {
      "entropy": 5.506007528305053,
      "epoch": 1.4718369180659292,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004793326343249144,
      "loss": 5.1163,
      "mean_token_accuracy": 0.20634000301361083,
      "num_tokens": 29236443.0,
      "step": 13105
    },
    {
      "entropy": 5.540659093856812,
      "epoch": 1.4723984949738866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004793158225824718,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19870007932186126,
      "num_tokens": 29247687.0,
      "step": 13110
    },
    {
      "entropy": 5.53632287979126,
      "epoch": 1.4729600718818443,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047929900433453423,
      "loss": 5.1653,
      "mean_token_accuracy": 0.19968986213207246,
      "num_tokens": 29259739.0,
      "step": 13115
    },
    {
      "entropy": 5.510801124572754,
      "epoch": 1.4735216487898017,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047928217958163703,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20789009779691697,
      "num_tokens": 29270834.0,
      "step": 13120
    },
    {
      "entropy": 5.491941642761231,
      "epoch": 1.4740832256977594,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047926534832431615,
      "loss": 5.1455,
      "mean_token_accuracy": 0.2055303454399109,
      "num_tokens": 29282291.0,
      "step": 13125
    },
    {
      "entropy": 5.520731115341187,
      "epoch": 1.4746448026057168,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004792485105631073,
      "loss": 5.1965,
      "mean_token_accuracy": 0.1997177168726921,
      "num_tokens": 29293836.0,
      "step": 13130
    },
    {
      "entropy": 5.4909240245819095,
      "epoch": 1.4752063795136743,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004792316662985467,
      "loss": 5.2003,
      "mean_token_accuracy": 0.19660606533288955,
      "num_tokens": 29305550.0,
      "step": 13135
    },
    {
      "entropy": 5.523720121383667,
      "epoch": 1.475767956421632,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047921481553117053,
      "loss": 5.1836,
      "mean_token_accuracy": 0.2036956235766411,
      "num_tokens": 29316973.0,
      "step": 13140
    },
    {
      "entropy": 5.476486587524414,
      "epoch": 1.4763295333295896,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047919795826151544,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2100991800427437,
      "num_tokens": 29327743.0,
      "step": 13145
    },
    {
      "entropy": 5.42023868560791,
      "epoch": 1.476891110237547,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004791810944901183,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20388011634349823,
      "num_tokens": 29338286.0,
      "step": 13150
    },
    {
      "entropy": 5.438933229446411,
      "epoch": 1.4774526871455045,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004791642242175158,
      "loss": 5.203,
      "mean_token_accuracy": 0.19936581403017045,
      "num_tokens": 29348877.0,
      "step": 13155
    },
    {
      "entropy": 5.410045480728149,
      "epoch": 1.4780142640534621,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047914734744424517,
      "loss": 4.9674,
      "mean_token_accuracy": 0.21752348393201829,
      "num_tokens": 29360072.0,
      "step": 13160
    },
    {
      "entropy": 5.513800144195557,
      "epoch": 1.4785758409614196,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047913046417084393,
      "loss": 5.2137,
      "mean_token_accuracy": 0.20630625784397125,
      "num_tokens": 29370744.0,
      "step": 13165
    },
    {
      "entropy": 5.522347116470337,
      "epoch": 1.4791374178693772,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047911357439784947,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20243710428476333,
      "num_tokens": 29381463.0,
      "step": 13170
    },
    {
      "entropy": 5.500744342803955,
      "epoch": 1.4796989947773347,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047909667812579964,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20441703498363495,
      "num_tokens": 29394227.0,
      "step": 13175
    },
    {
      "entropy": 5.426384735107422,
      "epoch": 1.4802605716852923,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047907977535523253,
      "loss": 5.0986,
      "mean_token_accuracy": 0.20749824345111847,
      "num_tokens": 29404754.0,
      "step": 13180
    },
    {
      "entropy": 5.423990058898926,
      "epoch": 1.4808221485932498,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004790628660866861,
      "loss": 5.142,
      "mean_token_accuracy": 0.21275347769260405,
      "num_tokens": 29416161.0,
      "step": 13185
    },
    {
      "entropy": 5.449864149093628,
      "epoch": 1.4813837255012074,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004790459503206989,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21738106459379197,
      "num_tokens": 29426587.0,
      "step": 13190
    },
    {
      "entropy": 5.494637823104858,
      "epoch": 1.4819453024091649,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004790290280578095,
      "loss": 5.1899,
      "mean_token_accuracy": 0.1964538276195526,
      "num_tokens": 29437488.0,
      "step": 13195
    },
    {
      "entropy": 5.451122140884399,
      "epoch": 1.4825068793171226,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047901209929855675,
      "loss": 5.158,
      "mean_token_accuracy": 0.20798053443431855,
      "num_tokens": 29448244.0,
      "step": 13200
    },
    {
      "entropy": 5.548648357391357,
      "epoch": 1.48306845622508,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004789951640434796,
      "loss": 5.2446,
      "mean_token_accuracy": 0.20025039315223694,
      "num_tokens": 29458260.0,
      "step": 13205
    },
    {
      "entropy": 5.603910875320435,
      "epoch": 1.4836300331330377,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004789782222931173,
      "loss": 5.1978,
      "mean_token_accuracy": 0.20077601224184036,
      "num_tokens": 29470874.0,
      "step": 13210
    },
    {
      "entropy": 5.41795244216919,
      "epoch": 1.484191610040995,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047896127404800935,
      "loss": 5.0994,
      "mean_token_accuracy": 0.20298130810260773,
      "num_tokens": 29481442.0,
      "step": 13215
    },
    {
      "entropy": 5.549015760421753,
      "epoch": 1.4847531869489528,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004789443193086953,
      "loss": 5.2512,
      "mean_token_accuracy": 0.2030337229371071,
      "num_tokens": 29492727.0,
      "step": 13220
    },
    {
      "entropy": 5.57514796257019,
      "epoch": 1.4853147638569102,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047892735807571505,
      "loss": 5.1879,
      "mean_token_accuracy": 0.19736639857292176,
      "num_tokens": 29502729.0,
      "step": 13225
    },
    {
      "entropy": 5.5533716678619385,
      "epoch": 1.4858763407648676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004789103903496086,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20643337666988373,
      "num_tokens": 29514315.0,
      "step": 13230
    },
    {
      "entropy": 5.490795993804932,
      "epoch": 1.4864379176728253,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004788934161309161,
      "loss": 5.0835,
      "mean_token_accuracy": 0.21079244315624238,
      "num_tokens": 29526030.0,
      "step": 13235
    },
    {
      "entropy": 5.45865387916565,
      "epoch": 1.486999494580783,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047887643542017836,
      "loss": 5.1081,
      "mean_token_accuracy": 0.20463404506444932,
      "num_tokens": 29536972.0,
      "step": 13240
    },
    {
      "entropy": 5.406656122207641,
      "epoch": 1.4875610714887404,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047885944821793574,
      "loss": 5.1241,
      "mean_token_accuracy": 0.20604779273271562,
      "num_tokens": 29547325.0,
      "step": 13245
    },
    {
      "entropy": 5.5179516792297365,
      "epoch": 1.4881226483966978,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047884245452472926,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20583217293024064,
      "num_tokens": 29557944.0,
      "step": 13250
    },
    {
      "entropy": 5.429768323898315,
      "epoch": 1.4886842253046555,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004788254543410999,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20905335396528243,
      "num_tokens": 29568618.0,
      "step": 13255
    },
    {
      "entropy": 5.499443531036377,
      "epoch": 1.489245802212613,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047880844766758914,
      "loss": 5.1496,
      "mean_token_accuracy": 0.2003599688410759,
      "num_tokens": 29580682.0,
      "step": 13260
    },
    {
      "entropy": 5.5412647247314455,
      "epoch": 1.4898073791205706,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047879143450473824,
      "loss": 5.204,
      "mean_token_accuracy": 0.20136342644691468,
      "num_tokens": 29592302.0,
      "step": 13265
    },
    {
      "entropy": 5.554577350616455,
      "epoch": 1.490368956028528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047877441485308917,
      "loss": 5.205,
      "mean_token_accuracy": 0.2014056697487831,
      "num_tokens": 29603591.0,
      "step": 13270
    },
    {
      "entropy": 5.447190618515014,
      "epoch": 1.4909305329364857,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047875738871318356,
      "loss": 5.1682,
      "mean_token_accuracy": 0.20276699662208558,
      "num_tokens": 29615542.0,
      "step": 13275
    },
    {
      "entropy": 5.551174640655518,
      "epoch": 1.4914921098444431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004787403560855637,
      "loss": 5.1888,
      "mean_token_accuracy": 0.21355810165405273,
      "num_tokens": 29626650.0,
      "step": 13280
    },
    {
      "entropy": 5.468151187896728,
      "epoch": 1.4920536867524008,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004787233169707719,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20220291763544082,
      "num_tokens": 29637135.0,
      "step": 13285
    },
    {
      "entropy": 5.492263269424439,
      "epoch": 1.4926152636603582,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004787062713693507,
      "loss": 5.1789,
      "mean_token_accuracy": 0.20043986588716506,
      "num_tokens": 29648862.0,
      "step": 13290
    },
    {
      "entropy": 5.472868776321411,
      "epoch": 1.493176840568316,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047868921928184276,
      "loss": 5.1482,
      "mean_token_accuracy": 0.2028573378920555,
      "num_tokens": 29661015.0,
      "step": 13295
    },
    {
      "entropy": 5.58891339302063,
      "epoch": 1.4937384174762733,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004786721607087911,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1992519497871399,
      "num_tokens": 29673521.0,
      "step": 13300
    },
    {
      "entropy": 5.544138336181641,
      "epoch": 1.494299994384231,
      "grad_norm": 0.875,
      "learning_rate": 0.0004786550956507389,
      "loss": 5.2147,
      "mean_token_accuracy": 0.2113247260451317,
      "num_tokens": 29685379.0,
      "step": 13305
    },
    {
      "entropy": 5.488175249099731,
      "epoch": 1.4948615712921884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004786380241082294,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20022862702608107,
      "num_tokens": 29696372.0,
      "step": 13310
    },
    {
      "entropy": 5.500126600265503,
      "epoch": 1.495423148200146,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047862094608180623,
      "loss": 5.1431,
      "mean_token_accuracy": 0.20842529982328414,
      "num_tokens": 29707401.0,
      "step": 13315
    },
    {
      "entropy": 5.431816911697387,
      "epoch": 1.4959847251081035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004786038615720132,
      "loss": 5.1222,
      "mean_token_accuracy": 0.2076442390680313,
      "num_tokens": 29717524.0,
      "step": 13320
    },
    {
      "entropy": 5.511350870132446,
      "epoch": 1.496546302016061,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047858677057939427,
      "loss": 5.2424,
      "mean_token_accuracy": 0.19728194177150726,
      "num_tokens": 29728235.0,
      "step": 13325
    },
    {
      "entropy": 5.586638069152832,
      "epoch": 1.4971078789240186,
      "grad_norm": 1.0,
      "learning_rate": 0.00047856967310449356,
      "loss": 5.24,
      "mean_token_accuracy": 0.19504618942737578,
      "num_tokens": 29738800.0,
      "step": 13330
    },
    {
      "entropy": 5.555594253540039,
      "epoch": 1.4976694558319763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047855256914785556,
      "loss": 5.1673,
      "mean_token_accuracy": 0.20222549438476561,
      "num_tokens": 29750346.0,
      "step": 13335
    },
    {
      "entropy": 5.454142665863037,
      "epoch": 1.4982310327399337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047853545871002476,
      "loss": 5.0805,
      "mean_token_accuracy": 0.2097135826945305,
      "num_tokens": 29761931.0,
      "step": 13340
    },
    {
      "entropy": 5.391069126129151,
      "epoch": 1.4987926096478912,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047851834179154595,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2105358362197876,
      "num_tokens": 29773333.0,
      "step": 13345
    },
    {
      "entropy": 5.414756536483765,
      "epoch": 1.4993541865558488,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004785012183929643,
      "loss": 5.0653,
      "mean_token_accuracy": 0.212247596681118,
      "num_tokens": 29785125.0,
      "step": 13350
    },
    {
      "entropy": 5.435869598388672,
      "epoch": 1.4999157634638065,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004784840885148249,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20986656248569488,
      "num_tokens": 29796288.0,
      "step": 13355
    },
    {
      "entropy": 5.5364643096923825,
      "epoch": 1.500477340371764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004784669521576732,
      "loss": 5.1953,
      "mean_token_accuracy": 0.2044151559472084,
      "num_tokens": 29806348.0,
      "step": 13360
    },
    {
      "entropy": 5.5060771942138675,
      "epoch": 1.5010389172797214,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004784498093220549,
      "loss": 5.0903,
      "mean_token_accuracy": 0.2109503373503685,
      "num_tokens": 29816342.0,
      "step": 13365
    },
    {
      "entropy": 5.406218242645264,
      "epoch": 1.501600494187679,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004784326600085157,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20413004904985427,
      "num_tokens": 29826789.0,
      "step": 13370
    },
    {
      "entropy": 5.496536493301392,
      "epoch": 1.5021620710956367,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047841550421760164,
      "loss": 5.1868,
      "mean_token_accuracy": 0.20478349030017853,
      "num_tokens": 29837633.0,
      "step": 13375
    },
    {
      "entropy": 5.5469011783599855,
      "epoch": 1.502723648003594,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047839834194985916,
      "loss": 5.2329,
      "mean_token_accuracy": 0.20279349386692047,
      "num_tokens": 29848498.0,
      "step": 13380
    },
    {
      "entropy": 5.54252142906189,
      "epoch": 1.5032852249115516,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004783811732058345,
      "loss": 5.2057,
      "mean_token_accuracy": 0.20127368122339248,
      "num_tokens": 29859777.0,
      "step": 13385
    },
    {
      "entropy": 5.718990612030029,
      "epoch": 1.5038468018195092,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047836399798607446,
      "loss": 5.3868,
      "mean_token_accuracy": 0.19448302388191224,
      "num_tokens": 29873373.0,
      "step": 13390
    },
    {
      "entropy": 5.546049451828003,
      "epoch": 1.5044083787274667,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047834681629112583,
      "loss": 5.1472,
      "mean_token_accuracy": 0.2054397717118263,
      "num_tokens": 29884749.0,
      "step": 13395
    },
    {
      "entropy": 5.415380764007568,
      "epoch": 1.5049699556354241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004783296281215357,
      "loss": 5.0754,
      "mean_token_accuracy": 0.20518286973237992,
      "num_tokens": 29895059.0,
      "step": 13400
    },
    {
      "entropy": 5.392477560043335,
      "epoch": 1.5055315325433818,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004783124334778513,
      "loss": 5.1289,
      "mean_token_accuracy": 0.2021928146481514,
      "num_tokens": 29906611.0,
      "step": 13405
    },
    {
      "entropy": 5.370140790939331,
      "epoch": 1.5060931094513395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004782952323606202,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21246108114719392,
      "num_tokens": 29918179.0,
      "step": 13410
    },
    {
      "entropy": 5.509697294235229,
      "epoch": 1.506654686359297,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004782780247703901,
      "loss": 5.1691,
      "mean_token_accuracy": 0.20233879536390303,
      "num_tokens": 29929612.0,
      "step": 13415
    },
    {
      "entropy": 5.491402673721313,
      "epoch": 1.5072162632672543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004782608107077088,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21546249240636825,
      "num_tokens": 29939569.0,
      "step": 13420
    },
    {
      "entropy": 5.478633022308349,
      "epoch": 1.507777840175212,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047824359017312445,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20860920548439027,
      "num_tokens": 29950344.0,
      "step": 13425
    },
    {
      "entropy": 5.435837888717652,
      "epoch": 1.5083394170831697,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004782263631671854,
      "loss": 5.1028,
      "mean_token_accuracy": 0.2105194792151451,
      "num_tokens": 29960748.0,
      "step": 13430
    },
    {
      "entropy": 5.588251638412475,
      "epoch": 1.508900993991127,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047820912969044014,
      "loss": 5.285,
      "mean_token_accuracy": 0.1966715157032013,
      "num_tokens": 29971678.0,
      "step": 13435
    },
    {
      "entropy": 5.416446542739868,
      "epoch": 1.5094625708990845,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047819188974343734,
      "loss": 5.1156,
      "mean_token_accuracy": 0.20558972209692,
      "num_tokens": 29982477.0,
      "step": 13440
    },
    {
      "entropy": 5.444994974136352,
      "epoch": 1.5100241478070422,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004781746433267259,
      "loss": 5.161,
      "mean_token_accuracy": 0.21009932905435563,
      "num_tokens": 29994539.0,
      "step": 13445
    },
    {
      "entropy": 5.513435792922974,
      "epoch": 1.5105857247149999,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047815739044085515,
      "loss": 5.108,
      "mean_token_accuracy": 0.21252892911434174,
      "num_tokens": 30004821.0,
      "step": 13450
    },
    {
      "entropy": 5.609660911560058,
      "epoch": 1.5111473016229573,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047814013108637423,
      "loss": 5.1657,
      "mean_token_accuracy": 0.20085231214761734,
      "num_tokens": 30016343.0,
      "step": 13455
    },
    {
      "entropy": 5.454044580459595,
      "epoch": 1.5117088785309147,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047812286526383274,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21451688855886458,
      "num_tokens": 30027719.0,
      "step": 13460
    },
    {
      "entropy": 5.426101160049439,
      "epoch": 1.5122704554388724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004781055929737804,
      "loss": 5.0707,
      "mean_token_accuracy": 0.2172236144542694,
      "num_tokens": 30037988.0,
      "step": 13465
    },
    {
      "entropy": 5.4582665920257565,
      "epoch": 1.51283203234683,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047808831421676724,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21147372126579284,
      "num_tokens": 30049736.0,
      "step": 13470
    },
    {
      "entropy": 5.537223529815674,
      "epoch": 1.5133936092547875,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004780710289933435,
      "loss": 5.2069,
      "mean_token_accuracy": 0.20294972658157348,
      "num_tokens": 30061963.0,
      "step": 13475
    },
    {
      "entropy": 5.569507360458374,
      "epoch": 1.513955186162745,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004780537373040593,
      "loss": 5.1057,
      "mean_token_accuracy": 0.21471622735261917,
      "num_tokens": 30072256.0,
      "step": 13480
    },
    {
      "entropy": 5.460702514648437,
      "epoch": 1.5145167630707026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047803643914946537,
      "loss": 5.1706,
      "mean_token_accuracy": 0.20563367158174514,
      "num_tokens": 30084804.0,
      "step": 13485
    },
    {
      "entropy": 5.479845857620239,
      "epoch": 1.51507833997866,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004780191345301125,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20428748577833175,
      "num_tokens": 30096938.0,
      "step": 13490
    },
    {
      "entropy": 5.505037069320679,
      "epoch": 1.5156399168866175,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047800182344655165,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20337493419647218,
      "num_tokens": 30108233.0,
      "step": 13495
    },
    {
      "entropy": 5.50295672416687,
      "epoch": 1.5162014937945751,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000477984505899334,
      "loss": 5.14,
      "mean_token_accuracy": 0.21044730097055436,
      "num_tokens": 30120246.0,
      "step": 13500
    },
    {
      "entropy": 5.458343076705932,
      "epoch": 1.5167630707025328,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047796718188901095,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21335820704698563,
      "num_tokens": 30131213.0,
      "step": 13505
    },
    {
      "entropy": 5.397097158432007,
      "epoch": 1.5173246476104902,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047794985141613413,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2153217688202858,
      "num_tokens": 30143300.0,
      "step": 13510
    },
    {
      "entropy": 5.437649822235107,
      "epoch": 1.5178862245184477,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047793251448125526,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20531620532274247,
      "num_tokens": 30153794.0,
      "step": 13515
    },
    {
      "entropy": 5.500084114074707,
      "epoch": 1.5184478014264053,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004779151710849265,
      "loss": 5.2161,
      "mean_token_accuracy": 0.19907034784555436,
      "num_tokens": 30165152.0,
      "step": 13520
    },
    {
      "entropy": 5.4851343631744385,
      "epoch": 1.519009378334363,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004778978212276999,
      "loss": 5.1327,
      "mean_token_accuracy": 0.20664205253124238,
      "num_tokens": 30176338.0,
      "step": 13525
    },
    {
      "entropy": 5.464123392105103,
      "epoch": 1.5195709552423204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004778804649101281,
      "loss": 5.1704,
      "mean_token_accuracy": 0.20037916749715806,
      "num_tokens": 30187666.0,
      "step": 13530
    },
    {
      "entropy": 5.48260989189148,
      "epoch": 1.5201325321502779,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004778631021327636,
      "loss": 5.2133,
      "mean_token_accuracy": 0.19810021072626113,
      "num_tokens": 30198892.0,
      "step": 13535
    },
    {
      "entropy": 5.453015661239624,
      "epoch": 1.5206941090582355,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004778457328961592,
      "loss": 5.1627,
      "mean_token_accuracy": 0.1998189389705658,
      "num_tokens": 30210677.0,
      "step": 13540
    },
    {
      "entropy": 5.519055509567261,
      "epoch": 1.5212556859661932,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047782835720086797,
      "loss": 5.0944,
      "mean_token_accuracy": 0.2124049633741379,
      "num_tokens": 30221217.0,
      "step": 13545
    },
    {
      "entropy": 5.560549926757813,
      "epoch": 1.5218172628741506,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004778109750474432,
      "loss": 5.236,
      "mean_token_accuracy": 0.20007369220256804,
      "num_tokens": 30233576.0,
      "step": 13550
    },
    {
      "entropy": 5.461468458175659,
      "epoch": 1.522378839782108,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004777935864364383,
      "loss": 5.0853,
      "mean_token_accuracy": 0.20346295982599258,
      "num_tokens": 30244787.0,
      "step": 13555
    },
    {
      "entropy": 5.4790794372558596,
      "epoch": 1.5229404166900657,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000477776191368407,
      "loss": 5.1388,
      "mean_token_accuracy": 0.20858580768108367,
      "num_tokens": 30256656.0,
      "step": 13560
    },
    {
      "entropy": 5.558883762359619,
      "epoch": 1.5235019935980234,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004777587898439031,
      "loss": 5.2076,
      "mean_token_accuracy": 0.20844527930021287,
      "num_tokens": 30267715.0,
      "step": 13565
    },
    {
      "entropy": 5.556528377532959,
      "epoch": 1.5240635705059808,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047774138186348073,
      "loss": 5.1963,
      "mean_token_accuracy": 0.2015288919210434,
      "num_tokens": 30278119.0,
      "step": 13570
    },
    {
      "entropy": 5.525978851318359,
      "epoch": 1.5246251474139383,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004777239674276941,
      "loss": 5.1146,
      "mean_token_accuracy": 0.20381217449903488,
      "num_tokens": 30289899.0,
      "step": 13575
    },
    {
      "entropy": 5.430830812454223,
      "epoch": 1.525186724321896,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004777065465370977,
      "loss": 5.1271,
      "mean_token_accuracy": 0.20787773430347442,
      "num_tokens": 30300799.0,
      "step": 13580
    },
    {
      "entropy": 5.5626091957092285,
      "epoch": 1.5257483012298534,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004776891191922463,
      "loss": 5.3249,
      "mean_token_accuracy": 0.20186955183744432,
      "num_tokens": 30312955.0,
      "step": 13585
    },
    {
      "entropy": 5.537404870986938,
      "epoch": 1.5263098781378108,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004776716853936947,
      "loss": 5.1053,
      "mean_token_accuracy": 0.205371630191803,
      "num_tokens": 30323089.0,
      "step": 13590
    },
    {
      "entropy": 5.489454984664917,
      "epoch": 1.5268714550457685,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000477654245141998,
      "loss": 5.0512,
      "mean_token_accuracy": 0.21588983833789827,
      "num_tokens": 30333926.0,
      "step": 13595
    },
    {
      "entropy": 5.4234906196594235,
      "epoch": 1.5274330319537262,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004776367984377116,
      "loss": 5.1617,
      "mean_token_accuracy": 0.20579976886510848,
      "num_tokens": 30346347.0,
      "step": 13600
    },
    {
      "entropy": 5.4525346755981445,
      "epoch": 1.5279946088616836,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047761934528139086,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1997246116399765,
      "num_tokens": 30357757.0,
      "step": 13605
    },
    {
      "entropy": 5.596718072891235,
      "epoch": 1.528556185769641,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004776018856735916,
      "loss": 5.1881,
      "mean_token_accuracy": 0.2084172859787941,
      "num_tokens": 30369323.0,
      "step": 13610
    },
    {
      "entropy": 5.464363765716553,
      "epoch": 1.5291177626775987,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004775844196148698,
      "loss": 5.0557,
      "mean_token_accuracy": 0.21078469157218932,
      "num_tokens": 30380335.0,
      "step": 13615
    },
    {
      "entropy": 5.402562236785888,
      "epoch": 1.5296793395855564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047756694710578146,
      "loss": 5.0762,
      "mean_token_accuracy": 0.2053290605545044,
      "num_tokens": 30391338.0,
      "step": 13620
    },
    {
      "entropy": 5.493469095230102,
      "epoch": 1.5302409164935138,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000477549468146883,
      "loss": 5.1995,
      "mean_token_accuracy": 0.1947053089737892,
      "num_tokens": 30402917.0,
      "step": 13625
    },
    {
      "entropy": 5.439848947525024,
      "epoch": 1.5308024934014712,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004775319827387309,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21650879383087157,
      "num_tokens": 30412690.0,
      "step": 13630
    },
    {
      "entropy": 5.448499393463135,
      "epoch": 1.531364070309429,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004775144908818818,
      "loss": 5.1884,
      "mean_token_accuracy": 0.19989157170057298,
      "num_tokens": 30423020.0,
      "step": 13635
    },
    {
      "entropy": 5.581812238693237,
      "epoch": 1.5319256472173866,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047749699257689296,
      "loss": 5.3771,
      "mean_token_accuracy": 0.18942070752382278,
      "num_tokens": 30434039.0,
      "step": 13640
    },
    {
      "entropy": 5.563929796218872,
      "epoch": 1.532487224125344,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004774794878243212,
      "loss": 5.1627,
      "mean_token_accuracy": 0.20131921023130417,
      "num_tokens": 30446743.0,
      "step": 13645
    },
    {
      "entropy": 5.518055868148804,
      "epoch": 1.5330488010333014,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000477461976624724,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20493233948946,
      "num_tokens": 30456969.0,
      "step": 13650
    },
    {
      "entropy": 5.398975658416748,
      "epoch": 1.533610377941259,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047744445897865905,
      "loss": 5.2021,
      "mean_token_accuracy": 0.19790499955415725,
      "num_tokens": 30467497.0,
      "step": 13655
    },
    {
      "entropy": 5.478356122970581,
      "epoch": 1.5341719548492168,
      "grad_norm": 0.921875,
      "learning_rate": 0.000477426934886684,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20651963502168655,
      "num_tokens": 30479428.0,
      "step": 13660
    },
    {
      "entropy": 5.526005125045776,
      "epoch": 1.5347335317571742,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047740940434935677,
      "loss": 5.1111,
      "mean_token_accuracy": 0.20874049216508866,
      "num_tokens": 30490193.0,
      "step": 13665
    },
    {
      "entropy": 5.522428846359253,
      "epoch": 1.5352951086651316,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004773918673672356,
      "loss": 5.1722,
      "mean_token_accuracy": 0.20612497478723527,
      "num_tokens": 30500311.0,
      "step": 13670
    },
    {
      "entropy": 5.3809489727020265,
      "epoch": 1.5358566855730893,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047737432394087887,
      "loss": 5.0778,
      "mean_token_accuracy": 0.21350422352552414,
      "num_tokens": 30512485.0,
      "step": 13675
    },
    {
      "entropy": 5.45035457611084,
      "epoch": 1.5364182624810467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047735677407084525,
      "loss": 5.0719,
      "mean_token_accuracy": 0.20744231790304185,
      "num_tokens": 30523495.0,
      "step": 13680
    },
    {
      "entropy": 5.458536672592163,
      "epoch": 1.5369798393890042,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004773392177576934,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20961540043354035,
      "num_tokens": 30533964.0,
      "step": 13685
    },
    {
      "entropy": 5.402696990966797,
      "epoch": 1.5375414162969618,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004773216550019824,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21492202281951905,
      "num_tokens": 30544798.0,
      "step": 13690
    },
    {
      "entropy": 5.47240047454834,
      "epoch": 1.5381029932049195,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047730408580427136,
      "loss": 5.2052,
      "mean_token_accuracy": 0.19995270669460297,
      "num_tokens": 30555991.0,
      "step": 13695
    },
    {
      "entropy": 5.499682283401489,
      "epoch": 1.538664570112877,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047728651016511994,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21119318455457686,
      "num_tokens": 30566848.0,
      "step": 13700
    },
    {
      "entropy": 5.491256618499756,
      "epoch": 1.5392261470208344,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047726892808508744,
      "loss": 5.07,
      "mean_token_accuracy": 0.21484928727149963,
      "num_tokens": 30577361.0,
      "step": 13705
    },
    {
      "entropy": 5.452073669433593,
      "epoch": 1.539787723928792,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004772513395647339,
      "loss": 5.1236,
      "mean_token_accuracy": 0.20703460276126862,
      "num_tokens": 30587685.0,
      "step": 13710
    },
    {
      "entropy": 5.447957944869995,
      "epoch": 1.5403493008367497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004772337446046192,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20866415947675704,
      "num_tokens": 30599864.0,
      "step": 13715
    },
    {
      "entropy": 5.489260244369507,
      "epoch": 1.5409108777447071,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047721614320530366,
      "loss": 5.1403,
      "mean_token_accuracy": 0.20871280878782272,
      "num_tokens": 30610885.0,
      "step": 13720
    },
    {
      "entropy": 5.468125247955323,
      "epoch": 1.5414724546526646,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047719853536734764,
      "loss": 5.1157,
      "mean_token_accuracy": 0.203450009226799,
      "num_tokens": 30622537.0,
      "step": 13725
    },
    {
      "entropy": 5.542632532119751,
      "epoch": 1.5420340315606222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004771809210913119,
      "loss": 5.253,
      "mean_token_accuracy": 0.19639279693365097,
      "num_tokens": 30635045.0,
      "step": 13730
    },
    {
      "entropy": 5.470077896118164,
      "epoch": 1.54259560846858,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047716330037775724,
      "loss": 5.0929,
      "mean_token_accuracy": 0.19972793608903885,
      "num_tokens": 30645998.0,
      "step": 13735
    },
    {
      "entropy": 5.455017995834351,
      "epoch": 1.5431571853765373,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004771456732272446,
      "loss": 5.137,
      "mean_token_accuracy": 0.20381448417901993,
      "num_tokens": 30658744.0,
      "step": 13740
    },
    {
      "entropy": 5.393810892105103,
      "epoch": 1.5437187622844948,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047712803964033545,
      "loss": 5.0143,
      "mean_token_accuracy": 0.21443181931972505,
      "num_tokens": 30669420.0,
      "step": 13745
    },
    {
      "entropy": 5.391561222076416,
      "epoch": 1.5442803391924524,
      "grad_norm": 1.0,
      "learning_rate": 0.00047711039961759107,
      "loss": 5.0946,
      "mean_token_accuracy": 0.2059645727276802,
      "num_tokens": 30680037.0,
      "step": 13750
    },
    {
      "entropy": 5.534412050247193,
      "epoch": 1.54484191610041,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004770927531595731,
      "loss": 5.1367,
      "mean_token_accuracy": 0.21158429682254792,
      "num_tokens": 30691650.0,
      "step": 13755
    },
    {
      "entropy": 5.451357793807984,
      "epoch": 1.5454034930083675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047707510026684353,
      "loss": 5.1424,
      "mean_token_accuracy": 0.20709216445684434,
      "num_tokens": 30701876.0,
      "step": 13760
    },
    {
      "entropy": 5.616465425491333,
      "epoch": 1.545965069916325,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047705744093996447,
      "loss": 5.3349,
      "mean_token_accuracy": 0.20009353905916213,
      "num_tokens": 30714484.0,
      "step": 13765
    },
    {
      "entropy": 5.50636305809021,
      "epoch": 1.5465266468242826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477039775179498,
      "loss": 5.1581,
      "mean_token_accuracy": 0.2003595307469368,
      "num_tokens": 30725185.0,
      "step": 13770
    },
    {
      "entropy": 5.472897386550903,
      "epoch": 1.54708822373224,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047702210298600684,
      "loss": 5.1346,
      "mean_token_accuracy": 0.2099891945719719,
      "num_tokens": 30735522.0,
      "step": 13775
    },
    {
      "entropy": 5.4250202655792235,
      "epoch": 1.5476498006401975,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047700442436005356,
      "loss": 5.1414,
      "mean_token_accuracy": 0.21014324426651002,
      "num_tokens": 30746353.0,
      "step": 13780
    },
    {
      "entropy": 5.479766178131103,
      "epoch": 1.5482113775481552,
      "grad_norm": 0.953125,
      "learning_rate": 0.000476986739302201,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20711538791656495,
      "num_tokens": 30756245.0,
      "step": 13785
    },
    {
      "entropy": 5.510843181610108,
      "epoch": 1.5487729544561128,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047696904781301233,
      "loss": 5.2278,
      "mean_token_accuracy": 0.20635064989328383,
      "num_tokens": 30767387.0,
      "step": 13790
    },
    {
      "entropy": 5.470912456512451,
      "epoch": 1.5493345313640703,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004769513498930508,
      "loss": 5.1819,
      "mean_token_accuracy": 0.21047525703907013,
      "num_tokens": 30780140.0,
      "step": 13795
    },
    {
      "entropy": 5.483899545669556,
      "epoch": 1.5498961082720277,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047693364554288,
      "loss": 5.0921,
      "mean_token_accuracy": 0.21013399809598923,
      "num_tokens": 30791137.0,
      "step": 13800
    },
    {
      "entropy": 5.4625725746154785,
      "epoch": 1.5504576851799854,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047691593476306373,
      "loss": 5.0913,
      "mean_token_accuracy": 0.2136094644665718,
      "num_tokens": 30801708.0,
      "step": 13805
    },
    {
      "entropy": 5.513564729690552,
      "epoch": 1.551019262087943,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004768982175541656,
      "loss": 5.1699,
      "mean_token_accuracy": 0.204835344851017,
      "num_tokens": 30813519.0,
      "step": 13810
    },
    {
      "entropy": 5.532550096511841,
      "epoch": 1.5515808389959005,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047688049391675006,
      "loss": 5.2338,
      "mean_token_accuracy": 0.20518066585063935,
      "num_tokens": 30823786.0,
      "step": 13815
    },
    {
      "entropy": 5.473483276367188,
      "epoch": 1.552142415903858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004768627638513813,
      "loss": 5.1036,
      "mean_token_accuracy": 0.21366893500089645,
      "num_tokens": 30834381.0,
      "step": 13820
    },
    {
      "entropy": 5.475982522964477,
      "epoch": 1.5527039928118156,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004768450273586237,
      "loss": 5.1069,
      "mean_token_accuracy": 0.2051657423377037,
      "num_tokens": 30844865.0,
      "step": 13825
    },
    {
      "entropy": 5.5269266128540036,
      "epoch": 1.5532655697197733,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047682728443904223,
      "loss": 5.1935,
      "mean_token_accuracy": 0.20764564573764802,
      "num_tokens": 30857088.0,
      "step": 13830
    },
    {
      "entropy": 5.444524192810059,
      "epoch": 1.5538271466277307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047680953509320176,
      "loss": 5.2268,
      "mean_token_accuracy": 0.19674571305513383,
      "num_tokens": 30868320.0,
      "step": 13835
    },
    {
      "entropy": 5.4890465259552,
      "epoch": 1.5543887235356881,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004767917793216674,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20868585258722305,
      "num_tokens": 30878598.0,
      "step": 13840
    },
    {
      "entropy": 5.474894428253174,
      "epoch": 1.5549503004436458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047677401712500455,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20976149886846543,
      "num_tokens": 30889254.0,
      "step": 13845
    },
    {
      "entropy": 5.519720125198364,
      "epoch": 1.5555118773516035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004767562485037787,
      "loss": 5.2016,
      "mean_token_accuracy": 0.20380555242300033,
      "num_tokens": 30901164.0,
      "step": 13850
    },
    {
      "entropy": 5.4564048767089846,
      "epoch": 1.556073454259561,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047673847345855566,
      "loss": 5.0943,
      "mean_token_accuracy": 0.20823153853416443,
      "num_tokens": 30911828.0,
      "step": 13855
    },
    {
      "entropy": 5.472623825073242,
      "epoch": 1.5566350311675183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004767206919899014,
      "loss": 4.971,
      "mean_token_accuracy": 0.2212482586503029,
      "num_tokens": 30922050.0,
      "step": 13860
    },
    {
      "entropy": 5.39045820236206,
      "epoch": 1.557196608075476,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000476702904098382,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21153525561094283,
      "num_tokens": 30932723.0,
      "step": 13865
    },
    {
      "entropy": 5.401630115509033,
      "epoch": 1.5577581849834334,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000476685109784564,
      "loss": 5.0649,
      "mean_token_accuracy": 0.21431580632925035,
      "num_tokens": 30943699.0,
      "step": 13870
    },
    {
      "entropy": 5.478217077255249,
      "epoch": 1.5583197618913909,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047666730904901377,
      "loss": 5.1425,
      "mean_token_accuracy": 0.20613633692264557,
      "num_tokens": 30954315.0,
      "step": 13875
    },
    {
      "entropy": 5.556294202804565,
      "epoch": 1.5588813387993485,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047664950189229833,
      "loss": 5.307,
      "mean_token_accuracy": 0.20023447126150132,
      "num_tokens": 30965840.0,
      "step": 13880
    },
    {
      "entropy": 5.567817831039429,
      "epoch": 1.5594429157073062,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004766316883149845,
      "loss": 5.2152,
      "mean_token_accuracy": 0.20502734780311585,
      "num_tokens": 30977051.0,
      "step": 13885
    },
    {
      "entropy": 5.43372950553894,
      "epoch": 1.5600044926152636,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047661386831763955,
      "loss": 5.0772,
      "mean_token_accuracy": 0.20623415857553482,
      "num_tokens": 30988730.0,
      "step": 13890
    },
    {
      "entropy": 5.4321348667144775,
      "epoch": 1.560566069523221,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047659604190083076,
      "loss": 5.0807,
      "mean_token_accuracy": 0.21048348546028137,
      "num_tokens": 31000571.0,
      "step": 13895
    },
    {
      "entropy": 5.4009092330932615,
      "epoch": 1.5611276464311787,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047657820906512585,
      "loss": 5.1008,
      "mean_token_accuracy": 0.21363420635461808,
      "num_tokens": 31011003.0,
      "step": 13900
    },
    {
      "entropy": 5.512424945831299,
      "epoch": 1.5616892233391364,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047656036981109256,
      "loss": 5.1044,
      "mean_token_accuracy": 0.2100786566734314,
      "num_tokens": 31022510.0,
      "step": 13905
    },
    {
      "entropy": 5.541799592971802,
      "epoch": 1.5622508002470938,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047654252413929896,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20794539600610734,
      "num_tokens": 31033149.0,
      "step": 13910
    },
    {
      "entropy": 5.52243390083313,
      "epoch": 1.5628123771550513,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047652467205031317,
      "loss": 5.2056,
      "mean_token_accuracy": 0.2090357467532158,
      "num_tokens": 31043626.0,
      "step": 13915
    },
    {
      "entropy": 5.539959287643432,
      "epoch": 1.563373954063009,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047650681354470363,
      "loss": 5.2478,
      "mean_token_accuracy": 0.2010113626718521,
      "num_tokens": 31055083.0,
      "step": 13920
    },
    {
      "entropy": 5.461220693588257,
      "epoch": 1.5639355309709666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047648894862303907,
      "loss": 5.0798,
      "mean_token_accuracy": 0.20491784065961838,
      "num_tokens": 31066763.0,
      "step": 13925
    },
    {
      "entropy": 5.473897457122803,
      "epoch": 1.564497107878924,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047647107728588826,
      "loss": 5.1596,
      "mean_token_accuracy": 0.20189221799373627,
      "num_tokens": 31078722.0,
      "step": 13930
    },
    {
      "entropy": 5.517248916625976,
      "epoch": 1.5650586847868815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047645319953382024,
      "loss": 5.1976,
      "mean_token_accuracy": 0.20537666529417037,
      "num_tokens": 31090207.0,
      "step": 13935
    },
    {
      "entropy": 5.496640539169311,
      "epoch": 1.5656202616948391,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047643531536740405,
      "loss": 5.2245,
      "mean_token_accuracy": 0.2039627730846405,
      "num_tokens": 31101943.0,
      "step": 13940
    },
    {
      "entropy": 5.481431484222412,
      "epoch": 1.5661818386027968,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004764174247872094,
      "loss": 5.2176,
      "mean_token_accuracy": 0.1974351242184639,
      "num_tokens": 31115032.0,
      "step": 13945
    },
    {
      "entropy": 5.553815650939941,
      "epoch": 1.5667434155107542,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047639952779380586,
      "loss": 5.1286,
      "mean_token_accuracy": 0.20633062720298767,
      "num_tokens": 31124965.0,
      "step": 13950
    },
    {
      "entropy": 5.527206802368164,
      "epoch": 1.5673049924187117,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004763816243877632,
      "loss": 5.1571,
      "mean_token_accuracy": 0.20625796616077424,
      "num_tokens": 31136692.0,
      "step": 13955
    },
    {
      "entropy": 5.490589666366577,
      "epoch": 1.5678665693266693,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004763637145696515,
      "loss": 5.11,
      "mean_token_accuracy": 0.20351085513830186,
      "num_tokens": 31148352.0,
      "step": 13960
    },
    {
      "entropy": 5.433886909484864,
      "epoch": 1.5684281462346268,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000476345798340041,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20816990435123445,
      "num_tokens": 31159892.0,
      "step": 13965
    },
    {
      "entropy": 5.5542994976043705,
      "epoch": 1.5689897231425842,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004763278756995022,
      "loss": 5.2357,
      "mean_token_accuracy": 0.20216581970453262,
      "num_tokens": 31170087.0,
      "step": 13970
    },
    {
      "entropy": 5.358810758590698,
      "epoch": 1.5695513000505419,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047630994664860576,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2090377017855644,
      "num_tokens": 31180687.0,
      "step": 13975
    },
    {
      "entropy": 5.513645839691162,
      "epoch": 1.5701128769584995,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004762920111879226,
      "loss": 5.2213,
      "mean_token_accuracy": 0.20003449618816377,
      "num_tokens": 31192513.0,
      "step": 13980
    },
    {
      "entropy": 5.50188980102539,
      "epoch": 1.570674453866457,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004762740693180237,
      "loss": 5.1193,
      "mean_token_accuracy": 0.2016259953379631,
      "num_tokens": 31203432.0,
      "step": 13985
    },
    {
      "entropy": 5.546356630325318,
      "epoch": 1.5712360307744144,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047625612103948035,
      "loss": 5.0958,
      "mean_token_accuracy": 0.2152412474155426,
      "num_tokens": 31214474.0,
      "step": 13990
    },
    {
      "entropy": 5.491615152359008,
      "epoch": 1.571797607682372,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004762381663528641,
      "loss": 5.1592,
      "mean_token_accuracy": 0.19755446761846543,
      "num_tokens": 31226489.0,
      "step": 13995
    },
    {
      "entropy": 5.383515691757202,
      "epoch": 1.5723591845903298,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004762202052587466,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20448143184185028,
      "num_tokens": 31237568.0,
      "step": 14000
    },
    {
      "entropy": 5.45446081161499,
      "epoch": 1.5729207614982872,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004762022377576996,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2189904808998108,
      "num_tokens": 31248587.0,
      "step": 14005
    },
    {
      "entropy": 5.511194324493408,
      "epoch": 1.5734823384062446,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004761842638502954,
      "loss": 5.177,
      "mean_token_accuracy": 0.1997183457016945,
      "num_tokens": 31259513.0,
      "step": 14010
    },
    {
      "entropy": 5.502513027191162,
      "epoch": 1.5740439153142023,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004761662835371062,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20309208929538727,
      "num_tokens": 31270984.0,
      "step": 14015
    },
    {
      "entropy": 5.439129066467285,
      "epoch": 1.57460549222216,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004761482968187045,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20797574520111084,
      "num_tokens": 31281907.0,
      "step": 14020
    },
    {
      "entropy": 5.403236055374146,
      "epoch": 1.5751670691301174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004761303036956631,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2057667925953865,
      "num_tokens": 31292903.0,
      "step": 14025
    },
    {
      "entropy": 5.526501417160034,
      "epoch": 1.5757286460380748,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004761123041685547,
      "loss": 5.1396,
      "mean_token_accuracy": 0.2085891082882881,
      "num_tokens": 31304185.0,
      "step": 14030
    },
    {
      "entropy": 5.473230457305908,
      "epoch": 1.5762902229460325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004760942982379526,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20867830067873,
      "num_tokens": 31314806.0,
      "step": 14035
    },
    {
      "entropy": 5.431953430175781,
      "epoch": 1.5768517998539902,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047607628590443004,
      "loss": 5.1661,
      "mean_token_accuracy": 0.20455605983734132,
      "num_tokens": 31326361.0,
      "step": 14040
    },
    {
      "entropy": 5.557656908035279,
      "epoch": 1.5774133767619476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047605826716856056,
      "loss": 5.141,
      "mean_token_accuracy": 0.20891171991825103,
      "num_tokens": 31336549.0,
      "step": 14045
    },
    {
      "entropy": 5.586280488967896,
      "epoch": 1.577974953669905,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004760402420309178,
      "loss": 5.2496,
      "mean_token_accuracy": 0.20191466957330703,
      "num_tokens": 31346994.0,
      "step": 14050
    },
    {
      "entropy": 5.502039051055908,
      "epoch": 1.5785365305778627,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004760222104920759,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20491855293512345,
      "num_tokens": 31358449.0,
      "step": 14055
    },
    {
      "entropy": 5.539928007125854,
      "epoch": 1.5790981074858201,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004760041725526088,
      "loss": 5.1835,
      "mean_token_accuracy": 0.20377593040466307,
      "num_tokens": 31368964.0,
      "step": 14060
    },
    {
      "entropy": 5.522278738021851,
      "epoch": 1.5796596843937776,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047598612821309077,
      "loss": 5.119,
      "mean_token_accuracy": 0.20832561254501342,
      "num_tokens": 31379388.0,
      "step": 14065
    },
    {
      "entropy": 5.444357109069824,
      "epoch": 1.5802212613017352,
      "grad_norm": 0.875,
      "learning_rate": 0.0004759680774740966,
      "loss": 5.1753,
      "mean_token_accuracy": 0.2067069321870804,
      "num_tokens": 31390757.0,
      "step": 14070
    },
    {
      "entropy": 5.436584568023681,
      "epoch": 1.580782838209693,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004759500203362008,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20713399946689606,
      "num_tokens": 31402309.0,
      "step": 14075
    },
    {
      "entropy": 5.53011794090271,
      "epoch": 1.5813444151176503,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004759319567999785,
      "loss": 5.1291,
      "mean_token_accuracy": 0.2070261210203171,
      "num_tokens": 31412559.0,
      "step": 14080
    },
    {
      "entropy": 5.49569354057312,
      "epoch": 1.5819059920256078,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004759138868660048,
      "loss": 5.2658,
      "mean_token_accuracy": 0.19862870872020721,
      "num_tokens": 31424982.0,
      "step": 14085
    },
    {
      "entropy": 5.440700101852417,
      "epoch": 1.5824675689335654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047589581053485494,
      "loss": 5.0118,
      "mean_token_accuracy": 0.20904837250709535,
      "num_tokens": 31434472.0,
      "step": 14090
    },
    {
      "entropy": 5.514515829086304,
      "epoch": 1.583029145841523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047587772780710455,
      "loss": 5.2239,
      "mean_token_accuracy": 0.20741783380508422,
      "num_tokens": 31445458.0,
      "step": 14095
    },
    {
      "entropy": 5.476921319961548,
      "epoch": 1.5835907227494805,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047585963868332945,
      "loss": 5.1038,
      "mean_token_accuracy": 0.2124158337712288,
      "num_tokens": 31455851.0,
      "step": 14100
    },
    {
      "entropy": 5.396849822998047,
      "epoch": 1.584152299657438,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004758415431641055,
      "loss": 5.0794,
      "mean_token_accuracy": 0.20829991549253463,
      "num_tokens": 31466279.0,
      "step": 14105
    },
    {
      "entropy": 5.529507398605347,
      "epoch": 1.5847138765653956,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004758234412500089,
      "loss": 5.1302,
      "mean_token_accuracy": 0.21033648997545243,
      "num_tokens": 31477933.0,
      "step": 14110
    },
    {
      "entropy": 5.535827016830444,
      "epoch": 1.5852754534733533,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047580533294161616,
      "loss": 5.1668,
      "mean_token_accuracy": 0.1991063192486763,
      "num_tokens": 31487640.0,
      "step": 14115
    },
    {
      "entropy": 5.569490623474121,
      "epoch": 1.5858370303813107,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004757872182395036,
      "loss": 5.1151,
      "mean_token_accuracy": 0.21425263881683348,
      "num_tokens": 31499059.0,
      "step": 14120
    },
    {
      "entropy": 5.525071239471435,
      "epoch": 1.5863986072892682,
      "grad_norm": 1.0,
      "learning_rate": 0.00047576909714424823,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20575840026140213,
      "num_tokens": 31510264.0,
      "step": 14125
    },
    {
      "entropy": 5.459005546569824,
      "epoch": 1.5869601841972258,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047575096965642687,
      "loss": 5.1201,
      "mean_token_accuracy": 0.20768028646707534,
      "num_tokens": 31520873.0,
      "step": 14130
    },
    {
      "entropy": 5.450346946716309,
      "epoch": 1.5875217611051835,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047573283577661684,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20521388202905655,
      "num_tokens": 31533836.0,
      "step": 14135
    },
    {
      "entropy": 5.463673448562622,
      "epoch": 1.588083338013141,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004757146955053955,
      "loss": 5.1787,
      "mean_token_accuracy": 0.20482372343540192,
      "num_tokens": 31545091.0,
      "step": 14140
    },
    {
      "entropy": 5.491623497009277,
      "epoch": 1.5886449149210984,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047569654884334036,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2070576786994934,
      "num_tokens": 31557139.0,
      "step": 14145
    },
    {
      "entropy": 5.522586679458618,
      "epoch": 1.589206491829056,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004756783957910292,
      "loss": 5.1498,
      "mean_token_accuracy": 0.20851533263921737,
      "num_tokens": 31567928.0,
      "step": 14150
    },
    {
      "entropy": 5.428273630142212,
      "epoch": 1.5897680687370135,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047566023634904015,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20564700663089752,
      "num_tokens": 31578081.0,
      "step": 14155
    },
    {
      "entropy": 5.421238803863526,
      "epoch": 1.590329645644971,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004756420705179513,
      "loss": 5.062,
      "mean_token_accuracy": 0.20919679403305053,
      "num_tokens": 31589648.0,
      "step": 14160
    },
    {
      "entropy": 5.480953645706177,
      "epoch": 1.5908912225529286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004756238982983412,
      "loss": 5.1463,
      "mean_token_accuracy": 0.203220696747303,
      "num_tokens": 31601085.0,
      "step": 14165
    },
    {
      "entropy": 5.486854982376099,
      "epoch": 1.5914527994608862,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004756057196907883,
      "loss": 5.201,
      "mean_token_accuracy": 0.20855775177478791,
      "num_tokens": 31612005.0,
      "step": 14170
    },
    {
      "entropy": 5.491432428359985,
      "epoch": 1.5920143763688437,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047558753469587154,
      "loss": 5.151,
      "mean_token_accuracy": 0.20160801112651824,
      "num_tokens": 31623427.0,
      "step": 14175
    },
    {
      "entropy": 5.632724189758301,
      "epoch": 1.5925759532768011,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047556934331416986,
      "loss": 5.3044,
      "mean_token_accuracy": 0.1924811854958534,
      "num_tokens": 31635086.0,
      "step": 14180
    },
    {
      "entropy": 5.565679883956909,
      "epoch": 1.5931375301847588,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004755511455462624,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20850237756967543,
      "num_tokens": 31646707.0,
      "step": 14185
    },
    {
      "entropy": 5.409745359420777,
      "epoch": 1.5936991070927164,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004755329413927287,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21165819168090821,
      "num_tokens": 31657327.0,
      "step": 14190
    },
    {
      "entropy": 5.461916923522949,
      "epoch": 1.5942606840006739,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004755147308541485,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20918234586715698,
      "num_tokens": 31668075.0,
      "step": 14195
    },
    {
      "entropy": 5.511706829071045,
      "epoch": 1.5948222609086313,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047549651393110145,
      "loss": 5.1757,
      "mean_token_accuracy": 0.19853594303131103,
      "num_tokens": 31680270.0,
      "step": 14200
    },
    {
      "entropy": 5.413037061691284,
      "epoch": 1.595383837816589,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047547829062416757,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21229496747255325,
      "num_tokens": 31691271.0,
      "step": 14205
    },
    {
      "entropy": 5.405726146697998,
      "epoch": 1.5959454147245467,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047546006093392726,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2144663840532303,
      "num_tokens": 31702982.0,
      "step": 14210
    },
    {
      "entropy": 5.468664169311523,
      "epoch": 1.596506991632504,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004754418248609608,
      "loss": 5.1565,
      "mean_token_accuracy": 0.2058590829372406,
      "num_tokens": 31715129.0,
      "step": 14215
    },
    {
      "entropy": 5.534321737289429,
      "epoch": 1.5970685685404615,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004754235824058489,
      "loss": 5.2237,
      "mean_token_accuracy": 0.20576290041208267,
      "num_tokens": 31726392.0,
      "step": 14220
    },
    {
      "entropy": 5.548418998718262,
      "epoch": 1.5976301454484192,
      "grad_norm": 1.0,
      "learning_rate": 0.0004754053335691724,
      "loss": 5.1866,
      "mean_token_accuracy": 0.2114136591553688,
      "num_tokens": 31738143.0,
      "step": 14225
    },
    {
      "entropy": 5.521101713180542,
      "epoch": 1.5981917223563769,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047538707835151237,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20609493404626847,
      "num_tokens": 31748510.0,
      "step": 14230
    },
    {
      "entropy": 5.372695684432983,
      "epoch": 1.5987532992643343,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047536881675345,
      "loss": 5.033,
      "mean_token_accuracy": 0.21007345020771026,
      "num_tokens": 31759378.0,
      "step": 14235
    },
    {
      "entropy": 5.586735010147095,
      "epoch": 1.5993148761722917,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004753505487755669,
      "loss": 5.2702,
      "mean_token_accuracy": 0.19648339450359345,
      "num_tokens": 31771627.0,
      "step": 14240
    },
    {
      "entropy": 5.464278268814087,
      "epoch": 1.5998764530802494,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047533227441844453,
      "loss": 5.0696,
      "mean_token_accuracy": 0.2142323836684227,
      "num_tokens": 31781990.0,
      "step": 14245
    },
    {
      "entropy": 5.4982294082641605,
      "epoch": 1.6004380299882068,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047531399368266485,
      "loss": 5.1313,
      "mean_token_accuracy": 0.2076670154929161,
      "num_tokens": 31792144.0,
      "step": 14250
    },
    {
      "entropy": 5.515278148651123,
      "epoch": 1.6009996068961643,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004752957065688099,
      "loss": 5.1827,
      "mean_token_accuracy": 0.20987604707479476,
      "num_tokens": 31804022.0,
      "step": 14255
    },
    {
      "entropy": 5.555851173400879,
      "epoch": 1.601561183804122,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000475277413077462,
      "loss": 5.2049,
      "mean_token_accuracy": 0.20333612263202666,
      "num_tokens": 31815108.0,
      "step": 14260
    },
    {
      "entropy": 5.429860401153564,
      "epoch": 1.6021227607120796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047525911320920363,
      "loss": 5.1719,
      "mean_token_accuracy": 0.20856576412916183,
      "num_tokens": 31825727.0,
      "step": 14265
    },
    {
      "entropy": 5.461634683609009,
      "epoch": 1.602684337620037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004752408069646174,
      "loss": 5.1202,
      "mean_token_accuracy": 0.21572770327329635,
      "num_tokens": 31836168.0,
      "step": 14270
    },
    {
      "entropy": 5.564251327514649,
      "epoch": 1.6032459145279945,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047522249434428623,
      "loss": 5.2198,
      "mean_token_accuracy": 0.2012143164873123,
      "num_tokens": 31847383.0,
      "step": 14275
    },
    {
      "entropy": 5.437561988830566,
      "epoch": 1.6038074914359521,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004752041753487931,
      "loss": 5.0445,
      "mean_token_accuracy": 0.2169059172272682,
      "num_tokens": 31858679.0,
      "step": 14280
    },
    {
      "entropy": 5.471265840530395,
      "epoch": 1.6043690683439098,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004751858499787215,
      "loss": 5.125,
      "mean_token_accuracy": 0.21637834310531617,
      "num_tokens": 31870195.0,
      "step": 14285
    },
    {
      "entropy": 5.55965633392334,
      "epoch": 1.6049306452518672,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004751675182346547,
      "loss": 5.1501,
      "mean_token_accuracy": 0.20644088089466095,
      "num_tokens": 31881175.0,
      "step": 14290
    },
    {
      "entropy": 5.449311542510986,
      "epoch": 1.6054922221598247,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047514918011717656,
      "loss": 5.0828,
      "mean_token_accuracy": 0.21454456746578215,
      "num_tokens": 31891199.0,
      "step": 14295
    },
    {
      "entropy": 5.5836633205413815,
      "epoch": 1.6060537990677823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047513083562687085,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19913967102766036,
      "num_tokens": 31903379.0,
      "step": 14300
    },
    {
      "entropy": 5.524294042587281,
      "epoch": 1.60661537597574,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004751124847643217,
      "loss": 5.0143,
      "mean_token_accuracy": 0.22051845937967302,
      "num_tokens": 31913958.0,
      "step": 14305
    },
    {
      "entropy": 5.398765325546265,
      "epoch": 1.6071769528836974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004750941275301135,
      "loss": 5.024,
      "mean_token_accuracy": 0.21078978329896927,
      "num_tokens": 31924547.0,
      "step": 14310
    },
    {
      "entropy": 5.536068439483643,
      "epoch": 1.6077385297916549,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004750757639248306,
      "loss": 5.2457,
      "mean_token_accuracy": 0.20039306133985518,
      "num_tokens": 31936987.0,
      "step": 14315
    },
    {
      "entropy": 5.473997306823731,
      "epoch": 1.6083001066996125,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047505739394905776,
      "loss": 5.1007,
      "mean_token_accuracy": 0.2050543561577797,
      "num_tokens": 31948165.0,
      "step": 14320
    },
    {
      "entropy": 5.52739839553833,
      "epoch": 1.6088616836075702,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047503901760338003,
      "loss": 5.1664,
      "mean_token_accuracy": 0.20757828950881957,
      "num_tokens": 31959206.0,
      "step": 14325
    },
    {
      "entropy": 5.507641983032227,
      "epoch": 1.6094232605155276,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047502063488838225,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20667380094528198,
      "num_tokens": 31970804.0,
      "step": 14330
    },
    {
      "entropy": 5.43414273262024,
      "epoch": 1.609984837423485,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047500224580464994,
      "loss": 5.0654,
      "mean_token_accuracy": 0.20945317447185516,
      "num_tokens": 31982961.0,
      "step": 14335
    },
    {
      "entropy": 5.493216896057129,
      "epoch": 1.6105464143314427,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047498385035276856,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20586385130882262,
      "num_tokens": 31993812.0,
      "step": 14340
    },
    {
      "entropy": 5.494396209716797,
      "epoch": 1.6111079912394002,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004749654485333238,
      "loss": 5.1722,
      "mean_token_accuracy": 0.20913249552249907,
      "num_tokens": 32004688.0,
      "step": 14345
    },
    {
      "entropy": 5.4740653991699215,
      "epoch": 1.6116695681473576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047494704034690163,
      "loss": 5.0515,
      "mean_token_accuracy": 0.20601202696561813,
      "num_tokens": 32014653.0,
      "step": 14350
    },
    {
      "entropy": 5.444654703140259,
      "epoch": 1.6122311450553153,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047492862579408816,
      "loss": 5.1168,
      "mean_token_accuracy": 0.2042867660522461,
      "num_tokens": 32025638.0,
      "step": 14355
    },
    {
      "entropy": 5.356553554534912,
      "epoch": 1.612792721963273,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047491020487546966,
      "loss": 5.0363,
      "mean_token_accuracy": 0.2104514479637146,
      "num_tokens": 32036316.0,
      "step": 14360
    },
    {
      "entropy": 5.453169488906861,
      "epoch": 1.6133542988712304,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004748917775916327,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21129222810268403,
      "num_tokens": 32047355.0,
      "step": 14365
    },
    {
      "entropy": 5.488507652282715,
      "epoch": 1.6139158757791878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047487334394316405,
      "loss": 5.1651,
      "mean_token_accuracy": 0.19742818176746368,
      "num_tokens": 32058897.0,
      "step": 14370
    },
    {
      "entropy": 5.5156721591949465,
      "epoch": 1.6144774526871455,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004748549039306506,
      "loss": 5.1539,
      "mean_token_accuracy": 0.2039370894432068,
      "num_tokens": 32069932.0,
      "step": 14375
    },
    {
      "entropy": 5.47552490234375,
      "epoch": 1.6150390295951031,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004748364575546795,
      "loss": 5.0533,
      "mean_token_accuracy": 0.2112406700849533,
      "num_tokens": 32081359.0,
      "step": 14380
    },
    {
      "entropy": 5.414346408843994,
      "epoch": 1.6156006065030606,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047481800481583806,
      "loss": 5.0265,
      "mean_token_accuracy": 0.20902552604675292,
      "num_tokens": 32093050.0,
      "step": 14385
    },
    {
      "entropy": 5.539455127716065,
      "epoch": 1.616162183411018,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004747995457147139,
      "loss": 5.1766,
      "mean_token_accuracy": 0.2029839888215065,
      "num_tokens": 32104623.0,
      "step": 14390
    },
    {
      "entropy": 5.440048742294311,
      "epoch": 1.6167237603189757,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047478108025189467,
      "loss": 5.1205,
      "mean_token_accuracy": 0.19632880985736847,
      "num_tokens": 32115845.0,
      "step": 14395
    },
    {
      "entropy": 5.460477352142334,
      "epoch": 1.6172853372269334,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047476260842796837,
      "loss": 5.084,
      "mean_token_accuracy": 0.20479002594947815,
      "num_tokens": 32127066.0,
      "step": 14400
    },
    {
      "entropy": 5.519058799743652,
      "epoch": 1.6178469141348908,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004747441302435231,
      "loss": 5.2198,
      "mean_token_accuracy": 0.201866115629673,
      "num_tokens": 32138344.0,
      "step": 14405
    },
    {
      "entropy": 5.5276977062225345,
      "epoch": 1.6184084910428482,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047472564569914726,
      "loss": 5.1703,
      "mean_token_accuracy": 0.198178967833519,
      "num_tokens": 32149054.0,
      "step": 14410
    },
    {
      "entropy": 5.483215618133545,
      "epoch": 1.618970067950806,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047470715479542945,
      "loss": 5.0635,
      "mean_token_accuracy": 0.20180138796567917,
      "num_tokens": 32160385.0,
      "step": 14415
    },
    {
      "entropy": 5.532918262481689,
      "epoch": 1.6195316448587636,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004746886575329584,
      "loss": 5.2201,
      "mean_token_accuracy": 0.20075257271528243,
      "num_tokens": 32171856.0,
      "step": 14420
    },
    {
      "entropy": 5.467215824127197,
      "epoch": 1.620093221766721,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047467015391232305,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20465515404939652,
      "num_tokens": 32182690.0,
      "step": 14425
    },
    {
      "entropy": 5.472300720214844,
      "epoch": 1.6206547986746784,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047465164393411245,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20631269216537476,
      "num_tokens": 32194045.0,
      "step": 14430
    },
    {
      "entropy": 5.468170261383056,
      "epoch": 1.621216375582636,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047463312759891615,
      "loss": 5.187,
      "mean_token_accuracy": 0.19822951257228852,
      "num_tokens": 32205699.0,
      "step": 14435
    },
    {
      "entropy": 5.553538513183594,
      "epoch": 1.6217779524905935,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004746146049073236,
      "loss": 5.1583,
      "mean_token_accuracy": 0.20397819727659225,
      "num_tokens": 32217974.0,
      "step": 14440
    },
    {
      "entropy": 5.385631847381592,
      "epoch": 1.622339529398551,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047459607585992466,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20507001280784606,
      "num_tokens": 32228723.0,
      "step": 14445
    },
    {
      "entropy": 5.47898588180542,
      "epoch": 1.6229011063065086,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047457754045730923,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20475900620222093,
      "num_tokens": 32241048.0,
      "step": 14450
    },
    {
      "entropy": 5.438021421432495,
      "epoch": 1.6234626832144663,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047455899870006745,
      "loss": 5.0878,
      "mean_token_accuracy": 0.21227079778909683,
      "num_tokens": 32251255.0,
      "step": 14455
    },
    {
      "entropy": 5.448459005355835,
      "epoch": 1.6240242601224237,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004745404505887899,
      "loss": 5.1182,
      "mean_token_accuracy": 0.20899780392646788,
      "num_tokens": 32262954.0,
      "step": 14460
    },
    {
      "entropy": 5.468794298171997,
      "epoch": 1.6245858370303812,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004745218961240669,
      "loss": 5.1327,
      "mean_token_accuracy": 0.20763630270957947,
      "num_tokens": 32275445.0,
      "step": 14465
    },
    {
      "entropy": 5.528776502609253,
      "epoch": 1.6251474139383388,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004745033353064893,
      "loss": 5.0874,
      "mean_token_accuracy": 0.2092886745929718,
      "num_tokens": 32285976.0,
      "step": 14470
    },
    {
      "entropy": 5.417572832107544,
      "epoch": 1.6257089908462965,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004744847681366482,
      "loss": 5.059,
      "mean_token_accuracy": 0.21153971701860427,
      "num_tokens": 32296304.0,
      "step": 14475
    },
    {
      "entropy": 5.433072662353515,
      "epoch": 1.626270567754254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047446619461513466,
      "loss": 5.11,
      "mean_token_accuracy": 0.21218177825212478,
      "num_tokens": 32307085.0,
      "step": 14480
    },
    {
      "entropy": 5.4493917465209964,
      "epoch": 1.6268321446622114,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004744476147425402,
      "loss": 5.0408,
      "mean_token_accuracy": 0.22019558101892472,
      "num_tokens": 32316814.0,
      "step": 14485
    },
    {
      "entropy": 5.452572202682495,
      "epoch": 1.627393721570169,
      "grad_norm": 1.0,
      "learning_rate": 0.00047442902851945634,
      "loss": 5.1103,
      "mean_token_accuracy": 0.20507177859544753,
      "num_tokens": 32327820.0,
      "step": 14490
    },
    {
      "entropy": 5.57509217262268,
      "epoch": 1.6279552984781267,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047441043594647486,
      "loss": 5.1156,
      "mean_token_accuracy": 0.2107645094394684,
      "num_tokens": 32338185.0,
      "step": 14495
    },
    {
      "entropy": 5.426065397262573,
      "epoch": 1.6285168753860841,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047439183702418765,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21276569366455078,
      "num_tokens": 32348537.0,
      "step": 14500
    },
    {
      "entropy": 5.456048107147216,
      "epoch": 1.6290784522940416,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047437323175318716,
      "loss": 5.1686,
      "mean_token_accuracy": 0.2014290764927864,
      "num_tokens": 32359094.0,
      "step": 14505
    },
    {
      "entropy": 5.381970405578613,
      "epoch": 1.6296400292019992,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047435462013406554,
      "loss": 4.988,
      "mean_token_accuracy": 0.2203885406255722,
      "num_tokens": 32370811.0,
      "step": 14510
    },
    {
      "entropy": 5.5212396621704105,
      "epoch": 1.630201606109957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004743360021674155,
      "loss": 5.1747,
      "mean_token_accuracy": 0.20606162399053574,
      "num_tokens": 32381522.0,
      "step": 14515
    },
    {
      "entropy": 5.4549699306488035,
      "epoch": 1.6307631830179143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047431737785382996,
      "loss": 5.1576,
      "mean_token_accuracy": 0.20090150386095046,
      "num_tokens": 32393221.0,
      "step": 14520
    },
    {
      "entropy": 5.518042850494385,
      "epoch": 1.6313247599258718,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047429874719390165,
      "loss": 5.1159,
      "mean_token_accuracy": 0.19962020516395568,
      "num_tokens": 32404007.0,
      "step": 14525
    },
    {
      "entropy": 5.523869371414184,
      "epoch": 1.6318863368338294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047428011018822404,
      "loss": 5.2026,
      "mean_token_accuracy": 0.20119993686676024,
      "num_tokens": 32414318.0,
      "step": 14530
    },
    {
      "entropy": 5.4379371166229244,
      "epoch": 1.6324479137417869,
      "grad_norm": 1.125,
      "learning_rate": 0.00047426146683739043,
      "loss": 5.1242,
      "mean_token_accuracy": 0.21090569943189622,
      "num_tokens": 32425493.0,
      "step": 14535
    },
    {
      "entropy": 5.446884393692017,
      "epoch": 1.6330094906497443,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047424281714199436,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20999828428030015,
      "num_tokens": 32435765.0,
      "step": 14540
    },
    {
      "entropy": 5.5830206871032715,
      "epoch": 1.633571067557702,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047422416110262975,
      "loss": 5.1719,
      "mean_token_accuracy": 0.2110847383737564,
      "num_tokens": 32446478.0,
      "step": 14545
    },
    {
      "entropy": 5.459376525878906,
      "epoch": 1.6341326444656596,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047420549871989053,
      "loss": 5.1765,
      "mean_token_accuracy": 0.20551348179578782,
      "num_tokens": 32456866.0,
      "step": 14550
    },
    {
      "entropy": 5.509270286560058,
      "epoch": 1.634694221373617,
      "grad_norm": 0.921875,
      "learning_rate": 0.000474186829994371,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21828665137290953,
      "num_tokens": 32466973.0,
      "step": 14555
    },
    {
      "entropy": 5.561855888366699,
      "epoch": 1.6352557982815745,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047416815492666564,
      "loss": 5.2379,
      "mean_token_accuracy": 0.1974301591515541,
      "num_tokens": 32478118.0,
      "step": 14560
    },
    {
      "entropy": 5.433518648147583,
      "epoch": 1.6358173751895322,
      "grad_norm": 1.0,
      "learning_rate": 0.0004741494735173689,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21908658593893052,
      "num_tokens": 32488255.0,
      "step": 14565
    },
    {
      "entropy": 5.518524789810181,
      "epoch": 1.6363789520974898,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004741307857670757,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20911531299352645,
      "num_tokens": 32498685.0,
      "step": 14570
    },
    {
      "entropy": 5.505541801452637,
      "epoch": 1.6369405290054473,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047411209167638106,
      "loss": 5.1465,
      "mean_token_accuracy": 0.20214727371931077,
      "num_tokens": 32510457.0,
      "step": 14575
    },
    {
      "entropy": 5.4434216022491455,
      "epoch": 1.6375021059134047,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004740933912458801,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20837015956640242,
      "num_tokens": 32522495.0,
      "step": 14580
    },
    {
      "entropy": 5.575459003448486,
      "epoch": 1.6380636828213624,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047407468447616845,
      "loss": 5.2001,
      "mean_token_accuracy": 0.20501646101474763,
      "num_tokens": 32534891.0,
      "step": 14585
    },
    {
      "entropy": 5.505761051177979,
      "epoch": 1.63862525972932,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047405597136784166,
      "loss": 5.0948,
      "mean_token_accuracy": 0.21115048825740815,
      "num_tokens": 32545555.0,
      "step": 14590
    },
    {
      "entropy": 5.442655324935913,
      "epoch": 1.6391868366372775,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004740372519214955,
      "loss": 5.1061,
      "mean_token_accuracy": 0.20640807300806047,
      "num_tokens": 32556851.0,
      "step": 14595
    },
    {
      "entropy": 5.56503210067749,
      "epoch": 1.639748413545235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047401852613772604,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19699703603982927,
      "num_tokens": 32569028.0,
      "step": 14600
    },
    {
      "entropy": 5.634431552886963,
      "epoch": 1.6403099904531926,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004739997940171295,
      "loss": 5.2103,
      "mean_token_accuracy": 0.21065267771482468,
      "num_tokens": 32581510.0,
      "step": 14605
    },
    {
      "entropy": 5.445714139938355,
      "epoch": 1.6408715673611503,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047398105556030234,
      "loss": 5.0799,
      "mean_token_accuracy": 0.20650525987148285,
      "num_tokens": 32592914.0,
      "step": 14610
    },
    {
      "entropy": 5.408236789703369,
      "epoch": 1.6414331442691077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047396231076784123,
      "loss": 5.0375,
      "mean_token_accuracy": 0.21460332870483398,
      "num_tokens": 32603931.0,
      "step": 14615
    },
    {
      "entropy": 5.5724873542785645,
      "epoch": 1.6419947211770651,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004739435596403429,
      "loss": 5.1816,
      "mean_token_accuracy": 0.20773862004280091,
      "num_tokens": 32614508.0,
      "step": 14620
    },
    {
      "entropy": 5.441210651397705,
      "epoch": 1.6425562980850228,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004739248021784046,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21945724338293077,
      "num_tokens": 32625393.0,
      "step": 14625
    },
    {
      "entropy": 5.362435245513916,
      "epoch": 1.6431178749929805,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004739060383826234,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21119230538606643,
      "num_tokens": 32636645.0,
      "step": 14630
    },
    {
      "entropy": 5.5764477252960205,
      "epoch": 1.6436794519009377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047388726825359673,
      "loss": 5.295,
      "mean_token_accuracy": 0.20065956860780715,
      "num_tokens": 32648792.0,
      "step": 14635
    },
    {
      "entropy": 5.530441379547119,
      "epoch": 1.6442410288088953,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047386849179192235,
      "loss": 5.1187,
      "mean_token_accuracy": 0.21125935018062592,
      "num_tokens": 32658825.0,
      "step": 14640
    },
    {
      "entropy": 5.5780431747436525,
      "epoch": 1.644802605716853,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047384970899819813,
      "loss": 5.135,
      "mean_token_accuracy": 0.21473995745182037,
      "num_tokens": 32670141.0,
      "step": 14645
    },
    {
      "entropy": 5.327546691894531,
      "epoch": 1.6453641826248104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004738309198730219,
      "loss": 4.9854,
      "mean_token_accuracy": 0.22079149931669234,
      "num_tokens": 32681804.0,
      "step": 14650
    },
    {
      "entropy": 5.5684239864349365,
      "epoch": 1.6459257595327679,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047381212441699217,
      "loss": 5.2163,
      "mean_token_accuracy": 0.2018399015069008,
      "num_tokens": 32692550.0,
      "step": 14655
    },
    {
      "entropy": 5.482094860076904,
      "epoch": 1.6464873364407255,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004737933226307072,
      "loss": 5.1196,
      "mean_token_accuracy": 0.212128646671772,
      "num_tokens": 32703833.0,
      "step": 14660
    },
    {
      "entropy": 5.455341005325318,
      "epoch": 1.6470489133486832,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047377451451476586,
      "loss": 5.1378,
      "mean_token_accuracy": 0.20528425574302672,
      "num_tokens": 32713928.0,
      "step": 14665
    },
    {
      "entropy": 5.505896949768067,
      "epoch": 1.6476104902566406,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004737557000697668,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20720291286706924,
      "num_tokens": 32725393.0,
      "step": 14670
    },
    {
      "entropy": 5.544742441177368,
      "epoch": 1.648172067164598,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004737368792963092,
      "loss": 5.1695,
      "mean_token_accuracy": 0.20412420034408568,
      "num_tokens": 32736047.0,
      "step": 14675
    },
    {
      "entropy": 5.487040567398071,
      "epoch": 1.6487336440725557,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047371805219499223,
      "loss": 5.1883,
      "mean_token_accuracy": 0.2028344303369522,
      "num_tokens": 32747960.0,
      "step": 14680
    },
    {
      "entropy": 5.443016624450683,
      "epoch": 1.6492952209805134,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047369921876641547,
      "loss": 5.1007,
      "mean_token_accuracy": 0.21083234548568724,
      "num_tokens": 32757851.0,
      "step": 14685
    },
    {
      "entropy": 5.464973783493042,
      "epoch": 1.6498567978884708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047368037901117854,
      "loss": 5.1258,
      "mean_token_accuracy": 0.21135618388652802,
      "num_tokens": 32767317.0,
      "step": 14690
    },
    {
      "entropy": 5.561215448379516,
      "epoch": 1.6504183747964283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004736615329298812,
      "loss": 5.1582,
      "mean_token_accuracy": 0.20844806581735612,
      "num_tokens": 32777729.0,
      "step": 14695
    },
    {
      "entropy": 5.513482284545899,
      "epoch": 1.650979951704386,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047364268052312367,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19741781651973725,
      "num_tokens": 32789547.0,
      "step": 14700
    },
    {
      "entropy": 5.489184713363647,
      "epoch": 1.6515415286123436,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004736238217915061,
      "loss": 5.1094,
      "mean_token_accuracy": 0.21159690171480178,
      "num_tokens": 32801487.0,
      "step": 14705
    },
    {
      "entropy": 5.5596295356750485,
      "epoch": 1.652103105520301,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047360495673562906,
      "loss": 5.1585,
      "mean_token_accuracy": 0.21734675467014314,
      "num_tokens": 32813042.0,
      "step": 14710
    },
    {
      "entropy": 5.549701070785522,
      "epoch": 1.6526646824282585,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004735860853560931,
      "loss": 5.1549,
      "mean_token_accuracy": 0.20712797939777375,
      "num_tokens": 32825467.0,
      "step": 14715
    },
    {
      "entropy": 5.480236768722534,
      "epoch": 1.6532262593362161,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047356720765349926,
      "loss": 5.1207,
      "mean_token_accuracy": 0.20850823819637299,
      "num_tokens": 32837113.0,
      "step": 14720
    },
    {
      "entropy": 5.577610826492309,
      "epoch": 1.6537878362441738,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004735483236284485,
      "loss": 5.2291,
      "mean_token_accuracy": 0.19821407943964003,
      "num_tokens": 32849779.0,
      "step": 14725
    },
    {
      "entropy": 5.445672798156738,
      "epoch": 1.654349413152131,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047352943328154203,
      "loss": 5.1422,
      "mean_token_accuracy": 0.206576469540596,
      "num_tokens": 32861626.0,
      "step": 14730
    },
    {
      "entropy": 5.512395715713501,
      "epoch": 1.6549109900600887,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047351053661338145,
      "loss": 5.1701,
      "mean_token_accuracy": 0.2074434831738472,
      "num_tokens": 32872875.0,
      "step": 14735
    },
    {
      "entropy": 5.477881383895874,
      "epoch": 1.6554725669680463,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004734916336245684,
      "loss": 5.0152,
      "mean_token_accuracy": 0.2107502579689026,
      "num_tokens": 32883899.0,
      "step": 14740
    },
    {
      "entropy": 5.416962766647339,
      "epoch": 1.6560341438760038,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004734727243157047,
      "loss": 5.0027,
      "mean_token_accuracy": 0.2155066102743149,
      "num_tokens": 32894255.0,
      "step": 14745
    },
    {
      "entropy": 5.502555274963379,
      "epoch": 1.6565957207839612,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047345380868739253,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20663869827985765,
      "num_tokens": 32903822.0,
      "step": 14750
    },
    {
      "entropy": 5.472291707992554,
      "epoch": 1.6571572976919189,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004734348867402341,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21425169110298156,
      "num_tokens": 32915879.0,
      "step": 14755
    },
    {
      "entropy": 5.410977840423584,
      "epoch": 1.6577188745998765,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004734159584748319,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20558079779148103,
      "num_tokens": 32928161.0,
      "step": 14760
    },
    {
      "entropy": 5.4737951278686525,
      "epoch": 1.658280451507834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047339702389178874,
      "loss": 5.0848,
      "mean_token_accuracy": 0.2081681475043297,
      "num_tokens": 32938399.0,
      "step": 14765
    },
    {
      "entropy": 5.451405191421509,
      "epoch": 1.6588420284157914,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047337808299170725,
      "loss": 5.0538,
      "mean_token_accuracy": 0.212621808052063,
      "num_tokens": 32948221.0,
      "step": 14770
    },
    {
      "entropy": 5.457606172561645,
      "epoch": 1.659403605323749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047335913577519075,
      "loss": 5.1153,
      "mean_token_accuracy": 0.21022116243839264,
      "num_tokens": 32959802.0,
      "step": 14775
    },
    {
      "entropy": 5.516530227661133,
      "epoch": 1.6599651822317067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047334018224284244,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20651769042015075,
      "num_tokens": 32970222.0,
      "step": 14780
    },
    {
      "entropy": 5.5132067680358885,
      "epoch": 1.6605267591396642,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004733212223952657,
      "loss": 5.0944,
      "mean_token_accuracy": 0.21669638454914092,
      "num_tokens": 32982241.0,
      "step": 14785
    },
    {
      "entropy": 5.460661935806274,
      "epoch": 1.6610883360476216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047330225623306443,
      "loss": 5.0388,
      "mean_token_accuracy": 0.21247430443763732,
      "num_tokens": 32992584.0,
      "step": 14790
    },
    {
      "entropy": 5.524060487747192,
      "epoch": 1.6616499129555793,
      "grad_norm": 1.0,
      "learning_rate": 0.00047328328375684237,
      "loss": 5.2286,
      "mean_token_accuracy": 0.20400873124599456,
      "num_tokens": 33004080.0,
      "step": 14795
    },
    {
      "entropy": 5.422363233566284,
      "epoch": 1.662211489863537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004732643049672036,
      "loss": 5.0597,
      "mean_token_accuracy": 0.21065025478601457,
      "num_tokens": 33015290.0,
      "step": 14800
    },
    {
      "entropy": 5.457514429092408,
      "epoch": 1.6627730667714944,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004732453198647525,
      "loss": 5.0394,
      "mean_token_accuracy": 0.2088946133852005,
      "num_tokens": 33026470.0,
      "step": 14805
    },
    {
      "entropy": 5.411339044570923,
      "epoch": 1.6633346436794518,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004732263284500936,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21732402443885804,
      "num_tokens": 33036766.0,
      "step": 14810
    },
    {
      "entropy": 5.417842006683349,
      "epoch": 1.6638962205874095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004732073307238314,
      "loss": 5.0562,
      "mean_token_accuracy": 0.2039196088910103,
      "num_tokens": 33047675.0,
      "step": 14815
    },
    {
      "entropy": 5.4877245903015135,
      "epoch": 1.6644577974953672,
      "grad_norm": 0.9375,
      "learning_rate": 0.000473188326686571,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21342210322618485,
      "num_tokens": 33057613.0,
      "step": 14820
    },
    {
      "entropy": 5.581669855117798,
      "epoch": 1.6650193744033244,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047316931633891736,
      "loss": 5.2239,
      "mean_token_accuracy": 0.199906887114048,
      "num_tokens": 33070498.0,
      "step": 14825
    },
    {
      "entropy": 5.472095394134522,
      "epoch": 1.665580951311282,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004731502996814759,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21086811274290085,
      "num_tokens": 33082026.0,
      "step": 14830
    },
    {
      "entropy": 5.475387144088745,
      "epoch": 1.6661425282192397,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047313127671485194,
      "loss": 5.0862,
      "mean_token_accuracy": 0.2051032692193985,
      "num_tokens": 33093841.0,
      "step": 14835
    },
    {
      "entropy": 5.537317705154419,
      "epoch": 1.6667041051271971,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004731122474396514,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20523856729269027,
      "num_tokens": 33103817.0,
      "step": 14840
    },
    {
      "entropy": 5.527795314788818,
      "epoch": 1.6672656820351546,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047309321185648,
      "loss": 5.1592,
      "mean_token_accuracy": 0.209054858982563,
      "num_tokens": 33116006.0,
      "step": 14845
    },
    {
      "entropy": 5.461910247802734,
      "epoch": 1.6678272589431122,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047307416996594386,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20389262586832047,
      "num_tokens": 33127649.0,
      "step": 14850
    },
    {
      "entropy": 5.378023672103882,
      "epoch": 1.66838883585107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004730551217686495,
      "loss": 4.9897,
      "mean_token_accuracy": 0.21935630738735198,
      "num_tokens": 33138748.0,
      "step": 14855
    },
    {
      "entropy": 5.53667516708374,
      "epoch": 1.6689504127590273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004730360672652031,
      "loss": 5.1761,
      "mean_token_accuracy": 0.20561397671699524,
      "num_tokens": 33150504.0,
      "step": 14860
    },
    {
      "entropy": 5.4520057201385494,
      "epoch": 1.6695119896669848,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047301700645621154,
      "loss": 5.1981,
      "mean_token_accuracy": 0.2012033224105835,
      "num_tokens": 33160625.0,
      "step": 14865
    },
    {
      "entropy": 5.50270266532898,
      "epoch": 1.6700735665749424,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004729979393422817,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2151162937283516,
      "num_tokens": 33172229.0,
      "step": 14870
    },
    {
      "entropy": 5.465377712249756,
      "epoch": 1.6706351434829,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004729788659240206,
      "loss": 5.1113,
      "mean_token_accuracy": 0.2084126889705658,
      "num_tokens": 33183234.0,
      "step": 14875
    },
    {
      "entropy": 5.494789981842041,
      "epoch": 1.6711967203908575,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004729597862020358,
      "loss": 5.1581,
      "mean_token_accuracy": 0.2057404786348343,
      "num_tokens": 33193976.0,
      "step": 14880
    },
    {
      "entropy": 5.493765211105346,
      "epoch": 1.671758297298815,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047294070017693453,
      "loss": 5.1168,
      "mean_token_accuracy": 0.21784327626228334,
      "num_tokens": 33205959.0,
      "step": 14885
    },
    {
      "entropy": 5.457241582870483,
      "epoch": 1.6723198742067726,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047292160784932454,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21431664675474166,
      "num_tokens": 33218405.0,
      "step": 14890
    },
    {
      "entropy": 5.443207406997681,
      "epoch": 1.6728814511147303,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047290250921981386,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20916522890329362,
      "num_tokens": 33230778.0,
      "step": 14895
    },
    {
      "entropy": 5.459868669509888,
      "epoch": 1.6734430280226877,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047288340428901053,
      "loss": 5.0599,
      "mean_token_accuracy": 0.209931942820549,
      "num_tokens": 33242069.0,
      "step": 14900
    },
    {
      "entropy": 5.477905321121216,
      "epoch": 1.6740046049306452,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047286429305752283,
      "loss": 5.1334,
      "mean_token_accuracy": 0.21028407216072081,
      "num_tokens": 33252942.0,
      "step": 14905
    },
    {
      "entropy": 5.586065673828125,
      "epoch": 1.6745661818386028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047284517552595933,
      "loss": 5.184,
      "mean_token_accuracy": 0.20406026840209962,
      "num_tokens": 33263892.0,
      "step": 14910
    },
    {
      "entropy": 5.429570007324219,
      "epoch": 1.6751277587465605,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004728260516949287,
      "loss": 5.0644,
      "mean_token_accuracy": 0.20687515437602996,
      "num_tokens": 33274383.0,
      "step": 14915
    },
    {
      "entropy": 5.50610237121582,
      "epoch": 1.6756893356545177,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004728069215650399,
      "loss": 5.0826,
      "mean_token_accuracy": 0.210039122402668,
      "num_tokens": 33285897.0,
      "step": 14920
    },
    {
      "entropy": 5.537733840942383,
      "epoch": 1.6762509125624754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004727877851369019,
      "loss": 5.1607,
      "mean_token_accuracy": 0.20547729581594468,
      "num_tokens": 33297192.0,
      "step": 14925
    },
    {
      "entropy": 5.443004941940307,
      "epoch": 1.676812489470433,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047276864241112417,
      "loss": 5.0644,
      "mean_token_accuracy": 0.21213503032922745,
      "num_tokens": 33308250.0,
      "step": 14930
    },
    {
      "entropy": 5.421862697601318,
      "epoch": 1.6773740663783905,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004727494933883161,
      "loss": 5.0995,
      "mean_token_accuracy": 0.2083527237176895,
      "num_tokens": 33318515.0,
      "step": 14935
    },
    {
      "entropy": 5.53689398765564,
      "epoch": 1.677935643286348,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004727303380690876,
      "loss": 5.196,
      "mean_token_accuracy": 0.2089026913046837,
      "num_tokens": 33329762.0,
      "step": 14940
    },
    {
      "entropy": 5.541815233230591,
      "epoch": 1.6784972201943056,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004727111764540484,
      "loss": 5.1991,
      "mean_token_accuracy": 0.2092477113008499,
      "num_tokens": 33341054.0,
      "step": 14945
    },
    {
      "entropy": 5.4353355884552,
      "epoch": 1.6790587971022632,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047269200854380854,
      "loss": 5.0536,
      "mean_token_accuracy": 0.2144552320241928,
      "num_tokens": 33351524.0,
      "step": 14950
    },
    {
      "entropy": 5.47586498260498,
      "epoch": 1.6796203740102207,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047267283433897854,
      "loss": 5.15,
      "mean_token_accuracy": 0.20320533514022826,
      "num_tokens": 33362953.0,
      "step": 14955
    },
    {
      "entropy": 5.3822277069091795,
      "epoch": 1.6801819509181781,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047265365384016887,
      "loss": 5.0234,
      "mean_token_accuracy": 0.21975015848875046,
      "num_tokens": 33373209.0,
      "step": 14960
    },
    {
      "entropy": 5.37195520401001,
      "epoch": 1.6807435278261358,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047263446704799017,
      "loss": 5.0311,
      "mean_token_accuracy": 0.20979778617620468,
      "num_tokens": 33384775.0,
      "step": 14965
    },
    {
      "entropy": 5.423147392272949,
      "epoch": 1.6813051047340934,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047261527396305343,
      "loss": 5.0762,
      "mean_token_accuracy": 0.20761286616325378,
      "num_tokens": 33395457.0,
      "step": 14970
    },
    {
      "entropy": 5.578161001205444,
      "epoch": 1.6818666816420509,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047259607458596966,
      "loss": 5.216,
      "mean_token_accuracy": 0.20890212059020996,
      "num_tokens": 33407506.0,
      "step": 14975
    },
    {
      "entropy": 5.5212071418762205,
      "epoch": 1.6824282585500083,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047257686891735023,
      "loss": 5.1077,
      "mean_token_accuracy": 0.21759800761938095,
      "num_tokens": 33418154.0,
      "step": 14980
    },
    {
      "entropy": 5.442839241027832,
      "epoch": 1.682989835457966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047255765695780665,
      "loss": 5.1294,
      "mean_token_accuracy": 0.21438449174165725,
      "num_tokens": 33428319.0,
      "step": 14985
    },
    {
      "entropy": 5.4437727451324465,
      "epoch": 1.6835514123659236,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004725384387079507,
      "loss": 5.049,
      "mean_token_accuracy": 0.20992681831121446,
      "num_tokens": 33440429.0,
      "step": 14990
    },
    {
      "entropy": 5.648414611816406,
      "epoch": 1.684112989273881,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047251921416839425,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19861510545015335,
      "num_tokens": 33451736.0,
      "step": 14995
    },
    {
      "entropy": 5.5062929630279545,
      "epoch": 1.6846745661818385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004724999833397494,
      "loss": 5.033,
      "mean_token_accuracy": 0.2050390586256981,
      "num_tokens": 33461991.0,
      "step": 15000
    },
    {
      "epoch": 1.6846745661818385,
      "eval_entropy": 5.232175178124421,
      "eval_loss": 5.187971591949463,
      "eval_mean_token_accuracy": 0.21247658614133602,
      "eval_num_tokens": 33461991.0,
      "eval_runtime": 18.3219,
      "eval_samples_per_second": 1692.673,
      "eval_steps_per_second": 211.605,
      "step": 15000
    },
    {
      "entropy": 5.44553918838501,
      "epoch": 1.6852361430897962,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004724807462226285,
      "loss": 5.1798,
      "mean_token_accuracy": 0.20041977167129515,
      "num_tokens": 33473646.0,
      "step": 15005
    },
    {
      "entropy": 5.432424688339234,
      "epoch": 1.6857977199977539,
      "grad_norm": 0.921875,
      "learning_rate": 0.000472461502817644,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20865580141544343,
      "num_tokens": 33485444.0,
      "step": 15010
    },
    {
      "entropy": 5.6253444194793705,
      "epoch": 1.6863592969057113,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004724422531254087,
      "loss": 5.2776,
      "mean_token_accuracy": 0.20015684515237808,
      "num_tokens": 33497199.0,
      "step": 15015
    },
    {
      "entropy": 5.578074884414673,
      "epoch": 1.6869208738136687,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004724229971465355,
      "loss": 5.2189,
      "mean_token_accuracy": 0.2022972285747528,
      "num_tokens": 33509045.0,
      "step": 15020
    },
    {
      "entropy": 5.448271703720093,
      "epoch": 1.6874824507216264,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004724037348816375,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21786488592624664,
      "num_tokens": 33519730.0,
      "step": 15025
    },
    {
      "entropy": 5.469680643081665,
      "epoch": 1.6880440276295838,
      "grad_norm": 1.09375,
      "learning_rate": 0.000472384466331328,
      "loss": 5.1991,
      "mean_token_accuracy": 0.20163263529539108,
      "num_tokens": 33530393.0,
      "step": 15030
    },
    {
      "entropy": 5.506126165390015,
      "epoch": 1.6886056045375413,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004723651914962205,
      "loss": 5.1577,
      "mean_token_accuracy": 0.2064269080758095,
      "num_tokens": 33543592.0,
      "step": 15035
    },
    {
      "entropy": 5.52133846282959,
      "epoch": 1.689167181445499,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047234591037692884,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20329122841358185,
      "num_tokens": 33554736.0,
      "step": 15040
    },
    {
      "entropy": 5.468460559844971,
      "epoch": 1.6897287583534566,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047232662297406683,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21586479395627975,
      "num_tokens": 33565428.0,
      "step": 15045
    },
    {
      "entropy": 5.553099727630615,
      "epoch": 1.690290335261414,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004723073292882485,
      "loss": 5.3281,
      "mean_token_accuracy": 0.2001742333173752,
      "num_tokens": 33577102.0,
      "step": 15050
    },
    {
      "entropy": 5.5601544857025145,
      "epoch": 1.6908519121693715,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047228802932008836,
      "loss": 5.1621,
      "mean_token_accuracy": 0.2123165100812912,
      "num_tokens": 33587325.0,
      "step": 15055
    },
    {
      "entropy": 5.5672807693481445,
      "epoch": 1.6914134890773291,
      "grad_norm": 1.0,
      "learning_rate": 0.0004722687230702009,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20312864035367967,
      "num_tokens": 33599294.0,
      "step": 15060
    },
    {
      "entropy": 5.5499523162841795,
      "epoch": 1.6919750659852868,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004722494105392007,
      "loss": 5.2524,
      "mean_token_accuracy": 0.2018488749861717,
      "num_tokens": 33609025.0,
      "step": 15065
    },
    {
      "entropy": 5.596439170837402,
      "epoch": 1.6925366428932442,
      "grad_norm": 1.0,
      "learning_rate": 0.0004722300917277029,
      "loss": 5.166,
      "mean_token_accuracy": 0.20324416309595109,
      "num_tokens": 33620881.0,
      "step": 15070
    },
    {
      "entropy": 5.527262496948242,
      "epoch": 1.6930982198012017,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004722107666363224,
      "loss": 5.1091,
      "mean_token_accuracy": 0.21146733462810516,
      "num_tokens": 33631505.0,
      "step": 15075
    },
    {
      "entropy": 5.510390186309815,
      "epoch": 1.6936597967091593,
      "grad_norm": 1.0,
      "learning_rate": 0.0004721914352656746,
      "loss": 5.2832,
      "mean_token_accuracy": 0.2002302035689354,
      "num_tokens": 33642560.0,
      "step": 15080
    },
    {
      "entropy": 5.553929042816162,
      "epoch": 1.694221373617117,
      "grad_norm": 1.015625,
      "learning_rate": 0.000472172097616375,
      "loss": 5.1528,
      "mean_token_accuracy": 0.20899198204278946,
      "num_tokens": 33654706.0,
      "step": 15085
    },
    {
      "entropy": 5.584413671493531,
      "epoch": 1.6947829505250744,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004721527536890394,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19373446553945542,
      "num_tokens": 33665392.0,
      "step": 15090
    },
    {
      "entropy": 5.367098474502564,
      "epoch": 1.6953445274330319,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004721334034842836,
      "loss": 5.0242,
      "mean_token_accuracy": 0.21408725529909134,
      "num_tokens": 33675504.0,
      "step": 15095
    },
    {
      "entropy": 5.5141242980957035,
      "epoch": 1.6959061043409895,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004721140470027238,
      "loss": 5.1549,
      "mean_token_accuracy": 0.20816438049077987,
      "num_tokens": 33686505.0,
      "step": 15100
    },
    {
      "entropy": 5.557978343963623,
      "epoch": 1.6964676812489472,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047209468424497627,
      "loss": 5.157,
      "mean_token_accuracy": 0.20750323981046676,
      "num_tokens": 33698752.0,
      "step": 15105
    },
    {
      "entropy": 5.562022304534912,
      "epoch": 1.6970292581569046,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004720753152116576,
      "loss": 5.2016,
      "mean_token_accuracy": 0.20154131650924684,
      "num_tokens": 33710323.0,
      "step": 15110
    },
    {
      "entropy": 5.593322467803955,
      "epoch": 1.697590835064862,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047205593990338443,
      "loss": 5.2761,
      "mean_token_accuracy": 0.19099510312080384,
      "num_tokens": 33721875.0,
      "step": 15115
    },
    {
      "entropy": 5.499043798446655,
      "epoch": 1.6981524119728197,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004720365583207737,
      "loss": 4.977,
      "mean_token_accuracy": 0.21811393201351165,
      "num_tokens": 33731520.0,
      "step": 15120
    },
    {
      "entropy": 5.554244136810302,
      "epoch": 1.6987139888807772,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004720171704644425,
      "loss": 5.2485,
      "mean_token_accuracy": 0.19921279102563857,
      "num_tokens": 33743351.0,
      "step": 15125
    },
    {
      "entropy": 5.546348237991333,
      "epoch": 1.6992755657887346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004719977763350083,
      "loss": 5.1606,
      "mean_token_accuracy": 0.21103411614894868,
      "num_tokens": 33754879.0,
      "step": 15130
    },
    {
      "entropy": 5.430835437774658,
      "epoch": 1.6998371426966923,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047197837593308833,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20844728499650955,
      "num_tokens": 33766671.0,
      "step": 15135
    },
    {
      "entropy": 5.527913570404053,
      "epoch": 1.70039871960465,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004719589692593006,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20766616612672806,
      "num_tokens": 33778236.0,
      "step": 15140
    },
    {
      "entropy": 5.481860828399658,
      "epoch": 1.7009602965126074,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004719395563142629,
      "loss": 5.0485,
      "mean_token_accuracy": 0.20752263814210892,
      "num_tokens": 33788898.0,
      "step": 15145
    },
    {
      "entropy": 5.503481388092041,
      "epoch": 1.7015218734205648,
      "grad_norm": 1.0,
      "learning_rate": 0.0004719201370985933,
      "loss": 5.2029,
      "mean_token_accuracy": 0.20233033299446107,
      "num_tokens": 33800641.0,
      "step": 15150
    },
    {
      "entropy": 5.553616380691528,
      "epoch": 1.7020834503285225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047190071161291007,
      "loss": 5.107,
      "mean_token_accuracy": 0.20703384429216384,
      "num_tokens": 33812581.0,
      "step": 15155
    },
    {
      "entropy": 5.441648197174072,
      "epoch": 1.7026450272364801,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047188127985783195,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21562660336494446,
      "num_tokens": 33823378.0,
      "step": 15160
    },
    {
      "entropy": 5.476008319854737,
      "epoch": 1.7032066041444376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047186184183397754,
      "loss": 5.1672,
      "mean_token_accuracy": 0.20512134283781053,
      "num_tokens": 33834804.0,
      "step": 15165
    },
    {
      "entropy": 5.449408864974975,
      "epoch": 1.703768181052395,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047184239754196564,
      "loss": 5.131,
      "mean_token_accuracy": 0.20889237374067307,
      "num_tokens": 33847044.0,
      "step": 15170
    },
    {
      "entropy": 5.5294944763183596,
      "epoch": 1.7043297579603527,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004718229469824155,
      "loss": 5.0855,
      "mean_token_accuracy": 0.2038659244775772,
      "num_tokens": 33858077.0,
      "step": 15175
    },
    {
      "entropy": 5.484337854385376,
      "epoch": 1.7048913348683103,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047180349015594634,
      "loss": 5.134,
      "mean_token_accuracy": 0.2047533482313156,
      "num_tokens": 33868765.0,
      "step": 15180
    },
    {
      "entropy": 5.43833475112915,
      "epoch": 1.7054529117762678,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004717840270631778,
      "loss": 5.0709,
      "mean_token_accuracy": 0.21182419508695602,
      "num_tokens": 33880415.0,
      "step": 15185
    },
    {
      "entropy": 5.487602281570434,
      "epoch": 1.7060144886842252,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047176455770472946,
      "loss": 5.1251,
      "mean_token_accuracy": 0.20670740008354188,
      "num_tokens": 33892081.0,
      "step": 15190
    },
    {
      "entropy": 5.575546455383301,
      "epoch": 1.706576065592183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047174508208122135,
      "loss": 5.2542,
      "mean_token_accuracy": 0.20220078229904176,
      "num_tokens": 33903433.0,
      "step": 15195
    },
    {
      "entropy": 5.546103191375733,
      "epoch": 1.7071376425001406,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004717256001932735,
      "loss": 5.1226,
      "mean_token_accuracy": 0.2091883361339569,
      "num_tokens": 33914706.0,
      "step": 15200
    },
    {
      "entropy": 5.480507898330688,
      "epoch": 1.707699219408098,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047170611204150624,
      "loss": 5.2388,
      "mean_token_accuracy": 0.20728666484355926,
      "num_tokens": 33927195.0,
      "step": 15205
    },
    {
      "entropy": 5.505918407440186,
      "epoch": 1.7082607963160554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004716866176265401,
      "loss": 5.1261,
      "mean_token_accuracy": 0.21205178052186965,
      "num_tokens": 33938422.0,
      "step": 15210
    },
    {
      "entropy": 5.579632568359375,
      "epoch": 1.708822373224013,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047166711694899573,
      "loss": 5.1436,
      "mean_token_accuracy": 0.21102875471115112,
      "num_tokens": 33949775.0,
      "step": 15215
    },
    {
      "entropy": 5.493266582489014,
      "epoch": 1.7093839501319705,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047164761000949404,
      "loss": 5.1162,
      "mean_token_accuracy": 0.20836064219474792,
      "num_tokens": 33960371.0,
      "step": 15220
    },
    {
      "entropy": 5.412378931045533,
      "epoch": 1.709945527039928,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004716280968086563,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20634318888187408,
      "num_tokens": 33971326.0,
      "step": 15225
    },
    {
      "entropy": 5.541870260238648,
      "epoch": 1.7105071039478856,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004716085773471036,
      "loss": 5.1858,
      "mean_token_accuracy": 0.20716300010681152,
      "num_tokens": 33982744.0,
      "step": 15230
    },
    {
      "entropy": 5.5545313358306885,
      "epoch": 1.7110686808558433,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047158905162545756,
      "loss": 5.2274,
      "mean_token_accuracy": 0.20111105293035508,
      "num_tokens": 33993647.0,
      "step": 15235
    },
    {
      "entropy": 5.457699489593506,
      "epoch": 1.7116302577638007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047156951964433984,
      "loss": 5.074,
      "mean_token_accuracy": 0.20902696698904039,
      "num_tokens": 34004026.0,
      "step": 15240
    },
    {
      "entropy": 5.475416469573974,
      "epoch": 1.7121918346717582,
      "grad_norm": 0.875,
      "learning_rate": 0.00047154998140437246,
      "loss": 5.1451,
      "mean_token_accuracy": 0.21045226454734803,
      "num_tokens": 34014882.0,
      "step": 15245
    },
    {
      "entropy": 5.473147296905518,
      "epoch": 1.7127534115797158,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047153043690617737,
      "loss": 5.1701,
      "mean_token_accuracy": 0.20746184885501862,
      "num_tokens": 34026548.0,
      "step": 15250
    },
    {
      "entropy": 5.456924057006836,
      "epoch": 1.7133149884876735,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000471510886150377,
      "loss": 5.0418,
      "mean_token_accuracy": 0.21107793897390364,
      "num_tokens": 34038393.0,
      "step": 15255
    },
    {
      "entropy": 5.5369831085205075,
      "epoch": 1.713876565395631,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004714913291375938,
      "loss": 5.1601,
      "mean_token_accuracy": 0.20325452834367752,
      "num_tokens": 34050257.0,
      "step": 15260
    },
    {
      "entropy": 5.507985544204712,
      "epoch": 1.7144381423035884,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047147176586845036,
      "loss": 5.1255,
      "mean_token_accuracy": 0.20985889434814453,
      "num_tokens": 34061860.0,
      "step": 15265
    },
    {
      "entropy": 5.550496912002563,
      "epoch": 1.714999719211546,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004714521963435698,
      "loss": 5.1786,
      "mean_token_accuracy": 0.202415731549263,
      "num_tokens": 34075247.0,
      "step": 15270
    },
    {
      "entropy": 5.512369871139526,
      "epoch": 1.7155612961195037,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047143262056357506,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20393330305814744,
      "num_tokens": 34085801.0,
      "step": 15275
    },
    {
      "entropy": 5.533247232437134,
      "epoch": 1.7161228730274611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047141303852908965,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20642355382442473,
      "num_tokens": 34095686.0,
      "step": 15280
    },
    {
      "entropy": 5.4551043033599855,
      "epoch": 1.7166844499354186,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004713934502407367,
      "loss": 5.0952,
      "mean_token_accuracy": 0.2121119812130928,
      "num_tokens": 34106518.0,
      "step": 15285
    },
    {
      "entropy": 5.502656698226929,
      "epoch": 1.7172460268433762,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004713738556991402,
      "loss": 5.1367,
      "mean_token_accuracy": 0.21238022446632385,
      "num_tokens": 34117793.0,
      "step": 15290
    },
    {
      "entropy": 5.543705701828003,
      "epoch": 1.717807603751334,
      "grad_norm": 1.0625,
      "learning_rate": 0.000471354254904924,
      "loss": 5.2046,
      "mean_token_accuracy": 0.20887165814638137,
      "num_tokens": 34128748.0,
      "step": 15295
    },
    {
      "entropy": 5.488872241973877,
      "epoch": 1.7183691806592913,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047133464785871216,
      "loss": 5.1311,
      "mean_token_accuracy": 0.21063185781240462,
      "num_tokens": 34140597.0,
      "step": 15300
    },
    {
      "entropy": 5.578644800186157,
      "epoch": 1.7189307575672488,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000471315034561129,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20001179575920106,
      "num_tokens": 34151823.0,
      "step": 15305
    },
    {
      "entropy": 5.4561553478240965,
      "epoch": 1.7194923344752064,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047129541501279897,
      "loss": 5.0687,
      "mean_token_accuracy": 0.2141604408621788,
      "num_tokens": 34163697.0,
      "step": 15310
    },
    {
      "entropy": 5.496334600448608,
      "epoch": 1.7200539113831639,
      "grad_norm": 1.0,
      "learning_rate": 0.0004712757892143468,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21214486807584762,
      "num_tokens": 34175058.0,
      "step": 15315
    },
    {
      "entropy": 5.550631427764893,
      "epoch": 1.7206154882911213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047125615716639744,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20638849288225175,
      "num_tokens": 34186883.0,
      "step": 15320
    },
    {
      "entropy": 5.465950059890747,
      "epoch": 1.721177065199079,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004712365188695758,
      "loss": 5.1005,
      "mean_token_accuracy": 0.21199408024549485,
      "num_tokens": 34198103.0,
      "step": 15325
    },
    {
      "entropy": 5.4663896560668945,
      "epoch": 1.7217386421070366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004712168743245074,
      "loss": 5.0781,
      "mean_token_accuracy": 0.21162348687648774,
      "num_tokens": 34208747.0,
      "step": 15330
    },
    {
      "entropy": 5.495966386795044,
      "epoch": 1.722300219014994,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047119722353181754,
      "loss": 5.2094,
      "mean_token_accuracy": 0.20605195164680482,
      "num_tokens": 34220775.0,
      "step": 15335
    },
    {
      "entropy": 5.5106219291687015,
      "epoch": 1.7228617959229515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047117756649213204,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21309981048107146,
      "num_tokens": 34231031.0,
      "step": 15340
    },
    {
      "entropy": 5.530784797668457,
      "epoch": 1.7234233728309092,
      "grad_norm": 1.0,
      "learning_rate": 0.0004711579032060767,
      "loss": 5.1828,
      "mean_token_accuracy": 0.2067065492272377,
      "num_tokens": 34241850.0,
      "step": 15345
    },
    {
      "entropy": 5.504141616821289,
      "epoch": 1.7239849497388668,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004711382336742777,
      "loss": 5.0622,
      "mean_token_accuracy": 0.21268633008003235,
      "num_tokens": 34252732.0,
      "step": 15350
    },
    {
      "entropy": 5.420687675476074,
      "epoch": 1.7245465266468243,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004711185578973612,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20594823956489564,
      "num_tokens": 34264066.0,
      "step": 15355
    },
    {
      "entropy": 5.562918663024902,
      "epoch": 1.7251081035547817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004710988758759538,
      "loss": 5.1934,
      "mean_token_accuracy": 0.19903359860181807,
      "num_tokens": 34275239.0,
      "step": 15360
    },
    {
      "entropy": 5.4476464748382565,
      "epoch": 1.7256696804627394,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004710791876106821,
      "loss": 5.0525,
      "mean_token_accuracy": 0.20853817611932754,
      "num_tokens": 34285723.0,
      "step": 15365
    },
    {
      "entropy": 5.497724485397339,
      "epoch": 1.726231257370697,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047105949310217304,
      "loss": 5.0993,
      "mean_token_accuracy": 0.212528495490551,
      "num_tokens": 34296292.0,
      "step": 15370
    },
    {
      "entropy": 5.4754719734191895,
      "epoch": 1.7267928342786545,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004710397923510537,
      "loss": 5.1212,
      "mean_token_accuracy": 0.21450895816087723,
      "num_tokens": 34307232.0,
      "step": 15375
    },
    {
      "entropy": 5.468585395812989,
      "epoch": 1.727354411186612,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047102008535795134,
      "loss": 5.1296,
      "mean_token_accuracy": 0.21698434203863143,
      "num_tokens": 34318933.0,
      "step": 15380
    },
    {
      "entropy": 5.518647527694702,
      "epoch": 1.7279159880945696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004710003721234934,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2157742887735367,
      "num_tokens": 34330431.0,
      "step": 15385
    },
    {
      "entropy": 5.541515827178955,
      "epoch": 1.7284775650025272,
      "grad_norm": 1.0,
      "learning_rate": 0.00047098065264830754,
      "loss": 5.1741,
      "mean_token_accuracy": 0.20666055381298065,
      "num_tokens": 34342006.0,
      "step": 15390
    },
    {
      "entropy": 5.526775598526001,
      "epoch": 1.7290391419104847,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047096092693302174,
      "loss": 5.0857,
      "mean_token_accuracy": 0.20858652591705323,
      "num_tokens": 34353390.0,
      "step": 15395
    },
    {
      "entropy": 5.407729434967041,
      "epoch": 1.7296007188184421,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047094119497826406,
      "loss": 5.0803,
      "mean_token_accuracy": 0.20938567072153091,
      "num_tokens": 34364211.0,
      "step": 15400
    },
    {
      "entropy": 5.485868453979492,
      "epoch": 1.7301622957263998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047092145678466273,
      "loss": 5.0868,
      "mean_token_accuracy": 0.2073808416724205,
      "num_tokens": 34375362.0,
      "step": 15405
    },
    {
      "entropy": 5.4084801197052,
      "epoch": 1.7307238726343572,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004709017123528462,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21943629831075667,
      "num_tokens": 34385511.0,
      "step": 15410
    },
    {
      "entropy": 5.488828182220459,
      "epoch": 1.7312854495423147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047088196168344315,
      "loss": 5.2217,
      "mean_token_accuracy": 0.20403611212968825,
      "num_tokens": 34397060.0,
      "step": 15415
    },
    {
      "entropy": 5.4864503860473635,
      "epoch": 1.7318470264502723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047086220477708256,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20844440013170243,
      "num_tokens": 34407703.0,
      "step": 15420
    },
    {
      "entropy": 5.49847297668457,
      "epoch": 1.73240860335823,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047084244163439335,
      "loss": 5.1447,
      "mean_token_accuracy": 0.21463810801506042,
      "num_tokens": 34418929.0,
      "step": 15425
    },
    {
      "entropy": 5.425997543334961,
      "epoch": 1.7329701802661874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004708226722560049,
      "loss": 5.0426,
      "mean_token_accuracy": 0.21104160249233245,
      "num_tokens": 34430693.0,
      "step": 15430
    },
    {
      "entropy": 5.448315000534057,
      "epoch": 1.7335317571741449,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004708028966425465,
      "loss": 5.0997,
      "mean_token_accuracy": 0.2069340541958809,
      "num_tokens": 34442299.0,
      "step": 15435
    },
    {
      "entropy": 5.512123918533325,
      "epoch": 1.7340933340821025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047078311479464797,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20963587760925292,
      "num_tokens": 34453426.0,
      "step": 15440
    },
    {
      "entropy": 5.466936779022217,
      "epoch": 1.7346549109900602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047076332671293917,
      "loss": 5.0861,
      "mean_token_accuracy": 0.21304211020469666,
      "num_tokens": 34463601.0,
      "step": 15445
    },
    {
      "entropy": 5.515047788619995,
      "epoch": 1.7352164878980176,
      "grad_norm": 1.0,
      "learning_rate": 0.00047074353239805005,
      "loss": 5.0716,
      "mean_token_accuracy": 0.21345555335283278,
      "num_tokens": 34474938.0,
      "step": 15450
    },
    {
      "entropy": 5.435859251022339,
      "epoch": 1.735778064805975,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004707237318506111,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21603488624095918,
      "num_tokens": 34485244.0,
      "step": 15455
    },
    {
      "entropy": 5.417296314239502,
      "epoch": 1.7363396417139327,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047070392507125245,
      "loss": 5.0233,
      "mean_token_accuracy": 0.21657561659812927,
      "num_tokens": 34496912.0,
      "step": 15460
    },
    {
      "entropy": 5.469209384918213,
      "epoch": 1.7369012186218904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000470684112060605,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20672615617513657,
      "num_tokens": 34508122.0,
      "step": 15465
    },
    {
      "entropy": 5.520959901809692,
      "epoch": 1.7374627955298478,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047066429281929944,
      "loss": 5.1022,
      "mean_token_accuracy": 0.21297637671232222,
      "num_tokens": 34519522.0,
      "step": 15470
    },
    {
      "entropy": 5.486133193969726,
      "epoch": 1.7380243724378053,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000470644467347967,
      "loss": 5.089,
      "mean_token_accuracy": 0.21493441313505174,
      "num_tokens": 34532832.0,
      "step": 15475
    },
    {
      "entropy": 5.499406433105468,
      "epoch": 1.738585949345763,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047062463564723884,
      "loss": 5.2114,
      "mean_token_accuracy": 0.2051677480340004,
      "num_tokens": 34544817.0,
      "step": 15480
    },
    {
      "entropy": 5.4390730381011965,
      "epoch": 1.7391475262537206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004706047977177464,
      "loss": 5.04,
      "mean_token_accuracy": 0.21157858967781068,
      "num_tokens": 34555391.0,
      "step": 15485
    },
    {
      "entropy": 5.424674701690674,
      "epoch": 1.739709103161678,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047058495356012135,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21034435629844667,
      "num_tokens": 34567531.0,
      "step": 15490
    },
    {
      "entropy": 5.505863666534424,
      "epoch": 1.7402706800696355,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047056510317499545,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20759210288524627,
      "num_tokens": 34578212.0,
      "step": 15495
    },
    {
      "entropy": 5.432086086273193,
      "epoch": 1.7408322569775931,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047054524656300087,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21777308583259583,
      "num_tokens": 34589331.0,
      "step": 15500
    },
    {
      "entropy": 5.418513917922974,
      "epoch": 1.7413938338855506,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047052538372476985,
      "loss": 5.109,
      "mean_token_accuracy": 0.21111343204975128,
      "num_tokens": 34601467.0,
      "step": 15505
    },
    {
      "entropy": 5.414806699752807,
      "epoch": 1.741955410793508,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047050551466093466,
      "loss": 5.0767,
      "mean_token_accuracy": 0.2150661826133728,
      "num_tokens": 34611608.0,
      "step": 15510
    },
    {
      "entropy": 5.475686407089233,
      "epoch": 1.7425169877014657,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004704856393721281,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21589724719524384,
      "num_tokens": 34622495.0,
      "step": 15515
    },
    {
      "entropy": 5.583570766448974,
      "epoch": 1.7430785646094233,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047046575785898297,
      "loss": 5.2391,
      "mean_token_accuracy": 0.20443634688854218,
      "num_tokens": 34633279.0,
      "step": 15520
    },
    {
      "entropy": 5.4607751846313475,
      "epoch": 1.7436401415173808,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004704458701221323,
      "loss": 5.0923,
      "mean_token_accuracy": 0.20940332263708114,
      "num_tokens": 34644030.0,
      "step": 15525
    },
    {
      "entropy": 5.589746570587158,
      "epoch": 1.7442017184253382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004704259761622094,
      "loss": 5.1854,
      "mean_token_accuracy": 0.1998036801815033,
      "num_tokens": 34655540.0,
      "step": 15530
    },
    {
      "entropy": 5.5024128437042235,
      "epoch": 1.7447632953332959,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047040607597984745,
      "loss": 5.1189,
      "mean_token_accuracy": 0.2082844704389572,
      "num_tokens": 34666752.0,
      "step": 15535
    },
    {
      "entropy": 5.498738765716553,
      "epoch": 1.7453248722412535,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004703861695756804,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20718310326337813,
      "num_tokens": 34678670.0,
      "step": 15540
    },
    {
      "entropy": 5.5130092144012455,
      "epoch": 1.745886449149211,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004703662569503419,
      "loss": 5.1373,
      "mean_token_accuracy": 0.20873839855194093,
      "num_tokens": 34690848.0,
      "step": 15545
    },
    {
      "entropy": 5.533270835876465,
      "epoch": 1.7464480260571684,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047034633810446604,
      "loss": 5.1441,
      "mean_token_accuracy": 0.2038038730621338,
      "num_tokens": 34703135.0,
      "step": 15550
    },
    {
      "entropy": 5.489834880828857,
      "epoch": 1.747009602965126,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004703264130386869,
      "loss": 5.0836,
      "mean_token_accuracy": 0.20878920704126358,
      "num_tokens": 34714329.0,
      "step": 15555
    },
    {
      "entropy": 5.515345478057862,
      "epoch": 1.7475711798730837,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047030648175363913,
      "loss": 5.1475,
      "mean_token_accuracy": 0.1987304389476776,
      "num_tokens": 34727032.0,
      "step": 15560
    },
    {
      "entropy": 5.530332756042481,
      "epoch": 1.7481327567810412,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004702865442499572,
      "loss": 5.1431,
      "mean_token_accuracy": 0.2069760039448738,
      "num_tokens": 34738600.0,
      "step": 15565
    },
    {
      "entropy": 5.551826143264771,
      "epoch": 1.7486943336889986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004702666005282759,
      "loss": 5.1849,
      "mean_token_accuracy": 0.20250886529684067,
      "num_tokens": 34749810.0,
      "step": 15570
    },
    {
      "entropy": 5.679680919647216,
      "epoch": 1.7492559105969563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004702466505892304,
      "loss": 5.3224,
      "mean_token_accuracy": 0.19772350043058395,
      "num_tokens": 34762029.0,
      "step": 15575
    },
    {
      "entropy": 5.4695165157318115,
      "epoch": 1.749817487504914,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047022669443345575,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20618197619915007,
      "num_tokens": 34773160.0,
      "step": 15580
    },
    {
      "entropy": 5.450992393493652,
      "epoch": 1.7503790644128714,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047020673206158743,
      "loss": 5.1564,
      "mean_token_accuracy": 0.20454892218112947,
      "num_tokens": 34784659.0,
      "step": 15585
    },
    {
      "entropy": 5.588104867935181,
      "epoch": 1.7509406413208288,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000470186763474261,
      "loss": 5.1343,
      "mean_token_accuracy": 0.21105916500091554,
      "num_tokens": 34796976.0,
      "step": 15590
    },
    {
      "entropy": 5.450671052932739,
      "epoch": 1.7515022182287865,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004701667886721125,
      "loss": 5.0721,
      "mean_token_accuracy": 0.2078850522637367,
      "num_tokens": 34809271.0,
      "step": 15595
    },
    {
      "entropy": 5.518737363815307,
      "epoch": 1.752063795136744,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004701468076557776,
      "loss": 5.1409,
      "mean_token_accuracy": 0.2118055209517479,
      "num_tokens": 34820291.0,
      "step": 15600
    },
    {
      "entropy": 5.428896713256836,
      "epoch": 1.7526253720447014,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047012682042589277,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21313309371471406,
      "num_tokens": 34831224.0,
      "step": 15605
    },
    {
      "entropy": 5.516469573974609,
      "epoch": 1.753186948952659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047010682698309423,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20943866819143295,
      "num_tokens": 34842947.0,
      "step": 15610
    },
    {
      "entropy": 5.467283105850219,
      "epoch": 1.7537485258606167,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004700868273280186,
      "loss": 5.0801,
      "mean_token_accuracy": 0.2154863640666008,
      "num_tokens": 34853422.0,
      "step": 15615
    },
    {
      "entropy": 5.441207838058472,
      "epoch": 1.7543101027685741,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047006682146130273,
      "loss": 5.0984,
      "mean_token_accuracy": 0.2049204781651497,
      "num_tokens": 34864082.0,
      "step": 15620
    },
    {
      "entropy": 5.4395575523376465,
      "epoch": 1.7548716796765316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004700468093835837,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21880912482738496,
      "num_tokens": 34874094.0,
      "step": 15625
    },
    {
      "entropy": 5.456491136550904,
      "epoch": 1.7554332565844892,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047002679109549864,
      "loss": 5.1019,
      "mean_token_accuracy": 0.21333929598331453,
      "num_tokens": 34884597.0,
      "step": 15630
    },
    {
      "entropy": 5.405950736999512,
      "epoch": 1.755994833492447,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047000676659768477,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21880512535572053,
      "num_tokens": 34894746.0,
      "step": 15635
    },
    {
      "entropy": 5.462916088104248,
      "epoch": 1.7565564104004043,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004699867358907799,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21646392196416855,
      "num_tokens": 34906250.0,
      "step": 15640
    },
    {
      "entropy": 5.5513204574584964,
      "epoch": 1.7571179873083618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004699666989754217,
      "loss": 5.2674,
      "mean_token_accuracy": 0.202106811106205,
      "num_tokens": 34917772.0,
      "step": 15645
    },
    {
      "entropy": 5.506875991821289,
      "epoch": 1.7576795642163194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004699466558522482,
      "loss": 5.1012,
      "mean_token_accuracy": 0.21430955976247787,
      "num_tokens": 34930243.0,
      "step": 15650
    },
    {
      "entropy": 5.457162189483642,
      "epoch": 1.758241141124277,
      "grad_norm": 1.0,
      "learning_rate": 0.00046992660652189755,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20771168917417526,
      "num_tokens": 34941357.0,
      "step": 15655
    },
    {
      "entropy": 5.581509637832641,
      "epoch": 1.7588027180322345,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046990655098500824,
      "loss": 5.2498,
      "mean_token_accuracy": 0.1956062912940979,
      "num_tokens": 34952116.0,
      "step": 15660
    },
    {
      "entropy": 5.529504013061524,
      "epoch": 1.759364294940192,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046988648924221866,
      "loss": 5.1422,
      "mean_token_accuracy": 0.20517128705978394,
      "num_tokens": 34963707.0,
      "step": 15665
    },
    {
      "entropy": 5.562208366394043,
      "epoch": 1.7599258718481496,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004698664212941677,
      "loss": 5.2446,
      "mean_token_accuracy": 0.20399125367403032,
      "num_tokens": 34974376.0,
      "step": 15670
    },
    {
      "entropy": 5.528583908081055,
      "epoch": 1.7604874487561073,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046984634714149437,
      "loss": 5.1281,
      "mean_token_accuracy": 0.2076116606593132,
      "num_tokens": 34986183.0,
      "step": 15675
    },
    {
      "entropy": 5.603015613555908,
      "epoch": 1.7610490256640647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004698262667848377,
      "loss": 5.1428,
      "mean_token_accuracy": 0.21122328341007232,
      "num_tokens": 34997137.0,
      "step": 15680
    },
    {
      "entropy": 5.578436374664307,
      "epoch": 1.7616106025720222,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046980618022483714,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19657762944698334,
      "num_tokens": 35008944.0,
      "step": 15685
    },
    {
      "entropy": 5.541048336029053,
      "epoch": 1.7621721794799798,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046978608746213224,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20120044350624083,
      "num_tokens": 35019972.0,
      "step": 15690
    },
    {
      "entropy": 5.476521062850952,
      "epoch": 1.7627337563879373,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004697659884973628,
      "loss": 5.064,
      "mean_token_accuracy": 0.2112387552857399,
      "num_tokens": 35031285.0,
      "step": 15695
    },
    {
      "entropy": 5.4324907779693605,
      "epoch": 1.7632953332958947,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004697458833311687,
      "loss": 5.0395,
      "mean_token_accuracy": 0.21114134788513184,
      "num_tokens": 35042553.0,
      "step": 15700
    },
    {
      "entropy": 5.59822416305542,
      "epoch": 1.7638569102038524,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004697257719641902,
      "loss": 5.2543,
      "mean_token_accuracy": 0.19547551721334458,
      "num_tokens": 35053637.0,
      "step": 15705
    },
    {
      "entropy": 5.531178522109985,
      "epoch": 1.76441848711181,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004697056543970676,
      "loss": 5.188,
      "mean_token_accuracy": 0.21026747673749924,
      "num_tokens": 35064754.0,
      "step": 15710
    },
    {
      "entropy": 5.539709615707397,
      "epoch": 1.7649800640197675,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046968553063044146,
      "loss": 5.0954,
      "mean_token_accuracy": 0.20780241936445237,
      "num_tokens": 35075816.0,
      "step": 15715
    },
    {
      "entropy": 5.546365880966187,
      "epoch": 1.765541640927725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046966540066495247,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20350608974695206,
      "num_tokens": 35088647.0,
      "step": 15720
    },
    {
      "entropy": 5.483157014846801,
      "epoch": 1.7661032178356826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046964526450124164,
      "loss": 5.0789,
      "mean_token_accuracy": 0.20694421231746674,
      "num_tokens": 35099372.0,
      "step": 15725
    },
    {
      "entropy": 5.530976581573486,
      "epoch": 1.7666647947436402,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046962512213995,
      "loss": 5.219,
      "mean_token_accuracy": 0.20368565917015075,
      "num_tokens": 35110663.0,
      "step": 15730
    },
    {
      "entropy": 5.688347959518433,
      "epoch": 1.7672263716515977,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046960497358171913,
      "loss": 5.3034,
      "mean_token_accuracy": 0.19481257051229478,
      "num_tokens": 35122406.0,
      "step": 15735
    },
    {
      "entropy": 5.46981725692749,
      "epoch": 1.7677879485595551,
      "grad_norm": 1.0,
      "learning_rate": 0.00046958481882719035,
      "loss": 5.069,
      "mean_token_accuracy": 0.21117494702339173,
      "num_tokens": 35133444.0,
      "step": 15740
    },
    {
      "entropy": 5.4243584156036375,
      "epoch": 1.7683495254675128,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046956465787700554,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20845517069101333,
      "num_tokens": 35143840.0,
      "step": 15745
    },
    {
      "entropy": 5.591473197937011,
      "epoch": 1.7689111023754704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046954449073180644,
      "loss": 5.2738,
      "mean_token_accuracy": 0.19550443440675735,
      "num_tokens": 35155050.0,
      "step": 15750
    },
    {
      "entropy": 5.544423151016235,
      "epoch": 1.7694726792834279,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004695243173922354,
      "loss": 5.1877,
      "mean_token_accuracy": 0.20177470296621322,
      "num_tokens": 35167649.0,
      "step": 15755
    },
    {
      "entropy": 5.491764163970947,
      "epoch": 1.7700342561913853,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046950413785893465,
      "loss": 5.1146,
      "mean_token_accuracy": 0.21210612058639527,
      "num_tokens": 35179238.0,
      "step": 15760
    },
    {
      "entropy": 5.558326864242554,
      "epoch": 1.770595833099343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046948395213254667,
      "loss": 5.1659,
      "mean_token_accuracy": 0.20976133346557618,
      "num_tokens": 35189510.0,
      "step": 15765
    },
    {
      "entropy": 5.563194131851196,
      "epoch": 1.7711574100073006,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004694637602137143,
      "loss": 5.1543,
      "mean_token_accuracy": 0.20395326614379883,
      "num_tokens": 35199727.0,
      "step": 15770
    },
    {
      "entropy": 5.467804193496704,
      "epoch": 1.771718986915258,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004694435621030803,
      "loss": 5.1097,
      "mean_token_accuracy": 0.21488792449235916,
      "num_tokens": 35210311.0,
      "step": 15775
    },
    {
      "entropy": 5.526813936233521,
      "epoch": 1.7722805638232155,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046942335780128785,
      "loss": 5.1397,
      "mean_token_accuracy": 0.212800133228302,
      "num_tokens": 35222755.0,
      "step": 15780
    },
    {
      "entropy": 5.510811376571655,
      "epoch": 1.7728421407311732,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046940314730898035,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20129836201667786,
      "num_tokens": 35234337.0,
      "step": 15785
    },
    {
      "entropy": 5.491893148422241,
      "epoch": 1.7734037176391306,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004693829306268012,
      "loss": 5.1254,
      "mean_token_accuracy": 0.20320425927639008,
      "num_tokens": 35245736.0,
      "step": 15790
    },
    {
      "entropy": 5.4725401401519775,
      "epoch": 1.773965294547088,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004693627077553942,
      "loss": 5.0821,
      "mean_token_accuracy": 0.21413436979055406,
      "num_tokens": 35256368.0,
      "step": 15795
    },
    {
      "entropy": 5.476490163803101,
      "epoch": 1.7745268714550457,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004693424786954031,
      "loss": 5.1542,
      "mean_token_accuracy": 0.20727033913135529,
      "num_tokens": 35267238.0,
      "step": 15800
    },
    {
      "entropy": 5.5224419116973875,
      "epoch": 1.7750884483630034,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046932224344747224,
      "loss": 5.1033,
      "mean_token_accuracy": 0.2042641207575798,
      "num_tokens": 35277113.0,
      "step": 15805
    },
    {
      "entropy": 5.4780128479003904,
      "epoch": 1.7756500252709608,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046930200201224573,
      "loss": 5.0415,
      "mean_token_accuracy": 0.2058327704668045,
      "num_tokens": 35288504.0,
      "step": 15810
    },
    {
      "entropy": 5.479607009887696,
      "epoch": 1.7762116021789183,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004692817543903681,
      "loss": 5.0657,
      "mean_token_accuracy": 0.21436280608177186,
      "num_tokens": 35301215.0,
      "step": 15815
    },
    {
      "entropy": 5.442288064956665,
      "epoch": 1.776773179086876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000469261500582484,
      "loss": 5.0728,
      "mean_token_accuracy": 0.2079812318086624,
      "num_tokens": 35311997.0,
      "step": 15820
    },
    {
      "entropy": 5.525636100769043,
      "epoch": 1.7773347559948336,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046924124058923843,
      "loss": 5.1464,
      "mean_token_accuracy": 0.20737137645483017,
      "num_tokens": 35323217.0,
      "step": 15825
    },
    {
      "entropy": 5.526032400131226,
      "epoch": 1.777896332902791,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004692209744112765,
      "loss": 5.1699,
      "mean_token_accuracy": 0.21151493042707442,
      "num_tokens": 35333102.0,
      "step": 15830
    },
    {
      "entropy": 5.562783288955688,
      "epoch": 1.7784579098107485,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004692007020492433,
      "loss": 5.1959,
      "mean_token_accuracy": 0.20669854283332825,
      "num_tokens": 35346394.0,
      "step": 15835
    },
    {
      "entropy": 5.415058898925781,
      "epoch": 1.7790194867187061,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004691804235037845,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21598448604345322,
      "num_tokens": 35357827.0,
      "step": 15840
    },
    {
      "entropy": 5.479741525650025,
      "epoch": 1.7795810636266638,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004691601387755456,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20664003044366835,
      "num_tokens": 35370517.0,
      "step": 15845
    },
    {
      "entropy": 5.5027184009552,
      "epoch": 1.7801426405346212,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046913984786517275,
      "loss": 5.1437,
      "mean_token_accuracy": 0.20750885754823684,
      "num_tokens": 35381830.0,
      "step": 15850
    },
    {
      "entropy": 5.584297227859497,
      "epoch": 1.7807042174425787,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046911955077331177,
      "loss": 5.1994,
      "mean_token_accuracy": 0.19884348511695862,
      "num_tokens": 35393161.0,
      "step": 15855
    },
    {
      "entropy": 5.5473456382751465,
      "epoch": 1.7812657943505363,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046909924750060893,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21679302603006362,
      "num_tokens": 35403338.0,
      "step": 15860
    },
    {
      "entropy": 5.535284900665284,
      "epoch": 1.781827371258494,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004690789380477109,
      "loss": 5.1797,
      "mean_token_accuracy": 0.2024092271924019,
      "num_tokens": 35415862.0,
      "step": 15865
    },
    {
      "entropy": 5.424515962600708,
      "epoch": 1.7823889481664514,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004690586224152641,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20500966012477875,
      "num_tokens": 35426696.0,
      "step": 15870
    },
    {
      "entropy": 5.494702529907227,
      "epoch": 1.7829505250744089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004690383006039156,
      "loss": 5.1068,
      "mean_token_accuracy": 0.21053788512945176,
      "num_tokens": 35438624.0,
      "step": 15875
    },
    {
      "entropy": 5.57385835647583,
      "epoch": 1.7835121019823665,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004690179726143123,
      "loss": 5.2522,
      "mean_token_accuracy": 0.20170057713985443,
      "num_tokens": 35449841.0,
      "step": 15880
    },
    {
      "entropy": 5.667948913574219,
      "epoch": 1.784073678890324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004689976384471014,
      "loss": 5.3342,
      "mean_token_accuracy": 0.19652565121650695,
      "num_tokens": 35462081.0,
      "step": 15885
    },
    {
      "entropy": 5.4623558044433596,
      "epoch": 1.7846352557982814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004689772981029306,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21292088627815248,
      "num_tokens": 35472834.0,
      "step": 15890
    },
    {
      "entropy": 5.444349384307861,
      "epoch": 1.785196832706239,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004689569515824473,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21588788628578187,
      "num_tokens": 35482975.0,
      "step": 15895
    },
    {
      "entropy": 5.409834623336792,
      "epoch": 1.7857584096141967,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004689365988862995,
      "loss": 5.1399,
      "mean_token_accuracy": 0.20853421837091446,
      "num_tokens": 35494470.0,
      "step": 15900
    },
    {
      "entropy": 5.457488107681274,
      "epoch": 1.7863199865221542,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046891624001513515,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21182076632976532,
      "num_tokens": 35505464.0,
      "step": 15905
    },
    {
      "entropy": 5.460780143737793,
      "epoch": 1.7868815634301116,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046889587496960257,
      "loss": 5.1109,
      "mean_token_accuracy": 0.20508903563022612,
      "num_tokens": 35517609.0,
      "step": 15910
    },
    {
      "entropy": 5.499519109725952,
      "epoch": 1.7874431403380693,
      "grad_norm": 1.0,
      "learning_rate": 0.00046887550375034997,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21565255671739578,
      "num_tokens": 35528994.0,
      "step": 15915
    },
    {
      "entropy": 5.573452138900757,
      "epoch": 1.788004717246027,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046885512635802627,
      "loss": 5.2198,
      "mean_token_accuracy": 0.20077406316995622,
      "num_tokens": 35541044.0,
      "step": 15920
    },
    {
      "entropy": 5.4689469814300535,
      "epoch": 1.7885662941539844,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004688347427932801,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21211225539445877,
      "num_tokens": 35552660.0,
      "step": 15925
    },
    {
      "entropy": 5.576786184310913,
      "epoch": 1.7891278710619418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046881435305676055,
      "loss": 5.1463,
      "mean_token_accuracy": 0.2024679332971573,
      "num_tokens": 35565144.0,
      "step": 15930
    },
    {
      "entropy": 5.540687227249146,
      "epoch": 1.7896894479698995,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046879395714911686,
      "loss": 5.096,
      "mean_token_accuracy": 0.21010931581258774,
      "num_tokens": 35577067.0,
      "step": 15935
    },
    {
      "entropy": 5.4537200927734375,
      "epoch": 1.7902510248778571,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046877355507099846,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21291753351688386,
      "num_tokens": 35588712.0,
      "step": 15940
    },
    {
      "entropy": 5.493366718292236,
      "epoch": 1.7908126017858146,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004687531468230549,
      "loss": 5.0116,
      "mean_token_accuracy": 0.21327860951423644,
      "num_tokens": 35599212.0,
      "step": 15945
    },
    {
      "entropy": 5.518341588973999,
      "epoch": 1.791374178693772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004687327324059359,
      "loss": 5.1059,
      "mean_token_accuracy": 0.21302772015333177,
      "num_tokens": 35609397.0,
      "step": 15950
    },
    {
      "entropy": 5.470599269866943,
      "epoch": 1.7919357556017297,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004687123118202916,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21037781089544297,
      "num_tokens": 35620615.0,
      "step": 15955
    },
    {
      "entropy": 5.545458459854126,
      "epoch": 1.7924973325096873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004686918850667722,
      "loss": 5.1114,
      "mean_token_accuracy": 0.21522961109876632,
      "num_tokens": 35632684.0,
      "step": 15960
    },
    {
      "entropy": 5.51549334526062,
      "epoch": 1.7930589094176448,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000468671452146028,
      "loss": 5.1458,
      "mean_token_accuracy": 0.21134270578622819,
      "num_tokens": 35643896.0,
      "step": 15965
    },
    {
      "entropy": 5.512136363983155,
      "epoch": 1.7936204863256022,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046865101305870965,
      "loss": 5.1097,
      "mean_token_accuracy": 0.2080790564417839,
      "num_tokens": 35655752.0,
      "step": 15970
    },
    {
      "entropy": 5.439818668365478,
      "epoch": 1.7941820632335599,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000468630567805468,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21380633860826492,
      "num_tokens": 35665870.0,
      "step": 15975
    },
    {
      "entropy": 5.487493705749512,
      "epoch": 1.7947436401415173,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000468610116386954,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20478664934635163,
      "num_tokens": 35677792.0,
      "step": 15980
    },
    {
      "entropy": 5.454782009124756,
      "epoch": 1.7953052170494748,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004685896588038187,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20663830637931824,
      "num_tokens": 35689059.0,
      "step": 15985
    },
    {
      "entropy": 5.506741619110107,
      "epoch": 1.7958667939574324,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685691950567136,
      "loss": 5.1567,
      "mean_token_accuracy": 0.2077423095703125,
      "num_tokens": 35699953.0,
      "step": 15990
    },
    {
      "entropy": 5.539091634750366,
      "epoch": 1.79642837086539,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046854872514629035,
      "loss": 5.0747,
      "mean_token_accuracy": 0.21003518849611283,
      "num_tokens": 35710894.0,
      "step": 15995
    },
    {
      "entropy": 5.4831499576568605,
      "epoch": 1.7969899477733475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004685282490732005,
      "loss": 5.1394,
      "mean_token_accuracy": 0.2100980430841446,
      "num_tokens": 35722440.0,
      "step": 16000
    },
    {
      "entropy": 5.473070573806763,
      "epoch": 1.797551524681305,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046850776683809623,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20651376843452454,
      "num_tokens": 35733768.0,
      "step": 16005
    },
    {
      "entropy": 5.490332460403442,
      "epoch": 1.7981131015892626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004684872784416295,
      "loss": 5.1103,
      "mean_token_accuracy": 0.20746628195047379,
      "num_tokens": 35744211.0,
      "step": 16010
    },
    {
      "entropy": 5.474593305587769,
      "epoch": 1.7986746784972203,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004684667838844529,
      "loss": 5.1112,
      "mean_token_accuracy": 0.20941847711801528,
      "num_tokens": 35755498.0,
      "step": 16015
    },
    {
      "entropy": 5.5333967208862305,
      "epoch": 1.7992362554051777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004684462831672188,
      "loss": 5.1329,
      "mean_token_accuracy": 0.2138266995549202,
      "num_tokens": 35767442.0,
      "step": 16020
    },
    {
      "entropy": 5.512209749221801,
      "epoch": 1.7997978323131352,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046842577629058006,
      "loss": 5.0799,
      "mean_token_accuracy": 0.21625768542289733,
      "num_tokens": 35778040.0,
      "step": 16025
    },
    {
      "entropy": 5.4857195854187015,
      "epoch": 1.8003594092210928,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004684052632551896,
      "loss": 5.0551,
      "mean_token_accuracy": 0.2176490068435669,
      "num_tokens": 35788139.0,
      "step": 16030
    },
    {
      "entropy": 5.4839424133300785,
      "epoch": 1.8009209861290505,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046838474406170046,
      "loss": 5.1017,
      "mean_token_accuracy": 0.20668245404958724,
      "num_tokens": 35799077.0,
      "step": 16035
    },
    {
      "entropy": 5.520399045944214,
      "epoch": 1.801482563037008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046836421871076616,
      "loss": 5.1436,
      "mean_token_accuracy": 0.2059086248278618,
      "num_tokens": 35809748.0,
      "step": 16040
    },
    {
      "entropy": 5.597242021560669,
      "epoch": 1.8020441399449654,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004683436872030401,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20932105481624602,
      "num_tokens": 35821621.0,
      "step": 16045
    },
    {
      "entropy": 5.522021913528443,
      "epoch": 1.802605716852923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046832314953917605,
      "loss": 5.1167,
      "mean_token_accuracy": 0.2089120164513588,
      "num_tokens": 35833497.0,
      "step": 16050
    },
    {
      "entropy": 5.586480188369751,
      "epoch": 1.8031672937608807,
      "grad_norm": 1.0,
      "learning_rate": 0.00046830260571982795,
      "loss": 5.2226,
      "mean_token_accuracy": 0.2081691727042198,
      "num_tokens": 35844266.0,
      "step": 16055
    },
    {
      "entropy": 5.577649021148682,
      "epoch": 1.8037288706688381,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004682820557456499,
      "loss": 5.2079,
      "mean_token_accuracy": 0.2062609001994133,
      "num_tokens": 35855260.0,
      "step": 16060
    },
    {
      "entropy": 5.528383874893189,
      "epoch": 1.8042904475767956,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046826149961729625,
      "loss": 5.1606,
      "mean_token_accuracy": 0.20990426391363143,
      "num_tokens": 35865950.0,
      "step": 16065
    },
    {
      "entropy": 5.561335754394531,
      "epoch": 1.8048520244847532,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046824093733542144,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19930185228586197,
      "num_tokens": 35877028.0,
      "step": 16070
    },
    {
      "entropy": 5.496836614608765,
      "epoch": 1.8054136013927107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046822036890068033,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20585497915744783,
      "num_tokens": 35887831.0,
      "step": 16075
    },
    {
      "entropy": 5.513091897964477,
      "epoch": 1.8059751783006681,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046819979431372773,
      "loss": 5.1105,
      "mean_token_accuracy": 0.20903973877429963,
      "num_tokens": 35897946.0,
      "step": 16080
    },
    {
      "entropy": 5.5629009246826175,
      "epoch": 1.8065367552086258,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046817921357521865,
      "loss": 5.2464,
      "mean_token_accuracy": 0.20251297801733018,
      "num_tokens": 35910973.0,
      "step": 16085
    },
    {
      "entropy": 5.55006775856018,
      "epoch": 1.8070983321165834,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046815862668580854,
      "loss": 5.18,
      "mean_token_accuracy": 0.20721229910850525,
      "num_tokens": 35921478.0,
      "step": 16090
    },
    {
      "entropy": 5.523074197769165,
      "epoch": 1.8076599090245409,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004681380336461528,
      "loss": 5.0508,
      "mean_token_accuracy": 0.21772347688674926,
      "num_tokens": 35932211.0,
      "step": 16095
    },
    {
      "entropy": 5.48393931388855,
      "epoch": 1.8082214859324983,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046811743445690724,
      "loss": 5.2093,
      "mean_token_accuracy": 0.20707104206085206,
      "num_tokens": 35944350.0,
      "step": 16100
    },
    {
      "entropy": 5.5646586418151855,
      "epoch": 1.808783062840456,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004680968291187276,
      "loss": 5.2008,
      "mean_token_accuracy": 0.20393282026052476,
      "num_tokens": 35955590.0,
      "step": 16105
    },
    {
      "entropy": 5.4751293659210205,
      "epoch": 1.8093446397484136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004680762176322701,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22462885230779647,
      "num_tokens": 35965872.0,
      "step": 16110
    },
    {
      "entropy": 5.426794910430909,
      "epoch": 1.809906216656371,
      "grad_norm": 1.0,
      "learning_rate": 0.00046805559999819093,
      "loss": 5.1048,
      "mean_token_accuracy": 0.203200101852417,
      "num_tokens": 35976851.0,
      "step": 16115
    },
    {
      "entropy": 5.51524076461792,
      "epoch": 1.8104677935643285,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004680349762171465,
      "loss": 5.0702,
      "mean_token_accuracy": 0.20514808744192123,
      "num_tokens": 35986984.0,
      "step": 16120
    },
    {
      "entropy": 5.527813577651978,
      "epoch": 1.8110293704722862,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004680143462897936,
      "loss": 5.1405,
      "mean_token_accuracy": 0.2088669568300247,
      "num_tokens": 35999349.0,
      "step": 16125
    },
    {
      "entropy": 5.467396688461304,
      "epoch": 1.8115909473802438,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004679937102167891,
      "loss": 5.1537,
      "mean_token_accuracy": 0.21317908465862273,
      "num_tokens": 36011122.0,
      "step": 16130
    },
    {
      "entropy": 5.517142963409424,
      "epoch": 1.8121525242882013,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046797306799879,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20882433652877808,
      "num_tokens": 36022729.0,
      "step": 16135
    },
    {
      "entropy": 5.473223686218262,
      "epoch": 1.8127141011961587,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046795241963645356,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20791100561618805,
      "num_tokens": 36034042.0,
      "step": 16140
    },
    {
      "entropy": 5.478951930999756,
      "epoch": 1.8132756781041164,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004679317651304373,
      "loss": 5.1306,
      "mean_token_accuracy": 0.20588796883821486,
      "num_tokens": 36045357.0,
      "step": 16145
    },
    {
      "entropy": 5.385507535934448,
      "epoch": 1.813837255012074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004679111044813987,
      "loss": 4.8927,
      "mean_token_accuracy": 0.21933177560567857,
      "num_tokens": 36055180.0,
      "step": 16150
    },
    {
      "entropy": 5.4356471538543705,
      "epoch": 1.8143988319200315,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046789043768999573,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21692266762256623,
      "num_tokens": 36065574.0,
      "step": 16155
    },
    {
      "entropy": 5.474743270874024,
      "epoch": 1.814960408827989,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004678697647568865,
      "loss": 5.1266,
      "mean_token_accuracy": 0.2107786700129509,
      "num_tokens": 36077316.0,
      "step": 16160
    },
    {
      "entropy": 5.461124753952026,
      "epoch": 1.8155219857359466,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046784908568272903,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20999224334955216,
      "num_tokens": 36088735.0,
      "step": 16165
    },
    {
      "entropy": 5.49382643699646,
      "epoch": 1.8160835626439042,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000467828400468182,
      "loss": 5.1175,
      "mean_token_accuracy": 0.20820717215538026,
      "num_tokens": 36098806.0,
      "step": 16170
    },
    {
      "entropy": 5.48393406867981,
      "epoch": 1.8166451395518615,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046780770911390387,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20794095396995543,
      "num_tokens": 36110333.0,
      "step": 16175
    },
    {
      "entropy": 5.5452765941619875,
      "epoch": 1.8172067164598191,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004677870116205534,
      "loss": 5.12,
      "mean_token_accuracy": 0.2085841938853264,
      "num_tokens": 36121719.0,
      "step": 16180
    },
    {
      "entropy": 5.491519832611084,
      "epoch": 1.8177682933677768,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046776630798878985,
      "loss": 5.0319,
      "mean_token_accuracy": 0.21408172845840454,
      "num_tokens": 36132129.0,
      "step": 16185
    },
    {
      "entropy": 5.586021089553833,
      "epoch": 1.8183298702757342,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004677455982192723,
      "loss": 5.3002,
      "mean_token_accuracy": 0.20487173050642013,
      "num_tokens": 36145189.0,
      "step": 16190
    },
    {
      "entropy": 5.474322319030762,
      "epoch": 1.8188914471836917,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046772488231266004,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20783973038196563,
      "num_tokens": 36157896.0,
      "step": 16195
    },
    {
      "entropy": 5.5622375965118405,
      "epoch": 1.8194530240916493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046770416026961284,
      "loss": 5.1676,
      "mean_token_accuracy": 0.2086881071329117,
      "num_tokens": 36168842.0,
      "step": 16200
    },
    {
      "entropy": 5.40425214767456,
      "epoch": 1.820014600999607,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004676834320907904,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21493154913187026,
      "num_tokens": 36180075.0,
      "step": 16205
    },
    {
      "entropy": 5.4317741870880125,
      "epoch": 1.8205761779075644,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004676626977768528,
      "loss": 5.0696,
      "mean_token_accuracy": 0.21408002227544784,
      "num_tokens": 36190589.0,
      "step": 16210
    },
    {
      "entropy": 5.448160791397095,
      "epoch": 1.8211377548155219,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046764195732846013,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21916674971580505,
      "num_tokens": 36201145.0,
      "step": 16215
    },
    {
      "entropy": 5.431320524215698,
      "epoch": 1.8216993317234795,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046762121074627285,
      "loss": 5.083,
      "mean_token_accuracy": 0.21231744140386583,
      "num_tokens": 36213243.0,
      "step": 16220
    },
    {
      "entropy": 5.372766590118408,
      "epoch": 1.8222609086314372,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046760045803095153,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21231216788291932,
      "num_tokens": 36224491.0,
      "step": 16225
    },
    {
      "entropy": 5.435123443603516,
      "epoch": 1.8228224855393946,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004675796991831569,
      "loss": 5.0989,
      "mean_token_accuracy": 0.2125253811478615,
      "num_tokens": 36236006.0,
      "step": 16230
    },
    {
      "entropy": 5.611803483963013,
      "epoch": 1.823384062447352,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046755893420355,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20481275618076325,
      "num_tokens": 36246925.0,
      "step": 16235
    },
    {
      "entropy": 5.529128360748291,
      "epoch": 1.8239456393553097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004675381630927919,
      "loss": 5.1179,
      "mean_token_accuracy": 0.2053695559501648,
      "num_tokens": 36258449.0,
      "step": 16240
    },
    {
      "entropy": 5.510815000534057,
      "epoch": 1.8245072162632674,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046751738585154405,
      "loss": 5.1142,
      "mean_token_accuracy": 0.21405246257781982,
      "num_tokens": 36270820.0,
      "step": 16245
    },
    {
      "entropy": 5.552372837066651,
      "epoch": 1.8250687931712248,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046749660248046794,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20924902707338333,
      "num_tokens": 36282346.0,
      "step": 16250
    },
    {
      "entropy": 5.551780843734742,
      "epoch": 1.8256303700791823,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046747581298022537,
      "loss": 5.2073,
      "mean_token_accuracy": 0.20106985569000244,
      "num_tokens": 36294111.0,
      "step": 16255
    },
    {
      "entropy": 5.476664686203003,
      "epoch": 1.82619194698714,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004674550173514782,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20898337811231613,
      "num_tokens": 36304899.0,
      "step": 16260
    },
    {
      "entropy": 5.503116321563721,
      "epoch": 1.8267535238950976,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046743421559488875,
      "loss": 5.195,
      "mean_token_accuracy": 0.206307815015316,
      "num_tokens": 36316272.0,
      "step": 16265
    },
    {
      "entropy": 5.474429273605347,
      "epoch": 1.8273151008030548,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046741340771111917,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21395708471536637,
      "num_tokens": 36326624.0,
      "step": 16270
    },
    {
      "entropy": 5.4546449184417725,
      "epoch": 1.8278766777110125,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046739259370083205,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21640014350414277,
      "num_tokens": 36338178.0,
      "step": 16275
    },
    {
      "entropy": 5.571144962310791,
      "epoch": 1.8284382546189701,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004673717735646902,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20726884603500367,
      "num_tokens": 36349993.0,
      "step": 16280
    },
    {
      "entropy": 5.575223827362061,
      "epoch": 1.8289998315269276,
      "grad_norm": 0.875,
      "learning_rate": 0.0004673509473033564,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20759999752044678,
      "num_tokens": 36361002.0,
      "step": 16285
    },
    {
      "entropy": 5.514489269256591,
      "epoch": 1.829561408434885,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046733011491749383,
      "loss": 5.1271,
      "mean_token_accuracy": 0.20994158685207367,
      "num_tokens": 36371688.0,
      "step": 16290
    },
    {
      "entropy": 5.434833145141601,
      "epoch": 1.8301229853428427,
      "grad_norm": 1.0,
      "learning_rate": 0.0004673092764077658,
      "loss": 5.0646,
      "mean_token_accuracy": 0.2152921661734581,
      "num_tokens": 36382683.0,
      "step": 16295
    },
    {
      "entropy": 5.5578593730926515,
      "epoch": 1.8306845622508003,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004672884317748359,
      "loss": 5.207,
      "mean_token_accuracy": 0.19801284670829772,
      "num_tokens": 36394170.0,
      "step": 16300
    },
    {
      "entropy": 5.535651779174804,
      "epoch": 1.8312461391587578,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004672675810193677,
      "loss": 5.1318,
      "mean_token_accuracy": 0.2031713142991066,
      "num_tokens": 36405613.0,
      "step": 16305
    },
    {
      "entropy": 5.547235202789307,
      "epoch": 1.8318077160667152,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046724672414202514,
      "loss": 5.1847,
      "mean_token_accuracy": 0.2065509647130966,
      "num_tokens": 36416282.0,
      "step": 16310
    },
    {
      "entropy": 5.534456539154053,
      "epoch": 1.8323692929746729,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004672258611434724,
      "loss": 5.1955,
      "mean_token_accuracy": 0.20271504819393157,
      "num_tokens": 36427624.0,
      "step": 16315
    },
    {
      "entropy": 5.480709505081177,
      "epoch": 1.8329308698826305,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046720499202437354,
      "loss": 5.0025,
      "mean_token_accuracy": 0.22154767364263533,
      "num_tokens": 36438230.0,
      "step": 16320
    },
    {
      "entropy": 5.499330759048462,
      "epoch": 1.833492446790588,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004671841167853933,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20689167082309723,
      "num_tokens": 36448854.0,
      "step": 16325
    },
    {
      "entropy": 5.523745155334472,
      "epoch": 1.8340540236985454,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004671632354271962,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20888641327619553,
      "num_tokens": 36460426.0,
      "step": 16330
    },
    {
      "entropy": 5.563951730728149,
      "epoch": 1.834615600606503,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046714234795044724,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20701346099376677,
      "num_tokens": 36470914.0,
      "step": 16335
    },
    {
      "entropy": 5.489262104034424,
      "epoch": 1.8351771775144607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004671214543558113,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2084989070892334,
      "num_tokens": 36480753.0,
      "step": 16340
    },
    {
      "entropy": 5.445824337005615,
      "epoch": 1.8357387544224182,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004671005546439538,
      "loss": 5.0406,
      "mean_token_accuracy": 0.2117574319243431,
      "num_tokens": 36492858.0,
      "step": 16345
    },
    {
      "entropy": 5.399022817611694,
      "epoch": 1.8363003313303756,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046707964881554005,
      "loss": 4.9979,
      "mean_token_accuracy": 0.2150909811258316,
      "num_tokens": 36503912.0,
      "step": 16350
    },
    {
      "entropy": 5.462345743179322,
      "epoch": 1.8368619082383333,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046705873687123586,
      "loss": 5.0765,
      "mean_token_accuracy": 0.2098548948764801,
      "num_tokens": 36515410.0,
      "step": 16355
    },
    {
      "entropy": 5.486423492431641,
      "epoch": 1.837423485146291,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046703781881170703,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21367933750152587,
      "num_tokens": 36526310.0,
      "step": 16360
    },
    {
      "entropy": 5.43696780204773,
      "epoch": 1.8379850620542482,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046701689463761947,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21390655040740966,
      "num_tokens": 36537081.0,
      "step": 16365
    },
    {
      "entropy": 5.450550699234009,
      "epoch": 1.8385466389622058,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046699596434963953,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21097376197576523,
      "num_tokens": 36549419.0,
      "step": 16370
    },
    {
      "entropy": 5.407167291641235,
      "epoch": 1.8391082158701635,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004669750279484337,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21898609548807144,
      "num_tokens": 36560842.0,
      "step": 16375
    },
    {
      "entropy": 5.468007230758667,
      "epoch": 1.839669792778121,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004669540854346685,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21070052534341813,
      "num_tokens": 36571717.0,
      "step": 16380
    },
    {
      "entropy": 5.503949975967407,
      "epoch": 1.8402313696860784,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046693313680901073,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21168841123580934,
      "num_tokens": 36581709.0,
      "step": 16385
    },
    {
      "entropy": 5.448351860046387,
      "epoch": 1.840792946594036,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046691218207212744,
      "loss": 5.1023,
      "mean_token_accuracy": 0.2082774430513382,
      "num_tokens": 36593033.0,
      "step": 16390
    },
    {
      "entropy": 5.498321866989135,
      "epoch": 1.8413545235019937,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004668912212246859,
      "loss": 5.1738,
      "mean_token_accuracy": 0.210365329682827,
      "num_tokens": 36603605.0,
      "step": 16395
    },
    {
      "entropy": 5.533682012557984,
      "epoch": 1.8419161004099511,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046687025426735335,
      "loss": 5.191,
      "mean_token_accuracy": 0.20624493509531022,
      "num_tokens": 36616219.0,
      "step": 16400
    },
    {
      "entropy": 5.605037593841553,
      "epoch": 1.8424776773179086,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004668492812007976,
      "loss": 5.1406,
      "mean_token_accuracy": 0.21203456223011016,
      "num_tokens": 36627777.0,
      "step": 16405
    },
    {
      "entropy": 5.495413827896118,
      "epoch": 1.8430392542258662,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004668283020256863,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2118648812174797,
      "num_tokens": 36638792.0,
      "step": 16410
    },
    {
      "entropy": 5.50250654220581,
      "epoch": 1.843600831133824,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046680731674268744,
      "loss": 5.1166,
      "mean_token_accuracy": 0.2120824784040451,
      "num_tokens": 36650218.0,
      "step": 16415
    },
    {
      "entropy": 5.504008388519287,
      "epoch": 1.8441624080417813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004667863253524692,
      "loss": 5.1004,
      "mean_token_accuracy": 0.209670090675354,
      "num_tokens": 36660987.0,
      "step": 16420
    },
    {
      "entropy": 5.484678792953491,
      "epoch": 1.8447239849497388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046676532785569995,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2035755068063736,
      "num_tokens": 36671270.0,
      "step": 16425
    },
    {
      "entropy": 5.462496709823609,
      "epoch": 1.8452855618576964,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004667443242530484,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21600109189748765,
      "num_tokens": 36682484.0,
      "step": 16430
    },
    {
      "entropy": 5.52449951171875,
      "epoch": 1.845847138765654,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046672331454518314,
      "loss": 5.1472,
      "mean_token_accuracy": 0.21022522151470185,
      "num_tokens": 36693723.0,
      "step": 16435
    },
    {
      "entropy": 5.423861646652222,
      "epoch": 1.8464087156736115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004667022987327732,
      "loss": 4.9707,
      "mean_token_accuracy": 0.2191660761833191,
      "num_tokens": 36704166.0,
      "step": 16440
    },
    {
      "entropy": 5.442873620986939,
      "epoch": 1.846970292581569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004666812768164877,
      "loss": 5.0316,
      "mean_token_accuracy": 0.2140076130628586,
      "num_tokens": 36714139.0,
      "step": 16445
    },
    {
      "entropy": 5.5038642406463625,
      "epoch": 1.8475318694895266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046666024879699606,
      "loss": 5.0678,
      "mean_token_accuracy": 0.21221577227115632,
      "num_tokens": 36724273.0,
      "step": 16450
    },
    {
      "entropy": 5.505935382843018,
      "epoch": 1.8480934463974843,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046663921467496766,
      "loss": 5.132,
      "mean_token_accuracy": 0.20436228066682816,
      "num_tokens": 36735176.0,
      "step": 16455
    },
    {
      "entropy": 5.422377729415894,
      "epoch": 1.8486550233054415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004666181744510724,
      "loss": 5.0635,
      "mean_token_accuracy": 0.21245591342449188,
      "num_tokens": 36746170.0,
      "step": 16460
    },
    {
      "entropy": 5.571506643295288,
      "epoch": 1.8492166002133992,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046659712812598017,
      "loss": 5.2286,
      "mean_token_accuracy": 0.20462593734264373,
      "num_tokens": 36756969.0,
      "step": 16465
    },
    {
      "entropy": 5.593419694900513,
      "epoch": 1.8497781771213568,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046657607570036103,
      "loss": 5.0746,
      "mean_token_accuracy": 0.21414605677127838,
      "num_tokens": 36767892.0,
      "step": 16470
    },
    {
      "entropy": 5.501107597351075,
      "epoch": 1.8503397540293143,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046655501717488533,
      "loss": 5.0969,
      "mean_token_accuracy": 0.21075148731470109,
      "num_tokens": 36778967.0,
      "step": 16475
    },
    {
      "entropy": 5.370328617095947,
      "epoch": 1.8509013309372717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004665339525502236,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2118169277906418,
      "num_tokens": 36789204.0,
      "step": 16480
    },
    {
      "entropy": 5.4745416164398195,
      "epoch": 1.8514629078452294,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046651288182704653,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21928381621837617,
      "num_tokens": 36798990.0,
      "step": 16485
    },
    {
      "entropy": 5.554029130935669,
      "epoch": 1.852024484753187,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004664918050060251,
      "loss": 5.138,
      "mean_token_accuracy": 0.2112487867474556,
      "num_tokens": 36809988.0,
      "step": 16490
    },
    {
      "entropy": 5.463080310821534,
      "epoch": 1.8525860616611445,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004664707220878302,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21656173020601271,
      "num_tokens": 36821579.0,
      "step": 16495
    },
    {
      "entropy": 5.562678432464599,
      "epoch": 1.853147638569102,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004664496330731334,
      "loss": 5.1856,
      "mean_token_accuracy": 0.21041040867567062,
      "num_tokens": 36833141.0,
      "step": 16500
    },
    {
      "entropy": 5.543489265441894,
      "epoch": 1.8537092154770596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004664285379626059,
      "loss": 5.1787,
      "mean_token_accuracy": 0.21282510459423065,
      "num_tokens": 36843347.0,
      "step": 16505
    },
    {
      "entropy": 5.5333380699157715,
      "epoch": 1.8542707923850172,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046640743675691954,
      "loss": 5.2258,
      "mean_token_accuracy": 0.1979595437645912,
      "num_tokens": 36854459.0,
      "step": 16510
    },
    {
      "entropy": 5.4644242286682125,
      "epoch": 1.8548323692929747,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004663863294567462,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21523868292570114,
      "num_tokens": 36864761.0,
      "step": 16515
    },
    {
      "entropy": 5.611645174026489,
      "epoch": 1.8553939462009321,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004663652160627579,
      "loss": 5.2336,
      "mean_token_accuracy": 0.2005768299102783,
      "num_tokens": 36875413.0,
      "step": 16520
    },
    {
      "entropy": 5.497026681900024,
      "epoch": 1.8559555231088898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046634409657562683,
      "loss": 5.1201,
      "mean_token_accuracy": 0.20424643754959107,
      "num_tokens": 36885350.0,
      "step": 16525
    },
    {
      "entropy": 5.477964973449707,
      "epoch": 1.8565171000168474,
      "grad_norm": 1.0,
      "learning_rate": 0.0004663229709960255,
      "loss": 5.0174,
      "mean_token_accuracy": 0.21442075818777084,
      "num_tokens": 36895355.0,
      "step": 16530
    },
    {
      "entropy": 5.558397912979126,
      "epoch": 1.8570786769248049,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046630183932462664,
      "loss": 5.1156,
      "mean_token_accuracy": 0.21687323451042176,
      "num_tokens": 36907454.0,
      "step": 16535
    },
    {
      "entropy": 5.508376884460449,
      "epoch": 1.8576402538327623,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004662807015621031,
      "loss": 5.1136,
      "mean_token_accuracy": 0.20426612347364426,
      "num_tokens": 36919199.0,
      "step": 16540
    },
    {
      "entropy": 5.569764471054077,
      "epoch": 1.85820183074072,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046625955770912764,
      "loss": 5.2022,
      "mean_token_accuracy": 0.2058103322982788,
      "num_tokens": 36930499.0,
      "step": 16545
    },
    {
      "entropy": 5.491767454147339,
      "epoch": 1.8587634076486776,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046623840776637384,
      "loss": 5.1276,
      "mean_token_accuracy": 0.2030834972858429,
      "num_tokens": 36943082.0,
      "step": 16550
    },
    {
      "entropy": 5.467373514175415,
      "epoch": 1.859324984556635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004662172517345149,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2111891523003578,
      "num_tokens": 36954591.0,
      "step": 16555
    },
    {
      "entropy": 5.481186199188232,
      "epoch": 1.8598865614645925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004661960896142245,
      "loss": 5.1174,
      "mean_token_accuracy": 0.21151324659585952,
      "num_tokens": 36967028.0,
      "step": 16560
    },
    {
      "entropy": 5.484088325500489,
      "epoch": 1.8604481383725502,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046617492140617654,
      "loss": 5.1595,
      "mean_token_accuracy": 0.20402714759111404,
      "num_tokens": 36978716.0,
      "step": 16565
    },
    {
      "entropy": 5.447563600540161,
      "epoch": 1.8610097152805076,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004661537471110448,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2113659292459488,
      "num_tokens": 36989416.0,
      "step": 16570
    },
    {
      "entropy": 5.493163204193115,
      "epoch": 1.861571292188465,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004661325667295037,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21947369277477263,
      "num_tokens": 37002197.0,
      "step": 16575
    },
    {
      "entropy": 5.635228061676026,
      "epoch": 1.8621328690964227,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004661113802622275,
      "loss": 5.1967,
      "mean_token_accuracy": 0.19943866431713103,
      "num_tokens": 37012977.0,
      "step": 16580
    },
    {
      "entropy": 5.492608690261841,
      "epoch": 1.8626944460043804,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004660901877098908,
      "loss": 5.107,
      "mean_token_accuracy": 0.20855230540037156,
      "num_tokens": 37024141.0,
      "step": 16585
    },
    {
      "entropy": 5.475650882720947,
      "epoch": 1.8632560229123378,
      "grad_norm": 1.0,
      "learning_rate": 0.0004660689890731685,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21789149492979049,
      "num_tokens": 37034471.0,
      "step": 16590
    },
    {
      "entropy": 5.519156265258789,
      "epoch": 1.8638175998202953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046604778435273546,
      "loss": 5.0897,
      "mean_token_accuracy": 0.2132817506790161,
      "num_tokens": 37046375.0,
      "step": 16595
    },
    {
      "entropy": 5.50539608001709,
      "epoch": 1.864379176728253,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046602657354926686,
      "loss": 5.1191,
      "mean_token_accuracy": 0.2123912066221237,
      "num_tokens": 37056617.0,
      "step": 16600
    },
    {
      "entropy": 5.383263063430786,
      "epoch": 1.8649407536362106,
      "grad_norm": 1.09375,
      "learning_rate": 0.000466005356663438,
      "loss": 4.9568,
      "mean_token_accuracy": 0.21887307614088058,
      "num_tokens": 37066477.0,
      "step": 16605
    },
    {
      "entropy": 5.4798565864562985,
      "epoch": 1.865502330544168,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046598413369592463,
      "loss": 5.135,
      "mean_token_accuracy": 0.2076435551047325,
      "num_tokens": 37077150.0,
      "step": 16610
    },
    {
      "entropy": 5.545440149307251,
      "epoch": 1.8660639074521255,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004659629046474023,
      "loss": 5.1636,
      "mean_token_accuracy": 0.21037871688604354,
      "num_tokens": 37089961.0,
      "step": 16615
    },
    {
      "entropy": 5.560234498977661,
      "epoch": 1.8666254843600831,
      "grad_norm": 0.984375,
      "learning_rate": 0.000465941669518547,
      "loss": 5.1731,
      "mean_token_accuracy": 0.20456912517547607,
      "num_tokens": 37100779.0,
      "step": 16620
    },
    {
      "entropy": 5.486607885360717,
      "epoch": 1.8671870612680408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004659204283100349,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2115027442574501,
      "num_tokens": 37111293.0,
      "step": 16625
    },
    {
      "entropy": 5.506576585769653,
      "epoch": 1.8677486381759982,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046589918102254237,
      "loss": 5.1206,
      "mean_token_accuracy": 0.20869044959545135,
      "num_tokens": 37123087.0,
      "step": 16630
    },
    {
      "entropy": 5.51568078994751,
      "epoch": 1.8683102150839557,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004658779276567458,
      "loss": 5.1069,
      "mean_token_accuracy": 0.21024658381938935,
      "num_tokens": 37135227.0,
      "step": 16635
    },
    {
      "entropy": 5.507825756072998,
      "epoch": 1.8688717919919133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046585666821332203,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21268498301506042,
      "num_tokens": 37146970.0,
      "step": 16640
    },
    {
      "entropy": 5.528319787979126,
      "epoch": 1.869433368899871,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046583540269294794,
      "loss": 5.0871,
      "mean_token_accuracy": 0.20527649372816087,
      "num_tokens": 37158144.0,
      "step": 16645
    },
    {
      "entropy": 5.54538927078247,
      "epoch": 1.8699949458078284,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004658141310963006,
      "loss": 5.1933,
      "mean_token_accuracy": 0.206640787422657,
      "num_tokens": 37169265.0,
      "step": 16650
    },
    {
      "entropy": 5.535696029663086,
      "epoch": 1.8705565227157859,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004657928534240573,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20693681091070176,
      "num_tokens": 37179186.0,
      "step": 16655
    },
    {
      "entropy": 5.588586902618408,
      "epoch": 1.8711180996237435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004657715696768955,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2120760276913643,
      "num_tokens": 37188906.0,
      "step": 16660
    },
    {
      "entropy": 5.452589893341065,
      "epoch": 1.871679676531701,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000465750279855493,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2205213263630867,
      "num_tokens": 37200879.0,
      "step": 16665
    },
    {
      "entropy": 5.486661577224732,
      "epoch": 1.8722412534396584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004657289839605276,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21644734889268874,
      "num_tokens": 37211648.0,
      "step": 16670
    },
    {
      "entropy": 5.582305097579956,
      "epoch": 1.872802830347616,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004657076819926774,
      "loss": 5.2132,
      "mean_token_accuracy": 0.21304583102464675,
      "num_tokens": 37223164.0,
      "step": 16675
    },
    {
      "entropy": 5.517575216293335,
      "epoch": 1.8733644072555737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004656863739526206,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20865258276462556,
      "num_tokens": 37233860.0,
      "step": 16680
    },
    {
      "entropy": 5.565133047103882,
      "epoch": 1.8739259841635312,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046566505984103575,
      "loss": 5.0521,
      "mean_token_accuracy": 0.2139038026332855,
      "num_tokens": 37245431.0,
      "step": 16685
    },
    {
      "entropy": 5.549070358276367,
      "epoch": 1.8744875610714886,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004656437396586014,
      "loss": 5.1869,
      "mean_token_accuracy": 0.20393299609422683,
      "num_tokens": 37256394.0,
      "step": 16690
    },
    {
      "entropy": 5.483378124237061,
      "epoch": 1.8750491379794463,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004656224134059964,
      "loss": 5.1557,
      "mean_token_accuracy": 0.20865960270166398,
      "num_tokens": 37268367.0,
      "step": 16695
    },
    {
      "entropy": 5.519014883041382,
      "epoch": 1.875610714887404,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046560108108389985,
      "loss": 5.1298,
      "mean_token_accuracy": 0.21005660891532899,
      "num_tokens": 37279019.0,
      "step": 16700
    },
    {
      "entropy": 5.55357871055603,
      "epoch": 1.8761722917953614,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000465579742692991,
      "loss": 5.0523,
      "mean_token_accuracy": 0.2132042810320854,
      "num_tokens": 37289806.0,
      "step": 16705
    },
    {
      "entropy": 5.420429039001465,
      "epoch": 1.8767338687033188,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004655583982339492,
      "loss": 5.1031,
      "mean_token_accuracy": 0.20848310440778733,
      "num_tokens": 37300607.0,
      "step": 16710
    },
    {
      "entropy": 5.4932701587677,
      "epoch": 1.8772954456112765,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004655370477074541,
      "loss": 5.1656,
      "mean_token_accuracy": 0.20755331814289094,
      "num_tokens": 37311966.0,
      "step": 16715
    },
    {
      "entropy": 5.414129447937012,
      "epoch": 1.8778570225192341,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004655156911141855,
      "loss": 5.0849,
      "mean_token_accuracy": 0.21028892397880555,
      "num_tokens": 37324459.0,
      "step": 16720
    },
    {
      "entropy": 5.528072500228882,
      "epoch": 1.8784185994271916,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004654943284548234,
      "loss": 5.0823,
      "mean_token_accuracy": 0.20873494893312455,
      "num_tokens": 37335276.0,
      "step": 16725
    },
    {
      "entropy": 5.498196125030518,
      "epoch": 1.878980176335149,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046547295973004807,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21564596891403198,
      "num_tokens": 37347274.0,
      "step": 16730
    },
    {
      "entropy": 5.469661903381348,
      "epoch": 1.8795417532431067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046545158494053966,
      "loss": 5.1333,
      "mean_token_accuracy": 0.20145306140184402,
      "num_tokens": 37359789.0,
      "step": 16735
    },
    {
      "entropy": 5.505513286590576,
      "epoch": 1.8801033301510643,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046543020408697905,
      "loss": 5.0895,
      "mean_token_accuracy": 0.21216436326503754,
      "num_tokens": 37369851.0,
      "step": 16740
    },
    {
      "entropy": 5.58355827331543,
      "epoch": 1.8806649070590218,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046540881717004685,
      "loss": 5.2496,
      "mean_token_accuracy": 0.2018754482269287,
      "num_tokens": 37381008.0,
      "step": 16745
    },
    {
      "entropy": 5.463074588775635,
      "epoch": 1.8812264839669792,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004653874241904241,
      "loss": 5.0948,
      "mean_token_accuracy": 0.20884383767843245,
      "num_tokens": 37392110.0,
      "step": 16750
    },
    {
      "entropy": 5.530291032791138,
      "epoch": 1.8817880608749369,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004653660251487919,
      "loss": 5.0907,
      "mean_token_accuracy": 0.2167768210172653,
      "num_tokens": 37402995.0,
      "step": 16755
    },
    {
      "entropy": 5.522855472564697,
      "epoch": 1.8823496377828943,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046534462004583154,
      "loss": 5.1128,
      "mean_token_accuracy": 0.2122217521071434,
      "num_tokens": 37415072.0,
      "step": 16760
    },
    {
      "entropy": 5.578861284255981,
      "epoch": 1.8829112146908518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004653232088822248,
      "loss": 5.2731,
      "mean_token_accuracy": 0.196064293384552,
      "num_tokens": 37427375.0,
      "step": 16765
    },
    {
      "entropy": 5.420341682434082,
      "epoch": 1.8834727915988094,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004653017916586532,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21561141759157182,
      "num_tokens": 37436829.0,
      "step": 16770
    },
    {
      "entropy": 5.523181486129761,
      "epoch": 1.884034368506767,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004652803683757987,
      "loss": 5.1476,
      "mean_token_accuracy": 0.19952462166547774,
      "num_tokens": 37447875.0,
      "step": 16775
    },
    {
      "entropy": 5.442384910583496,
      "epoch": 1.8845959454147245,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004652589390343436,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21096417605876921,
      "num_tokens": 37458408.0,
      "step": 16780
    },
    {
      "entropy": 5.465321254730225,
      "epoch": 1.885157522322682,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046523750363497,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20720913559198378,
      "num_tokens": 37468437.0,
      "step": 16785
    },
    {
      "entropy": 5.572143793106079,
      "epoch": 1.8857190992306396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004652160621783605,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20912869423627853,
      "num_tokens": 37479359.0,
      "step": 16790
    },
    {
      "entropy": 5.463993740081787,
      "epoch": 1.8862806761385973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046519461466519785,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21512609273195266,
      "num_tokens": 37491321.0,
      "step": 16795
    },
    {
      "entropy": 5.549176549911499,
      "epoch": 1.8868422530465547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004651731610961649,
      "loss": 5.0916,
      "mean_token_accuracy": 0.21092260330915452,
      "num_tokens": 37501828.0,
      "step": 16800
    },
    {
      "entropy": 5.613136863708496,
      "epoch": 1.8874038299545122,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004651517014719446,
      "loss": 5.2364,
      "mean_token_accuracy": 0.20484791845083236,
      "num_tokens": 37514525.0,
      "step": 16805
    },
    {
      "entropy": 5.514158487319946,
      "epoch": 1.8879654068624698,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004651302357932206,
      "loss": 5.035,
      "mean_token_accuracy": 0.20797241926193238,
      "num_tokens": 37525588.0,
      "step": 16810
    },
    {
      "entropy": 5.467323303222656,
      "epoch": 1.8885269837704275,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000465108764060676,
      "loss": 5.0659,
      "mean_token_accuracy": 0.21335967779159545,
      "num_tokens": 37537941.0,
      "step": 16815
    },
    {
      "entropy": 5.5093378067016605,
      "epoch": 1.889088560678385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046508728627499464,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20967621803283693,
      "num_tokens": 37549663.0,
      "step": 16820
    },
    {
      "entropy": 5.563655233383178,
      "epoch": 1.8896501375863424,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004650658024368604,
      "loss": 5.1653,
      "mean_token_accuracy": 0.21119212955236435,
      "num_tokens": 37559793.0,
      "step": 16825
    },
    {
      "entropy": 5.479906988143921,
      "epoch": 1.8902117144943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004650443125469573,
      "loss": 5.0316,
      "mean_token_accuracy": 0.21199899017810822,
      "num_tokens": 37570203.0,
      "step": 16830
    },
    {
      "entropy": 5.43759937286377,
      "epoch": 1.8907732914022577,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004650228166059695,
      "loss": 5.1249,
      "mean_token_accuracy": 0.21079009026288986,
      "num_tokens": 37581858.0,
      "step": 16835
    },
    {
      "entropy": 5.470770835876465,
      "epoch": 1.8913348683102151,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004650013146145816,
      "loss": 5.0701,
      "mean_token_accuracy": 0.2109155386686325,
      "num_tokens": 37592311.0,
      "step": 16840
    },
    {
      "entropy": 5.594961833953858,
      "epoch": 1.8918964452181726,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004649798065734781,
      "loss": 5.2281,
      "mean_token_accuracy": 0.20618222951889037,
      "num_tokens": 37604721.0,
      "step": 16845
    },
    {
      "entropy": 5.6119483470916744,
      "epoch": 1.8924580221261302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046495829248334393,
      "loss": 5.2977,
      "mean_token_accuracy": 0.2009274885058403,
      "num_tokens": 37616044.0,
      "step": 16850
    },
    {
      "entropy": 5.494494915008545,
      "epoch": 1.8930195990340877,
      "grad_norm": 1.0,
      "learning_rate": 0.0004649367723448641,
      "loss": 5.1662,
      "mean_token_accuracy": 0.21000045388936997,
      "num_tokens": 37626501.0,
      "step": 16855
    },
    {
      "entropy": 5.579296350479126,
      "epoch": 1.8935811759420451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046491524615872366,
      "loss": 5.1485,
      "mean_token_accuracy": 0.2096484810113907,
      "num_tokens": 37637861.0,
      "step": 16860
    },
    {
      "entropy": 5.455595636367798,
      "epoch": 1.8941427528500028,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046489371392560816,
      "loss": 5.002,
      "mean_token_accuracy": 0.21168086528778077,
      "num_tokens": 37649288.0,
      "step": 16865
    },
    {
      "entropy": 5.422808980941772,
      "epoch": 1.8947043297579604,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046487217564620317,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21992223411798478,
      "num_tokens": 37659700.0,
      "step": 16870
    },
    {
      "entropy": 5.500381565093994,
      "epoch": 1.8952659066659179,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004648506313211945,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20493011325597763,
      "num_tokens": 37671739.0,
      "step": 16875
    },
    {
      "entropy": 5.544111347198486,
      "epoch": 1.8958274835738753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046482908095126803,
      "loss": 5.0508,
      "mean_token_accuracy": 0.22107810527086258,
      "num_tokens": 37682250.0,
      "step": 16880
    },
    {
      "entropy": 5.484669542312622,
      "epoch": 1.896389060481833,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046480752453711,
      "loss": 5.1283,
      "mean_token_accuracy": 0.21625462025403977,
      "num_tokens": 37694052.0,
      "step": 16885
    },
    {
      "entropy": 5.5489319324493405,
      "epoch": 1.8969506373897906,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004647859620794068,
      "loss": 5.173,
      "mean_token_accuracy": 0.21634069830179214,
      "num_tokens": 37704818.0,
      "step": 16890
    },
    {
      "entropy": 5.605900812149048,
      "epoch": 1.897512214297748,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046476439357884497,
      "loss": 5.2306,
      "mean_token_accuracy": 0.20108183324337006,
      "num_tokens": 37717679.0,
      "step": 16895
    },
    {
      "entropy": 5.4793174266815186,
      "epoch": 1.8980737912057055,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046474281903611126,
      "loss": 5.0801,
      "mean_token_accuracy": 0.20846800059080123,
      "num_tokens": 37728711.0,
      "step": 16900
    },
    {
      "entropy": 5.512612533569336,
      "epoch": 1.8986353681136632,
      "grad_norm": 1.0,
      "learning_rate": 0.00046472123845189253,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21195953339338303,
      "num_tokens": 37740082.0,
      "step": 16905
    },
    {
      "entropy": 5.580585861206055,
      "epoch": 1.8991969450216208,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000464699651826876,
      "loss": 5.204,
      "mean_token_accuracy": 0.20177604258060455,
      "num_tokens": 37750715.0,
      "step": 16910
    },
    {
      "entropy": 5.51938829421997,
      "epoch": 1.8997585219295783,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046467805916174907,
      "loss": 5.1692,
      "mean_token_accuracy": 0.20542775988578796,
      "num_tokens": 37763230.0,
      "step": 16915
    },
    {
      "entropy": 5.488376808166504,
      "epoch": 1.9003200988375357,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046465646045719915,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20421547889709474,
      "num_tokens": 37773621.0,
      "step": 16920
    },
    {
      "entropy": 5.543055629730224,
      "epoch": 1.9008816757454934,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004646348557139139,
      "loss": 5.1444,
      "mean_token_accuracy": 0.20714546144008636,
      "num_tokens": 37785068.0,
      "step": 16925
    },
    {
      "entropy": 5.5287432193756105,
      "epoch": 1.901443252653451,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046461324493258133,
      "loss": 5.138,
      "mean_token_accuracy": 0.2144142910838127,
      "num_tokens": 37796925.0,
      "step": 16930
    },
    {
      "entropy": 5.471051931381226,
      "epoch": 1.9020048295614085,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046459162811388955,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20846734642982484,
      "num_tokens": 37808291.0,
      "step": 16935
    },
    {
      "entropy": 5.430331563949585,
      "epoch": 1.902566406469366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004645700052585267,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2222292333841324,
      "num_tokens": 37819337.0,
      "step": 16940
    },
    {
      "entropy": 5.538673162460327,
      "epoch": 1.9031279833773236,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004645483763671815,
      "loss": 5.1737,
      "mean_token_accuracy": 0.2089215874671936,
      "num_tokens": 37831832.0,
      "step": 16945
    },
    {
      "entropy": 5.53937611579895,
      "epoch": 1.903689560285281,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046452674144054236,
      "loss": 5.1123,
      "mean_token_accuracy": 0.20860581248998641,
      "num_tokens": 37842586.0,
      "step": 16950
    },
    {
      "entropy": 5.490165376663208,
      "epoch": 1.9042511371932385,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046450510047929826,
      "loss": 5.0319,
      "mean_token_accuracy": 0.22052785456180574,
      "num_tokens": 37852198.0,
      "step": 16955
    },
    {
      "entropy": 5.49484486579895,
      "epoch": 1.9048127141011961,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046448345348413833,
      "loss": 5.0825,
      "mean_token_accuracy": 0.21098519414663314,
      "num_tokens": 37862967.0,
      "step": 16960
    },
    {
      "entropy": 5.492881011962891,
      "epoch": 1.9053742910091538,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004644618004557517,
      "loss": 5.2105,
      "mean_token_accuracy": 0.19650280177593232,
      "num_tokens": 37874501.0,
      "step": 16965
    },
    {
      "entropy": 5.462063789367676,
      "epoch": 1.9059358679171112,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004644401413948279,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2200288638472557,
      "num_tokens": 37886404.0,
      "step": 16970
    },
    {
      "entropy": 5.571914911270142,
      "epoch": 1.9064974448250687,
      "grad_norm": 1.0,
      "learning_rate": 0.0004644184763020565,
      "loss": 5.1889,
      "mean_token_accuracy": 0.2027203232049942,
      "num_tokens": 37897344.0,
      "step": 16975
    },
    {
      "entropy": 5.448255729675293,
      "epoch": 1.9070590217330263,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004643968051781273,
      "loss": 5.0247,
      "mean_token_accuracy": 0.22057299613952636,
      "num_tokens": 37909226.0,
      "step": 16980
    },
    {
      "entropy": 5.535170078277588,
      "epoch": 1.907620598640984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004643751280237305,
      "loss": 5.1534,
      "mean_token_accuracy": 0.20810251981019973,
      "num_tokens": 37920578.0,
      "step": 16985
    },
    {
      "entropy": 5.549870586395263,
      "epoch": 1.9081821755489414,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004643534448395561,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20492344945669175,
      "num_tokens": 37931814.0,
      "step": 16990
    },
    {
      "entropy": 5.491148042678833,
      "epoch": 1.9087437524568989,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004643317556262946,
      "loss": 5.0791,
      "mean_token_accuracy": 0.2090654417872429,
      "num_tokens": 37942345.0,
      "step": 16995
    },
    {
      "entropy": 5.4702905178070065,
      "epoch": 1.9093053293648565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046431006038463653,
      "loss": 4.9955,
      "mean_token_accuracy": 0.21522418707609176,
      "num_tokens": 37953027.0,
      "step": 17000
    },
    {
      "entropy": 5.531443023681641,
      "epoch": 1.9098669062728142,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004642883591152727,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20836733728647233,
      "num_tokens": 37963290.0,
      "step": 17005
    },
    {
      "entropy": 5.513917016983032,
      "epoch": 1.9104284831807716,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004642666518188942,
      "loss": 5.2477,
      "mean_token_accuracy": 0.20172259211540222,
      "num_tokens": 37974424.0,
      "step": 17010
    },
    {
      "entropy": 5.614091968536377,
      "epoch": 1.910990060088729,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00046424493849619205,
      "loss": 5.2333,
      "mean_token_accuracy": 0.20718204081058503,
      "num_tokens": 37987993.0,
      "step": 17015
    },
    {
      "entropy": 5.542656326293946,
      "epoch": 1.9115516369966867,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046422321914785757,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21496725380420684,
      "num_tokens": 37999340.0,
      "step": 17020
    },
    {
      "entropy": 5.477643394470215,
      "epoch": 1.9121132139046444,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046420149377458255,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20814482122659683,
      "num_tokens": 38009737.0,
      "step": 17025
    },
    {
      "entropy": 5.489460563659668,
      "epoch": 1.9126747908126018,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046417976237705846,
      "loss": 5.0735,
      "mean_token_accuracy": 0.20965756326913834,
      "num_tokens": 38021283.0,
      "step": 17030
    },
    {
      "entropy": 5.635855054855346,
      "epoch": 1.9132363677205593,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046415802495597744,
      "loss": 5.2081,
      "mean_token_accuracy": 0.2019257128238678,
      "num_tokens": 38034044.0,
      "step": 17035
    },
    {
      "entropy": 5.49278507232666,
      "epoch": 1.913797944628517,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046413628151203154,
      "loss": 5.0624,
      "mean_token_accuracy": 0.21443053781986238,
      "num_tokens": 38044585.0,
      "step": 17040
    },
    {
      "entropy": 5.431963920593262,
      "epoch": 1.9143595215364744,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046411453204591303,
      "loss": 5.0313,
      "mean_token_accuracy": 0.21053629964590073,
      "num_tokens": 38055771.0,
      "step": 17045
    },
    {
      "entropy": 5.468029403686524,
      "epoch": 1.9149210984444318,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046409277655831455,
      "loss": 5.102,
      "mean_token_accuracy": 0.20692533850669861,
      "num_tokens": 38066344.0,
      "step": 17050
    },
    {
      "entropy": 5.509651517868042,
      "epoch": 1.9154826753523895,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004640710150499287,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2124193713068962,
      "num_tokens": 38077785.0,
      "step": 17055
    },
    {
      "entropy": 5.486874103546143,
      "epoch": 1.9160442522603471,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046404924752144834,
      "loss": 5.1488,
      "mean_token_accuracy": 0.20800398886203766,
      "num_tokens": 38089023.0,
      "step": 17060
    },
    {
      "entropy": 5.517941474914551,
      "epoch": 1.9166058291683046,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046402747397356666,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2164747402071953,
      "num_tokens": 38099428.0,
      "step": 17065
    },
    {
      "entropy": 5.440216398239135,
      "epoch": 1.917167406076262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004640056944069768,
      "loss": 5.0234,
      "mean_token_accuracy": 0.21029316484928132,
      "num_tokens": 38110757.0,
      "step": 17070
    },
    {
      "entropy": 5.529173088073731,
      "epoch": 1.9177289829842197,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004639839088223725,
      "loss": 5.1115,
      "mean_token_accuracy": 0.20887204855680466,
      "num_tokens": 38121442.0,
      "step": 17075
    },
    {
      "entropy": 5.526953315734863,
      "epoch": 1.9182905598921773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046396211722044716,
      "loss": 5.0814,
      "mean_token_accuracy": 0.2084110200405121,
      "num_tokens": 38132917.0,
      "step": 17080
    },
    {
      "entropy": 5.43894248008728,
      "epoch": 1.9188521368001348,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046394031960189473,
      "loss": 5.0299,
      "mean_token_accuracy": 0.2077005013823509,
      "num_tokens": 38143188.0,
      "step": 17085
    },
    {
      "entropy": 5.601272678375244,
      "epoch": 1.9194137137080922,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004639185159674093,
      "loss": 5.1661,
      "mean_token_accuracy": 0.2141815021634102,
      "num_tokens": 38154300.0,
      "step": 17090
    },
    {
      "entropy": 5.543551874160767,
      "epoch": 1.9199752906160499,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046389670631768494,
      "loss": 5.195,
      "mean_token_accuracy": 0.20772186368703843,
      "num_tokens": 38165697.0,
      "step": 17095
    },
    {
      "entropy": 5.505008029937744,
      "epoch": 1.9205368675240075,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004638748906534162,
      "loss": 5.0932,
      "mean_token_accuracy": 0.2104604259133339,
      "num_tokens": 38176601.0,
      "step": 17100
    },
    {
      "entropy": 5.552993869781494,
      "epoch": 1.921098444431965,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004638530689752977,
      "loss": 5.1965,
      "mean_token_accuracy": 0.2072519689798355,
      "num_tokens": 38188334.0,
      "step": 17105
    },
    {
      "entropy": 5.571506690979004,
      "epoch": 1.9216600213399224,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046383124128402433,
      "loss": 5.1233,
      "mean_token_accuracy": 0.21032475382089616,
      "num_tokens": 38199483.0,
      "step": 17110
    },
    {
      "entropy": 5.543706893920898,
      "epoch": 1.92222159824788,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046380940758029095,
      "loss": 5.169,
      "mean_token_accuracy": 0.20641636997461318,
      "num_tokens": 38211391.0,
      "step": 17115
    },
    {
      "entropy": 5.511992883682251,
      "epoch": 1.9227831751558377,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046378756786479274,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21290345638990402,
      "num_tokens": 38223415.0,
      "step": 17120
    },
    {
      "entropy": 5.499290513992309,
      "epoch": 1.9233447520637952,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004637657221382252,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21650653183460236,
      "num_tokens": 38235093.0,
      "step": 17125
    },
    {
      "entropy": 5.486071968078614,
      "epoch": 1.9239063289717526,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046374387040128384,
      "loss": 5.1675,
      "mean_token_accuracy": 0.2031059518456459,
      "num_tokens": 38247192.0,
      "step": 17130
    },
    {
      "entropy": 5.556863594055176,
      "epoch": 1.9244679058797103,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046372201265466444,
      "loss": 5.1282,
      "mean_token_accuracy": 0.2048423707485199,
      "num_tokens": 38257965.0,
      "step": 17135
    },
    {
      "entropy": 5.593494033813476,
      "epoch": 1.9250294827876677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000463700148899063,
      "loss": 5.2313,
      "mean_token_accuracy": 0.20051195621490478,
      "num_tokens": 38268529.0,
      "step": 17140
    },
    {
      "entropy": 5.49454894065857,
      "epoch": 1.9255910596956252,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004636782791351756,
      "loss": 5.099,
      "mean_token_accuracy": 0.20805274248123168,
      "num_tokens": 38279480.0,
      "step": 17145
    },
    {
      "entropy": 5.558289480209351,
      "epoch": 1.9261526366035828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004636564033636987,
      "loss": 5.1723,
      "mean_token_accuracy": 0.20879970192909242,
      "num_tokens": 38291670.0,
      "step": 17150
    },
    {
      "entropy": 5.4374833583831785,
      "epoch": 1.9267142135115405,
      "grad_norm": 1.0,
      "learning_rate": 0.00046363452158532857,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22452065944671631,
      "num_tokens": 38303073.0,
      "step": 17155
    },
    {
      "entropy": 5.55133285522461,
      "epoch": 1.927275790419498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004636126338007622,
      "loss": 5.1136,
      "mean_token_accuracy": 0.20449230670928956,
      "num_tokens": 38313915.0,
      "step": 17160
    },
    {
      "entropy": 5.488015031814575,
      "epoch": 1.9278373673274554,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004635907400106965,
      "loss": 5.0869,
      "mean_token_accuracy": 0.20664833337068558,
      "num_tokens": 38324845.0,
      "step": 17165
    },
    {
      "entropy": 5.496645975112915,
      "epoch": 1.928398944235413,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004635688402158285,
      "loss": 5.0799,
      "mean_token_accuracy": 0.21642663478851318,
      "num_tokens": 38335096.0,
      "step": 17170
    },
    {
      "entropy": 5.5423304557800295,
      "epoch": 1.9289605211433707,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046354693441685537,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20953764468431474,
      "num_tokens": 38345559.0,
      "step": 17175
    },
    {
      "entropy": 5.454735851287841,
      "epoch": 1.9295220980513281,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046352502261447475,
      "loss": 5.0179,
      "mean_token_accuracy": 0.22581692785024643,
      "num_tokens": 38356318.0,
      "step": 17180
    },
    {
      "entropy": 5.4991730690002445,
      "epoch": 1.9300836749592856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004635031048093843,
      "loss": 5.0756,
      "mean_token_accuracy": 0.21269574761390686,
      "num_tokens": 38367246.0,
      "step": 17185
    },
    {
      "entropy": 5.50422101020813,
      "epoch": 1.9306452518672432,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004634811810022819,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20993321239948273,
      "num_tokens": 38378767.0,
      "step": 17190
    },
    {
      "entropy": 5.511542892456054,
      "epoch": 1.931206828775201,
      "grad_norm": 1.0,
      "learning_rate": 0.00046345925119386556,
      "loss": 5.0493,
      "mean_token_accuracy": 0.21711223125457763,
      "num_tokens": 38390414.0,
      "step": 17195
    },
    {
      "entropy": 5.504837942123413,
      "epoch": 1.9317684056831583,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004634373153848336,
      "loss": 5.0552,
      "mean_token_accuracy": 0.21715765595436096,
      "num_tokens": 38401569.0,
      "step": 17200
    },
    {
      "entropy": 5.57008376121521,
      "epoch": 1.9323299825911158,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004634153735758845,
      "loss": 5.1973,
      "mean_token_accuracy": 0.20708859115839004,
      "num_tokens": 38413561.0,
      "step": 17205
    },
    {
      "entropy": 5.5646172046661375,
      "epoch": 1.9328915594990734,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004633934257677167,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2140567660331726,
      "num_tokens": 38423965.0,
      "step": 17210
    },
    {
      "entropy": 5.485130405426025,
      "epoch": 1.933453136407031,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004633714719610292,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21017589420080185,
      "num_tokens": 38434158.0,
      "step": 17215
    },
    {
      "entropy": 5.456086492538452,
      "epoch": 1.9340147133149885,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046334951215652097,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2119516596198082,
      "num_tokens": 38445449.0,
      "step": 17220
    },
    {
      "entropy": 5.549007987976074,
      "epoch": 1.934576290222946,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046332754635489126,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20987604409456254,
      "num_tokens": 38457601.0,
      "step": 17225
    },
    {
      "entropy": 5.514615774154663,
      "epoch": 1.9351378671309036,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004633055745568394,
      "loss": 5.0637,
      "mean_token_accuracy": 0.2161819726228714,
      "num_tokens": 38468631.0,
      "step": 17230
    },
    {
      "entropy": 5.4328948020935055,
      "epoch": 1.935699444038861,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046328359676306506,
      "loss": 5.1187,
      "mean_token_accuracy": 0.21032307893037797,
      "num_tokens": 38479567.0,
      "step": 17235
    },
    {
      "entropy": 5.459002304077148,
      "epoch": 1.9362610209468185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000463261612974268,
      "loss": 5.0743,
      "mean_token_accuracy": 0.21308864206075667,
      "num_tokens": 38489934.0,
      "step": 17240
    },
    {
      "entropy": 5.614971017837524,
      "epoch": 1.9368225978547762,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004632396231911481,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20101860761642457,
      "num_tokens": 38502214.0,
      "step": 17245
    },
    {
      "entropy": 5.559498596191406,
      "epoch": 1.9373841747627338,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004632176274144056,
      "loss": 5.0951,
      "mean_token_accuracy": 0.20539983808994294,
      "num_tokens": 38513915.0,
      "step": 17250
    },
    {
      "entropy": 5.499972009658814,
      "epoch": 1.9379457516706913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046319562564474083,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21868704557418822,
      "num_tokens": 38525039.0,
      "step": 17255
    },
    {
      "entropy": 5.4255469799041744,
      "epoch": 1.9385073285786487,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046317361788285436,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21705400943756104,
      "num_tokens": 38537264.0,
      "step": 17260
    },
    {
      "entropy": 5.50873908996582,
      "epoch": 1.9390689054866064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004631516041294469,
      "loss": 5.1256,
      "mean_token_accuracy": 0.21171692460775376,
      "num_tokens": 38547926.0,
      "step": 17265
    },
    {
      "entropy": 5.533066987991333,
      "epoch": 1.939630482394564,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004631295843852194,
      "loss": 5.1168,
      "mean_token_accuracy": 0.2153423473238945,
      "num_tokens": 38558745.0,
      "step": 17270
    },
    {
      "entropy": 5.48057599067688,
      "epoch": 1.9401920593025215,
      "grad_norm": 1.015625,
      "learning_rate": 0.000463107558650873,
      "loss": 5.0598,
      "mean_token_accuracy": 0.21696562319993973,
      "num_tokens": 38568190.0,
      "step": 17275
    },
    {
      "entropy": 5.564253044128418,
      "epoch": 1.940753636210479,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046308552692710895,
      "loss": 5.1462,
      "mean_token_accuracy": 0.20852722674608232,
      "num_tokens": 38578932.0,
      "step": 17280
    },
    {
      "entropy": 5.5262538433074955,
      "epoch": 1.9413152131184366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046306348921462876,
      "loss": 5.1237,
      "mean_token_accuracy": 0.20576460808515548,
      "num_tokens": 38590501.0,
      "step": 17285
    },
    {
      "entropy": 5.463390731811524,
      "epoch": 1.9418767900263942,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004630414455141342,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2190954715013504,
      "num_tokens": 38600498.0,
      "step": 17290
    },
    {
      "entropy": 5.461698818206787,
      "epoch": 1.9424383669343517,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000463019395826327,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2127041131258011,
      "num_tokens": 38611991.0,
      "step": 17295
    },
    {
      "entropy": 5.4473169326782225,
      "epoch": 1.9429999438423091,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046299734015190934,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21374461948871612,
      "num_tokens": 38622284.0,
      "step": 17300
    },
    {
      "entropy": 5.520754480361939,
      "epoch": 1.9435615207502668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046297527849158344,
      "loss": 5.1563,
      "mean_token_accuracy": 0.2070332422852516,
      "num_tokens": 38633777.0,
      "step": 17305
    },
    {
      "entropy": 5.554642200469971,
      "epoch": 1.9441230976582244,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004629532108460519,
      "loss": 5.1143,
      "mean_token_accuracy": 0.20368158221244811,
      "num_tokens": 38644360.0,
      "step": 17310
    },
    {
      "entropy": 5.466087102890015,
      "epoch": 1.9446846745661819,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046293113721601706,
      "loss": 5.1133,
      "mean_token_accuracy": 0.21449712216854094,
      "num_tokens": 38655242.0,
      "step": 17315
    },
    {
      "entropy": 5.512748861312867,
      "epoch": 1.9452462514741393,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046290905760218193,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2246527001261711,
      "num_tokens": 38667792.0,
      "step": 17320
    },
    {
      "entropy": 5.441223621368408,
      "epoch": 1.945807828382097,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046288697200524964,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2176685616374016,
      "num_tokens": 38677735.0,
      "step": 17325
    },
    {
      "entropy": 5.484303903579712,
      "epoch": 1.9463694052900544,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046286488042592325,
      "loss": 5.0774,
      "mean_token_accuracy": 0.21318703293800353,
      "num_tokens": 38689556.0,
      "step": 17330
    },
    {
      "entropy": 5.492072629928589,
      "epoch": 1.9469309821980119,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004628427828649062,
      "loss": 5.0996,
      "mean_token_accuracy": 0.21750720739364623,
      "num_tokens": 38700556.0,
      "step": 17335
    },
    {
      "entropy": 5.486243724822998,
      "epoch": 1.9474925591059695,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046282067932290204,
      "loss": 5.0731,
      "mean_token_accuracy": 0.20691203474998474,
      "num_tokens": 38711019.0,
      "step": 17340
    },
    {
      "entropy": 5.462958526611328,
      "epoch": 1.9480541360139272,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004627985698006146,
      "loss": 5.0865,
      "mean_token_accuracy": 0.21032506078481675,
      "num_tokens": 38722228.0,
      "step": 17345
    },
    {
      "entropy": 5.429541873931885,
      "epoch": 1.9486157129218846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000462776454298748,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21790332794189454,
      "num_tokens": 38732183.0,
      "step": 17350
    },
    {
      "entropy": 5.4275336265563965,
      "epoch": 1.949177289829842,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046275433281800613,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21502691954374314,
      "num_tokens": 38743659.0,
      "step": 17355
    },
    {
      "entropy": 5.512891578674316,
      "epoch": 1.9497388667377997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004627322053590935,
      "loss": 5.1499,
      "mean_token_accuracy": 0.2083219036459923,
      "num_tokens": 38755032.0,
      "step": 17360
    },
    {
      "entropy": 5.514606952667236,
      "epoch": 1.9503004436457574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004627100719227147,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21565451472997665,
      "num_tokens": 38766126.0,
      "step": 17365
    },
    {
      "entropy": 5.56097822189331,
      "epoch": 1.9508620205537148,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046268793250957437,
      "loss": 5.1852,
      "mean_token_accuracy": 0.209283684194088,
      "num_tokens": 38776921.0,
      "step": 17370
    },
    {
      "entropy": 5.459462976455688,
      "epoch": 1.9514235974616723,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046266578712037747,
      "loss": 5.067,
      "mean_token_accuracy": 0.21421228498220443,
      "num_tokens": 38788884.0,
      "step": 17375
    },
    {
      "entropy": 5.61196460723877,
      "epoch": 1.95198517436963,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046264363575582916,
      "loss": 5.2128,
      "mean_token_accuracy": 0.20367702692747117,
      "num_tokens": 38800505.0,
      "step": 17380
    },
    {
      "entropy": 5.499144554138184,
      "epoch": 1.9525467512775876,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004626214784166347,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20915404707193375,
      "num_tokens": 38811132.0,
      "step": 17385
    },
    {
      "entropy": 5.573312902450562,
      "epoch": 1.953108328185545,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004625993151034995,
      "loss": 5.1246,
      "mean_token_accuracy": 0.21404734998941422,
      "num_tokens": 38821349.0,
      "step": 17390
    },
    {
      "entropy": 5.511688709259033,
      "epoch": 1.9536699050935025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004625771458171294,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20949177891016008,
      "num_tokens": 38832970.0,
      "step": 17395
    },
    {
      "entropy": 5.458802509307861,
      "epoch": 1.9542314820014601,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046255497055823025,
      "loss": 5.1095,
      "mean_token_accuracy": 0.209658282995224,
      "num_tokens": 38844220.0,
      "step": 17400
    },
    {
      "entropy": 5.480219030380249,
      "epoch": 1.9547930589094178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004625327893275082,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20576697140932082,
      "num_tokens": 38854848.0,
      "step": 17405
    },
    {
      "entropy": 5.52643232345581,
      "epoch": 1.9553546358173752,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046251060212566927,
      "loss": 5.2009,
      "mean_token_accuracy": 0.20566636621952056,
      "num_tokens": 38864827.0,
      "step": 17410
    },
    {
      "entropy": 5.531147289276123,
      "epoch": 1.9559162127253327,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004624884089534201,
      "loss": 5.047,
      "mean_token_accuracy": 0.2071419432759285,
      "num_tokens": 38875914.0,
      "step": 17415
    },
    {
      "entropy": 5.415796232223511,
      "epoch": 1.9564777896332903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004624662098114672,
      "loss": 5.034,
      "mean_token_accuracy": 0.2120719686150551,
      "num_tokens": 38886565.0,
      "step": 17420
    },
    {
      "entropy": 5.53605375289917,
      "epoch": 1.9570393665412478,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046244400470051754,
      "loss": 5.2878,
      "mean_token_accuracy": 0.1968618482351303,
      "num_tokens": 38897319.0,
      "step": 17425
    },
    {
      "entropy": 5.488940906524658,
      "epoch": 1.9576009434492052,
      "grad_norm": 1.015625,
      "learning_rate": 0.000462421793621278,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21463557779788972,
      "num_tokens": 38908824.0,
      "step": 17430
    },
    {
      "entropy": 5.5308998107910154,
      "epoch": 1.9581625203571629,
      "grad_norm": 0.9375,
      "learning_rate": 0.000462399576574456,
      "loss": 5.1367,
      "mean_token_accuracy": 0.21122062504291533,
      "num_tokens": 38921800.0,
      "step": 17435
    },
    {
      "entropy": 5.498126220703125,
      "epoch": 1.9587240972651205,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004623773535607587,
      "loss": 5.0839,
      "mean_token_accuracy": 0.21036441177129744,
      "num_tokens": 38932284.0,
      "step": 17440
    },
    {
      "entropy": 5.494027376174927,
      "epoch": 1.959285674173078,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004623551245808938,
      "loss": 5.0903,
      "mean_token_accuracy": 0.21248753666877745,
      "num_tokens": 38943632.0,
      "step": 17445
    },
    {
      "entropy": 5.493687105178833,
      "epoch": 1.9598472510810354,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004623328896355691,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21455607414245606,
      "num_tokens": 38954701.0,
      "step": 17450
    },
    {
      "entropy": 5.5608988285064695,
      "epoch": 1.960408827988993,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046231064872549244,
      "loss": 5.1293,
      "mean_token_accuracy": 0.2087258979678154,
      "num_tokens": 38965625.0,
      "step": 17455
    },
    {
      "entropy": 5.513734006881714,
      "epoch": 1.9609704048969507,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046228840185137215,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21236360520124437,
      "num_tokens": 38976989.0,
      "step": 17460
    },
    {
      "entropy": 5.416563224792481,
      "epoch": 1.9615319818049082,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004622661490139165,
      "loss": 5.1203,
      "mean_token_accuracy": 0.21698066741228103,
      "num_tokens": 38987661.0,
      "step": 17465
    },
    {
      "entropy": 5.491742849349976,
      "epoch": 1.9620935587128656,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004622438902138341,
      "loss": 5.0934,
      "mean_token_accuracy": 0.2114655703306198,
      "num_tokens": 38998014.0,
      "step": 17470
    },
    {
      "entropy": 5.488398838043213,
      "epoch": 1.9626551356208233,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046222162545183357,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21660214066505432,
      "num_tokens": 39008843.0,
      "step": 17475
    },
    {
      "entropy": 5.524238061904907,
      "epoch": 1.963216712528781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046219935472862386,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20438172817230224,
      "num_tokens": 39020173.0,
      "step": 17480
    },
    {
      "entropy": 5.5069468975067135,
      "epoch": 1.9637782894367384,
      "grad_norm": 1.0,
      "learning_rate": 0.00046217707804491404,
      "loss": 5.1228,
      "mean_token_accuracy": 0.2057792380452156,
      "num_tokens": 39033087.0,
      "step": 17485
    },
    {
      "entropy": 5.581241750717163,
      "epoch": 1.9643398663446958,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004621547954014135,
      "loss": 5.1038,
      "mean_token_accuracy": 0.21628159135580063,
      "num_tokens": 39045252.0,
      "step": 17490
    },
    {
      "entropy": 5.54075984954834,
      "epoch": 1.9649014432526535,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004621325067988317,
      "loss": 5.2374,
      "mean_token_accuracy": 0.20619835704565048,
      "num_tokens": 39056755.0,
      "step": 17495
    },
    {
      "entropy": 5.461740446090698,
      "epoch": 1.9654630201606111,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046211021223787827,
      "loss": 5.0201,
      "mean_token_accuracy": 0.21625028252601625,
      "num_tokens": 39066993.0,
      "step": 17500
    },
    {
      "entropy": 5.463070344924927,
      "epoch": 1.9660245970685686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046208791171926315,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21087120473384857,
      "num_tokens": 39078671.0,
      "step": 17505
    },
    {
      "entropy": 5.555288648605346,
      "epoch": 1.966586173976526,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046206560524369636,
      "loss": 5.09,
      "mean_token_accuracy": 0.2068958342075348,
      "num_tokens": 39089563.0,
      "step": 17510
    },
    {
      "entropy": 5.609184217453003,
      "epoch": 1.9671477508844837,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046204329281188816,
      "loss": 5.147,
      "mean_token_accuracy": 0.20618539601564406,
      "num_tokens": 39099837.0,
      "step": 17515
    },
    {
      "entropy": 5.551881122589111,
      "epoch": 1.9677093277924411,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004620209744245489,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20492150634527206,
      "num_tokens": 39112061.0,
      "step": 17520
    },
    {
      "entropy": 5.486149740219116,
      "epoch": 1.9682709047003986,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046199865008238937,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21601057946681976,
      "num_tokens": 39122777.0,
      "step": 17525
    },
    {
      "entropy": 5.433051681518554,
      "epoch": 1.9688324816083562,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004619763197861203,
      "loss": 5.0473,
      "mean_token_accuracy": 0.21456861793994902,
      "num_tokens": 39134827.0,
      "step": 17530
    },
    {
      "entropy": 5.502239608764649,
      "epoch": 1.9693940585163139,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046195398353645256,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20950087457895278,
      "num_tokens": 39144352.0,
      "step": 17535
    },
    {
      "entropy": 5.447834491729736,
      "epoch": 1.9699556354242713,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004619316413340976,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21413307189941405,
      "num_tokens": 39155610.0,
      "step": 17540
    },
    {
      "entropy": 5.435008716583252,
      "epoch": 1.9705172123322288,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046190929317976664,
      "loss": 5.0797,
      "mean_token_accuracy": 0.21072807908058167,
      "num_tokens": 39167058.0,
      "step": 17545
    },
    {
      "entropy": 5.5347373485565186,
      "epoch": 1.9710787892401864,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004618869390741713,
      "loss": 5.1895,
      "mean_token_accuracy": 0.20344768166542054,
      "num_tokens": 39178797.0,
      "step": 17550
    },
    {
      "entropy": 5.521516418457031,
      "epoch": 1.971640366148144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004618645790180233,
      "loss": 5.0327,
      "mean_token_accuracy": 0.2178642213344574,
      "num_tokens": 39188980.0,
      "step": 17555
    },
    {
      "entropy": 5.476399850845337,
      "epoch": 1.9722019430561015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004618422130120347,
      "loss": 5.0385,
      "mean_token_accuracy": 0.20950965881347655,
      "num_tokens": 39200058.0,
      "step": 17560
    },
    {
      "entropy": 5.456131839752198,
      "epoch": 1.972763519964059,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004618198410569175,
      "loss": 5.1211,
      "mean_token_accuracy": 0.21067933589220048,
      "num_tokens": 39211369.0,
      "step": 17565
    },
    {
      "entropy": 5.506629180908203,
      "epoch": 1.9733250968720166,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046179746315338414,
      "loss": 5.1673,
      "mean_token_accuracy": 0.20458319783210754,
      "num_tokens": 39222438.0,
      "step": 17570
    },
    {
      "entropy": 5.540224647521972,
      "epoch": 1.9738866737799743,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046177507930214714,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20280339270830156,
      "num_tokens": 39233047.0,
      "step": 17575
    },
    {
      "entropy": 5.501207304000855,
      "epoch": 1.9744482506879317,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004617526895039191,
      "loss": 5.0527,
      "mean_token_accuracy": 0.21280987709760665,
      "num_tokens": 39244547.0,
      "step": 17580
    },
    {
      "entropy": 5.5037909030914305,
      "epoch": 1.9750098275958892,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046173029375941305,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21062546819448472,
      "num_tokens": 39255549.0,
      "step": 17585
    },
    {
      "entropy": 5.4590222358703615,
      "epoch": 1.9755714045038468,
      "grad_norm": 1.140625,
      "learning_rate": 0.000461707892069342,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21513874977827072,
      "num_tokens": 39266527.0,
      "step": 17590
    },
    {
      "entropy": 5.474949550628662,
      "epoch": 1.9761329814118045,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046168548443441926,
      "loss": 5.1033,
      "mean_token_accuracy": 0.2163463920354843,
      "num_tokens": 39278013.0,
      "step": 17595
    },
    {
      "entropy": 5.500087070465088,
      "epoch": 1.976694558319762,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004616630708553583,
      "loss": 5.0801,
      "mean_token_accuracy": 0.2105663910508156,
      "num_tokens": 39289349.0,
      "step": 17600
    },
    {
      "entropy": 5.458623790740967,
      "epoch": 1.9772561352277194,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046164065133287276,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21383445262908934,
      "num_tokens": 39301213.0,
      "step": 17605
    },
    {
      "entropy": 5.414368629455566,
      "epoch": 1.977817712135677,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004616182258676765,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2220798745751381,
      "num_tokens": 39310960.0,
      "step": 17610
    },
    {
      "entropy": 5.5328442573547365,
      "epoch": 1.9783792890436345,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004615957944604835,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20953653007745743,
      "num_tokens": 39322603.0,
      "step": 17615
    },
    {
      "entropy": 5.52403883934021,
      "epoch": 1.978940865951592,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004615733571120081,
      "loss": 5.0922,
      "mean_token_accuracy": 0.21249038875102996,
      "num_tokens": 39333811.0,
      "step": 17620
    },
    {
      "entropy": 5.5889006614685055,
      "epoch": 1.9795024428595496,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004615509138229646,
      "loss": 5.1141,
      "mean_token_accuracy": 0.2104892373085022,
      "num_tokens": 39345218.0,
      "step": 17625
    },
    {
      "entropy": 5.44272518157959,
      "epoch": 1.9800640197675072,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046152846459406765,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21633495688438414,
      "num_tokens": 39356777.0,
      "step": 17630
    },
    {
      "entropy": 5.400483083724976,
      "epoch": 1.9806255966754647,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004615060094260321,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20853884369134904,
      "num_tokens": 39368288.0,
      "step": 17635
    },
    {
      "entropy": 5.54973258972168,
      "epoch": 1.9811871735834221,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046148354831957285,
      "loss": 5.0824,
      "mean_token_accuracy": 0.21646516919136047,
      "num_tokens": 39379398.0,
      "step": 17640
    },
    {
      "entropy": 5.458340120315552,
      "epoch": 1.9817487504913798,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046146108127540506,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21528061777353286,
      "num_tokens": 39391129.0,
      "step": 17645
    },
    {
      "entropy": 5.505998659133911,
      "epoch": 1.9823103273993374,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004614386082942442,
      "loss": 5.1455,
      "mean_token_accuracy": 0.20194572508335112,
      "num_tokens": 39402442.0,
      "step": 17650
    },
    {
      "entropy": 5.587158727645874,
      "epoch": 1.9828719043072949,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004614161293768057,
      "loss": 5.1018,
      "mean_token_accuracy": 0.2137361690402031,
      "num_tokens": 39413044.0,
      "step": 17655
    },
    {
      "entropy": 5.512194681167602,
      "epoch": 1.9834334812152523,
      "grad_norm": 1.0,
      "learning_rate": 0.00046139364452380536,
      "loss": 5.0689,
      "mean_token_accuracy": 0.21499471813440324,
      "num_tokens": 39423739.0,
      "step": 17660
    },
    {
      "entropy": 5.518840885162353,
      "epoch": 1.98399505812321,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000461371153735959,
      "loss": 5.2039,
      "mean_token_accuracy": 0.20920741260051728,
      "num_tokens": 39436199.0,
      "step": 17665
    },
    {
      "entropy": 5.546721792221069,
      "epoch": 1.9845566350311676,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00046134865701398287,
      "loss": 5.0958,
      "mean_token_accuracy": 0.2184235170483589,
      "num_tokens": 39449302.0,
      "step": 17670
    },
    {
      "entropy": 5.5443761348724365,
      "epoch": 1.985118211939125,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046132615435859326,
      "loss": 5.127,
      "mean_token_accuracy": 0.20702789574861527,
      "num_tokens": 39460723.0,
      "step": 17675
    },
    {
      "entropy": 5.4941198348999025,
      "epoch": 1.9856797888470825,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046130364577050655,
      "loss": 5.0571,
      "mean_token_accuracy": 0.2145622879266739,
      "num_tokens": 39471251.0,
      "step": 17680
    },
    {
      "entropy": 5.514349174499512,
      "epoch": 1.9862413657550402,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046128113125043957,
      "loss": 5.0563,
      "mean_token_accuracy": 0.21840719729661942,
      "num_tokens": 39482613.0,
      "step": 17685
    },
    {
      "entropy": 5.493435478210449,
      "epoch": 1.9868029426629978,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046125861079910913,
      "loss": 5.087,
      "mean_token_accuracy": 0.2208486169576645,
      "num_tokens": 39494626.0,
      "step": 17690
    },
    {
      "entropy": 5.429390716552734,
      "epoch": 1.9873645195709553,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004612360844172322,
      "loss": 5.02,
      "mean_token_accuracy": 0.2112290158867836,
      "num_tokens": 39505302.0,
      "step": 17695
    },
    {
      "entropy": 5.4902407169342045,
      "epoch": 1.9879260964789127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046121355210552626,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2135127902030945,
      "num_tokens": 39515854.0,
      "step": 17700
    },
    {
      "entropy": 5.542556095123291,
      "epoch": 1.9884876733868704,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004611910138647085,
      "loss": 5.1687,
      "mean_token_accuracy": 0.20248592495918274,
      "num_tokens": 39527504.0,
      "step": 17705
    },
    {
      "entropy": 5.535033655166626,
      "epoch": 1.989049250294828,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046116846969549666,
      "loss": 5.058,
      "mean_token_accuracy": 0.21445262730121611,
      "num_tokens": 39538885.0,
      "step": 17710
    },
    {
      "entropy": 5.46994571685791,
      "epoch": 1.9896108272027853,
      "grad_norm": 1.0,
      "learning_rate": 0.0004611459195986085,
      "loss": 5.0519,
      "mean_token_accuracy": 0.21494306921958922,
      "num_tokens": 39549401.0,
      "step": 17715
    },
    {
      "entropy": 5.48551573753357,
      "epoch": 1.990172404110743,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046112336357476217,
      "loss": 5.1302,
      "mean_token_accuracy": 0.2062358573079109,
      "num_tokens": 39560814.0,
      "step": 17720
    },
    {
      "entropy": 5.503226613998413,
      "epoch": 1.9907339810187006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004611008016246757,
      "loss": 5.0823,
      "mean_token_accuracy": 0.20804405957460403,
      "num_tokens": 39571490.0,
      "step": 17725
    },
    {
      "entropy": 5.4573742866516115,
      "epoch": 1.991295557926658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046107823374906754,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21922310888767244,
      "num_tokens": 39582815.0,
      "step": 17730
    },
    {
      "entropy": 5.509768152236939,
      "epoch": 1.9918571348346155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046105565994865614,
      "loss": 5.0887,
      "mean_token_accuracy": 0.21390316486358643,
      "num_tokens": 39593712.0,
      "step": 17735
    },
    {
      "entropy": 5.499211645126342,
      "epoch": 1.9924187117425731,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004610330802241605,
      "loss": 5.0776,
      "mean_token_accuracy": 0.21099428534507753,
      "num_tokens": 39604504.0,
      "step": 17740
    },
    {
      "entropy": 5.499092245101929,
      "epoch": 1.9929802886505308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004610104945762994,
      "loss": 5.0987,
      "mean_token_accuracy": 0.21465860158205033,
      "num_tokens": 39614971.0,
      "step": 17745
    },
    {
      "entropy": 5.5031129837036135,
      "epoch": 1.9935418655584882,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000460987903005792,
      "loss": 5.0477,
      "mean_token_accuracy": 0.21754294335842134,
      "num_tokens": 39626647.0,
      "step": 17750
    },
    {
      "entropy": 5.518901109695435,
      "epoch": 1.9941034424664457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004609653055133576,
      "loss": 5.1586,
      "mean_token_accuracy": 0.20872440040111542,
      "num_tokens": 39636432.0,
      "step": 17755
    },
    {
      "entropy": 5.5340526580810545,
      "epoch": 1.9946650193744033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046094270209971576,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2147448793053627,
      "num_tokens": 39647055.0,
      "step": 17760
    },
    {
      "entropy": 5.534825611114502,
      "epoch": 1.995226596282361,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004609200927655862,
      "loss": 5.079,
      "mean_token_accuracy": 0.21150288581848145,
      "num_tokens": 39657891.0,
      "step": 17765
    },
    {
      "entropy": 5.4525188446044925,
      "epoch": 1.9957881731903184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004608974775116888,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22385076135396959,
      "num_tokens": 39668023.0,
      "step": 17770
    },
    {
      "entropy": 5.463602590560913,
      "epoch": 1.9963497500982759,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004608748563387436,
      "loss": 5.0909,
      "mean_token_accuracy": 0.2162756472826004,
      "num_tokens": 39678398.0,
      "step": 17775
    },
    {
      "entropy": 5.456797409057617,
      "epoch": 1.9969113270062335,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004608522292474708,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20805446803569794,
      "num_tokens": 39690619.0,
      "step": 17780
    },
    {
      "entropy": 5.396885967254638,
      "epoch": 1.9974729039141912,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004608295962385911,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21658187210559846,
      "num_tokens": 39701690.0,
      "step": 17785
    },
    {
      "entropy": 5.471135950088501,
      "epoch": 1.9980344808221486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046080695731282484,
      "loss": 5.1418,
      "mean_token_accuracy": 0.20364563763141633,
      "num_tokens": 39713878.0,
      "step": 17790
    },
    {
      "entropy": 5.55912766456604,
      "epoch": 1.998596057730106,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000460784312470893,
      "loss": 5.0874,
      "mean_token_accuracy": 0.21329856514930726,
      "num_tokens": 39725164.0,
      "step": 17795
    },
    {
      "entropy": 5.487591648101807,
      "epoch": 1.9991576346380637,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046076166171351673,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2124040186405182,
      "num_tokens": 39735354.0,
      "step": 17800
    },
    {
      "entropy": 5.567241048812866,
      "epoch": 1.9997192115460214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000460739005041417,
      "loss": 5.1786,
      "mean_token_accuracy": 0.20215678811073304,
      "num_tokens": 39746618.0,
      "step": 17805
    },
    {
      "entropy": 5.5056837399800616,
      "epoch": 2.000224630763183,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004607163424553153,
      "loss": 5.1008,
      "mean_token_accuracy": 0.2133440590567059,
      "num_tokens": 39755697.0,
      "step": 17810
    },
    {
      "entropy": 5.430418491363525,
      "epoch": 2.0007862076711405,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046069367395593325,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2154828578233719,
      "num_tokens": 39766311.0,
      "step": 17815
    },
    {
      "entropy": 5.548388433456421,
      "epoch": 2.001347784579098,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004606709995439926,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21145301163196564,
      "num_tokens": 39777504.0,
      "step": 17820
    },
    {
      "entropy": 5.533293628692627,
      "epoch": 2.001909361487056,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046064831922021534,
      "loss": 4.9901,
      "mean_token_accuracy": 0.2146280139684677,
      "num_tokens": 39789190.0,
      "step": 17825
    },
    {
      "entropy": 5.466887331008911,
      "epoch": 2.002470938395013,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004606256329853235,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2125431478023529,
      "num_tokens": 39801534.0,
      "step": 17830
    },
    {
      "entropy": 5.519063568115234,
      "epoch": 2.0030325153029707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004606029408400396,
      "loss": 5.0209,
      "mean_token_accuracy": 0.21773672103881836,
      "num_tokens": 39812353.0,
      "step": 17835
    },
    {
      "entropy": 5.469810628890992,
      "epoch": 2.0035940922109283,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004605802427850861,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22153539210557938,
      "num_tokens": 39823313.0,
      "step": 17840
    },
    {
      "entropy": 5.43530421257019,
      "epoch": 2.004155669118886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046055753882118564,
      "loss": 4.9288,
      "mean_token_accuracy": 0.2247951865196228,
      "num_tokens": 39833966.0,
      "step": 17845
    },
    {
      "entropy": 5.499657249450683,
      "epoch": 2.004717246026843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004605348289490611,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21671060621738433,
      "num_tokens": 39843963.0,
      "step": 17850
    },
    {
      "entropy": 5.506442499160767,
      "epoch": 2.005278822934801,
      "grad_norm": 1.0,
      "learning_rate": 0.0004605121131694358,
      "loss": 5.0194,
      "mean_token_accuracy": 0.21268052905797957,
      "num_tokens": 39855237.0,
      "step": 17855
    },
    {
      "entropy": 5.545067882537841,
      "epoch": 2.0058403998427585,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004604893914830328,
      "loss": 5.1389,
      "mean_token_accuracy": 0.21175451576709747,
      "num_tokens": 39867270.0,
      "step": 17860
    },
    {
      "entropy": 5.583662652969361,
      "epoch": 2.006401976750716,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046046666389057563,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21966953277587892,
      "num_tokens": 39878103.0,
      "step": 17865
    },
    {
      "entropy": 5.468788051605225,
      "epoch": 2.0069635536586734,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000460443930392788,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2135985389351845,
      "num_tokens": 39888945.0,
      "step": 17870
    },
    {
      "entropy": 5.534671211242676,
      "epoch": 2.007525130566631,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046042119099039367,
      "loss": 5.1748,
      "mean_token_accuracy": 0.2087427407503128,
      "num_tokens": 39899848.0,
      "step": 17875
    },
    {
      "entropy": 5.524389171600342,
      "epoch": 2.0080867074745887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046039844568411667,
      "loss": 5.0287,
      "mean_token_accuracy": 0.21321588903665542,
      "num_tokens": 39911824.0,
      "step": 17880
    },
    {
      "entropy": 5.523957443237305,
      "epoch": 2.0086482843825464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004603756944746813,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21783199310302734,
      "num_tokens": 39923257.0,
      "step": 17885
    },
    {
      "entropy": 5.4722949981689455,
      "epoch": 2.0092098612905036,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046035293736281185,
      "loss": 4.9452,
      "mean_token_accuracy": 0.22077374905347824,
      "num_tokens": 39935312.0,
      "step": 17890
    },
    {
      "entropy": 5.4918975830078125,
      "epoch": 2.0097714381984613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004603301743492331,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21507604867219926,
      "num_tokens": 39946468.0,
      "step": 17895
    },
    {
      "entropy": 5.362382125854492,
      "epoch": 2.010333015106419,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004603074054346696,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22720238268375398,
      "num_tokens": 39957517.0,
      "step": 17900
    },
    {
      "entropy": 5.457093572616577,
      "epoch": 2.0108945920143766,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004602846306198465,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21380143463611603,
      "num_tokens": 39968385.0,
      "step": 17905
    },
    {
      "entropy": 5.458067083358765,
      "epoch": 2.011456168922334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004602618499054889,
      "loss": 4.932,
      "mean_token_accuracy": 0.2189737692475319,
      "num_tokens": 39979393.0,
      "step": 17910
    },
    {
      "entropy": 5.5300391674041744,
      "epoch": 2.0120177458302915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004602390632923221,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21148758232593537,
      "num_tokens": 39989709.0,
      "step": 17915
    },
    {
      "entropy": 5.511027097702026,
      "epoch": 2.012579322738249,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046021627078107177,
      "loss": 4.9337,
      "mean_token_accuracy": 0.21188617795705794,
      "num_tokens": 40000472.0,
      "step": 17920
    },
    {
      "entropy": 5.355703210830688,
      "epoch": 2.0131408996462063,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046019347237246343,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22474419474601745,
      "num_tokens": 40011062.0,
      "step": 17925
    },
    {
      "entropy": 5.552822828292847,
      "epoch": 2.013702476554164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004601706680672232,
      "loss": 5.1041,
      "mean_token_accuracy": 0.20089901089668274,
      "num_tokens": 40022430.0,
      "step": 17930
    },
    {
      "entropy": 5.460572147369385,
      "epoch": 2.0142640534621217,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004601478578660771,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2144003242254257,
      "num_tokens": 40033077.0,
      "step": 17935
    },
    {
      "entropy": 5.531789493560791,
      "epoch": 2.0148256303700793,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004601250417697513,
      "loss": 4.999,
      "mean_token_accuracy": 0.2227654606103897,
      "num_tokens": 40044258.0,
      "step": 17940
    },
    {
      "entropy": 5.549270439147949,
      "epoch": 2.0153872072780366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004601022197789724,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21001608222723006,
      "num_tokens": 40055876.0,
      "step": 17945
    },
    {
      "entropy": 5.449124050140381,
      "epoch": 2.015948784185994,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004600793918944671,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2259718045592308,
      "num_tokens": 40066691.0,
      "step": 17950
    },
    {
      "entropy": 5.463275718688965,
      "epoch": 2.016510361093952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004600565581169621,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21821629554033278,
      "num_tokens": 40077335.0,
      "step": 17955
    },
    {
      "entropy": 5.472330617904663,
      "epoch": 2.0170719380019095,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046003371844718455,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20822831988334656,
      "num_tokens": 40089053.0,
      "step": 17960
    },
    {
      "entropy": 5.4755579948425295,
      "epoch": 2.0176335149098668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046001087288586164,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22442007064819336,
      "num_tokens": 40099371.0,
      "step": 17965
    },
    {
      "entropy": 5.504808044433593,
      "epoch": 2.0181950918178244,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004599880214337208,
      "loss": 5.0635,
      "mean_token_accuracy": 0.21030860990285874,
      "num_tokens": 40110274.0,
      "step": 17970
    },
    {
      "entropy": 5.529354095458984,
      "epoch": 2.018756668725782,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004599651640914896,
      "loss": 5.073,
      "mean_token_accuracy": 0.21026033014059067,
      "num_tokens": 40120364.0,
      "step": 17975
    },
    {
      "entropy": 5.482689714431762,
      "epoch": 2.0193182456337397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004599423008598957,
      "loss": 5.0306,
      "mean_token_accuracy": 0.21393043100833892,
      "num_tokens": 40132400.0,
      "step": 17980
    },
    {
      "entropy": 5.506116580963135,
      "epoch": 2.019879822541697,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045991943173966744,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21296658217906952,
      "num_tokens": 40142820.0,
      "step": 17985
    },
    {
      "entropy": 5.52550196647644,
      "epoch": 2.0204413994496546,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004598965567315326,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21463638991117479,
      "num_tokens": 40153620.0,
      "step": 17990
    },
    {
      "entropy": 5.514501762390137,
      "epoch": 2.0210029763576123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004598736758362198,
      "loss": 4.9981,
      "mean_token_accuracy": 0.21377922594547272,
      "num_tokens": 40164250.0,
      "step": 17995
    },
    {
      "entropy": 5.471903991699219,
      "epoch": 2.02156455326557,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045985078905445734,
      "loss": 4.9292,
      "mean_token_accuracy": 0.2171599328517914,
      "num_tokens": 40175809.0,
      "step": 18000
    },
    {
      "epoch": 2.02156455326557,
      "eval_entropy": 5.288749482664276,
      "eval_loss": 5.150600433349609,
      "eval_mean_token_accuracy": 0.2166442650487766,
      "eval_num_tokens": 40175809.0,
      "eval_runtime": 18.4969,
      "eval_samples_per_second": 1676.663,
      "eval_steps_per_second": 209.603,
      "step": 18000
    },
    {
      "entropy": 5.379497289657593,
      "epoch": 2.022126130173527,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045982789638697414,
      "loss": 4.891,
      "mean_token_accuracy": 0.2242031827569008,
      "num_tokens": 40187855.0,
      "step": 18005
    },
    {
      "entropy": 5.4945213317871096,
      "epoch": 2.022687707081485,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045980499783449897,
      "loss": 4.9695,
      "mean_token_accuracy": 0.22087526321411133,
      "num_tokens": 40198802.0,
      "step": 18010
    },
    {
      "entropy": 5.603531408309936,
      "epoch": 2.0232492839894425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045978209339776104,
      "loss": 5.1675,
      "mean_token_accuracy": 0.20049773007631302,
      "num_tokens": 40210636.0,
      "step": 18015
    },
    {
      "entropy": 5.544476652145386,
      "epoch": 2.0238108608973997,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045975918307748965,
      "loss": 5.0369,
      "mean_token_accuracy": 0.21504127234220505,
      "num_tokens": 40221216.0,
      "step": 18020
    },
    {
      "entropy": 5.445092487335205,
      "epoch": 2.0243724378053574,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004597362668744141,
      "loss": 4.8992,
      "mean_token_accuracy": 0.22452698200941085,
      "num_tokens": 40231988.0,
      "step": 18025
    },
    {
      "entropy": 5.595971870422363,
      "epoch": 2.024934014713315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004597133447892642,
      "loss": 5.2083,
      "mean_token_accuracy": 0.20227197408676148,
      "num_tokens": 40243789.0,
      "step": 18030
    },
    {
      "entropy": 5.531242895126343,
      "epoch": 2.0254955916212727,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004596904168227698,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2195182576775551,
      "num_tokens": 40254771.0,
      "step": 18035
    },
    {
      "entropy": 5.502043533325195,
      "epoch": 2.02605716852923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045966748297566085,
      "loss": 4.9918,
      "mean_token_accuracy": 0.22369295358657837,
      "num_tokens": 40265444.0,
      "step": 18040
    },
    {
      "entropy": 5.499016380310058,
      "epoch": 2.0266187454371876,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045964454324866767,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2110731765627861,
      "num_tokens": 40277146.0,
      "step": 18045
    },
    {
      "entropy": 5.493988943099976,
      "epoch": 2.0271803223451452,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045962159764252055,
      "loss": 5.0899,
      "mean_token_accuracy": 0.2147793397307396,
      "num_tokens": 40286493.0,
      "step": 18050
    },
    {
      "entropy": 5.541953277587891,
      "epoch": 2.027741899253103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004595986461579502,
      "loss": 5.0027,
      "mean_token_accuracy": 0.21921555399894715,
      "num_tokens": 40297209.0,
      "step": 18055
    },
    {
      "entropy": 5.444691371917725,
      "epoch": 2.02830347616106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045957568879568736,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21676986068487167,
      "num_tokens": 40308666.0,
      "step": 18060
    },
    {
      "entropy": 5.5048667907714846,
      "epoch": 2.0288650530690178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045955272555646306,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20855429619550706,
      "num_tokens": 40321145.0,
      "step": 18065
    },
    {
      "entropy": 5.543835020065307,
      "epoch": 2.0294266299769754,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004595297564410083,
      "loss": 5.0162,
      "mean_token_accuracy": 0.218657548725605,
      "num_tokens": 40332418.0,
      "step": 18070
    },
    {
      "entropy": 5.442637205123901,
      "epoch": 2.029988206884933,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045950678145005456,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21859144866466523,
      "num_tokens": 40342838.0,
      "step": 18075
    },
    {
      "entropy": 5.4588264465332035,
      "epoch": 2.0305497837928903,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004594838005843333,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22099916338920594,
      "num_tokens": 40352940.0,
      "step": 18080
    },
    {
      "entropy": 5.570397138595581,
      "epoch": 2.031111360700848,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045946081384457633,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20862097889184952,
      "num_tokens": 40364273.0,
      "step": 18085
    },
    {
      "entropy": 5.5007225513458256,
      "epoch": 2.0316729376088056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045943782123151556,
      "loss": 4.9415,
      "mean_token_accuracy": 0.2262819305062294,
      "num_tokens": 40374617.0,
      "step": 18090
    },
    {
      "entropy": 5.4458253383636475,
      "epoch": 2.0322345145167633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004594148227458829,
      "loss": 5.0613,
      "mean_token_accuracy": 0.2112346425652504,
      "num_tokens": 40385398.0,
      "step": 18095
    },
    {
      "entropy": 5.603299331665039,
      "epoch": 2.0327960914247205,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004593918183884108,
      "loss": 5.1097,
      "mean_token_accuracy": 0.21055574119091033,
      "num_tokens": 40397030.0,
      "step": 18100
    },
    {
      "entropy": 5.569251441955567,
      "epoch": 2.033357668332678,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004593688081598318,
      "loss": 5.1002,
      "mean_token_accuracy": 0.20993017554283142,
      "num_tokens": 40408027.0,
      "step": 18105
    },
    {
      "entropy": 5.572751474380493,
      "epoch": 2.033919245240636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045934579206087837,
      "loss": 5.0328,
      "mean_token_accuracy": 0.21790675520896913,
      "num_tokens": 40418755.0,
      "step": 18110
    },
    {
      "entropy": 5.528239679336548,
      "epoch": 2.034480822148593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045932277009228336,
      "loss": 5.054,
      "mean_token_accuracy": 0.22020067870616913,
      "num_tokens": 40430694.0,
      "step": 18115
    },
    {
      "entropy": 5.5167765617370605,
      "epoch": 2.0350423990565507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045929974225477986,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21925848722457886,
      "num_tokens": 40442019.0,
      "step": 18120
    },
    {
      "entropy": 5.491471290588379,
      "epoch": 2.0356039759645084,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045927670854910117,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20773710310459137,
      "num_tokens": 40454120.0,
      "step": 18125
    },
    {
      "entropy": 5.481816244125366,
      "epoch": 2.036165552872466,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004592536689759806,
      "loss": 5.1573,
      "mean_token_accuracy": 0.21080696880817412,
      "num_tokens": 40467445.0,
      "step": 18130
    },
    {
      "entropy": 5.47070083618164,
      "epoch": 2.0367271297804233,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045923062353615167,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21739886999130248,
      "num_tokens": 40477512.0,
      "step": 18135
    },
    {
      "entropy": 5.430739879608154,
      "epoch": 2.037288706688381,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045920757223034833,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2223648324608803,
      "num_tokens": 40488209.0,
      "step": 18140
    },
    {
      "entropy": 5.544326257705689,
      "epoch": 2.0378502835963386,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004591845150593044,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21513805091381072,
      "num_tokens": 40499793.0,
      "step": 18145
    },
    {
      "entropy": 5.547037553787232,
      "epoch": 2.0384118605042962,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045916145202375407,
      "loss": 5.1225,
      "mean_token_accuracy": 0.20552191138267517,
      "num_tokens": 40510598.0,
      "step": 18150
    },
    {
      "entropy": 5.566928291320801,
      "epoch": 2.0389734374122535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004591383831244317,
      "loss": 5.0629,
      "mean_token_accuracy": 0.2161805659532547,
      "num_tokens": 40521289.0,
      "step": 18155
    },
    {
      "entropy": 5.6003069400787355,
      "epoch": 2.039535014320211,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045911530836207173,
      "loss": 5.0936,
      "mean_token_accuracy": 0.21076880991458893,
      "num_tokens": 40532048.0,
      "step": 18160
    },
    {
      "entropy": 5.516372156143189,
      "epoch": 2.0400965912281688,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045909222773740904,
      "loss": 5.068,
      "mean_token_accuracy": 0.21474754214286804,
      "num_tokens": 40543699.0,
      "step": 18165
    },
    {
      "entropy": 5.54240403175354,
      "epoch": 2.0406581681361264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045906914125117833,
      "loss": 5.0942,
      "mean_token_accuracy": 0.215559983253479,
      "num_tokens": 40554626.0,
      "step": 18170
    },
    {
      "entropy": 5.5363201141357425,
      "epoch": 2.0412197450440837,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045904604890411476,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2124529168009758,
      "num_tokens": 40567046.0,
      "step": 18175
    },
    {
      "entropy": 5.641685533523559,
      "epoch": 2.0417813219520413,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045902295069695373,
      "loss": 5.126,
      "mean_token_accuracy": 0.2017991915345192,
      "num_tokens": 40578472.0,
      "step": 18180
    },
    {
      "entropy": 5.526861095428467,
      "epoch": 2.042342898859999,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004589998466304305,
      "loss": 5.0408,
      "mean_token_accuracy": 0.21494200825691223,
      "num_tokens": 40589012.0,
      "step": 18185
    },
    {
      "entropy": 5.442477560043335,
      "epoch": 2.0429044757679566,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045897673670528086,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21925510764122008,
      "num_tokens": 40600807.0,
      "step": 18190
    },
    {
      "entropy": 5.6104436874389645,
      "epoch": 2.043466052675914,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004589536209222404,
      "loss": 5.125,
      "mean_token_accuracy": 0.21372979432344436,
      "num_tokens": 40612578.0,
      "step": 18195
    },
    {
      "entropy": 5.475525665283203,
      "epoch": 2.0440276295838715,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004589304992820454,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22448095977306365,
      "num_tokens": 40623504.0,
      "step": 18200
    },
    {
      "entropy": 5.537428760528565,
      "epoch": 2.044589206491829,
      "grad_norm": 1.03125,
      "learning_rate": 0.000458907371785432,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21251678317785264,
      "num_tokens": 40633676.0,
      "step": 18205
    },
    {
      "entropy": 5.5287086963653564,
      "epoch": 2.0451507833997864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004588842384331366,
      "loss": 5.1117,
      "mean_token_accuracy": 0.21183071434497833,
      "num_tokens": 40645991.0,
      "step": 18210
    },
    {
      "entropy": 5.4644937992095945,
      "epoch": 2.045712360307744,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045886109922589564,
      "loss": 4.9526,
      "mean_token_accuracy": 0.2195266768336296,
      "num_tokens": 40656617.0,
      "step": 18215
    },
    {
      "entropy": 5.516165924072266,
      "epoch": 2.0462739372157017,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045883795416444594,
      "loss": 5.0579,
      "mean_token_accuracy": 0.22006842344999314,
      "num_tokens": 40669708.0,
      "step": 18220
    },
    {
      "entropy": 5.496811628341675,
      "epoch": 2.0468355141236594,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00045881480324952455,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20650033801794052,
      "num_tokens": 40680248.0,
      "step": 18225
    },
    {
      "entropy": 5.593188047409058,
      "epoch": 2.0473970910316166,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045879164648186853,
      "loss": 5.1206,
      "mean_token_accuracy": 0.2094391867518425,
      "num_tokens": 40692704.0,
      "step": 18230
    },
    {
      "entropy": 5.535466051101684,
      "epoch": 2.0479586679395743,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045876848386221513,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2209407165646553,
      "num_tokens": 40703156.0,
      "step": 18235
    },
    {
      "entropy": 5.469901371002197,
      "epoch": 2.048520244847532,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045874531539130204,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2193508118391037,
      "num_tokens": 40713996.0,
      "step": 18240
    },
    {
      "entropy": 5.487784290313721,
      "epoch": 2.0490818217554896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045872214106986684,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22373228818178176,
      "num_tokens": 40724532.0,
      "step": 18245
    },
    {
      "entropy": 5.538501787185669,
      "epoch": 2.049643398663447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045869896089864733,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2152457982301712,
      "num_tokens": 40736045.0,
      "step": 18250
    },
    {
      "entropy": 5.38323130607605,
      "epoch": 2.0502049755714045,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004586757748783817,
      "loss": 4.919,
      "mean_token_accuracy": 0.21768129616975784,
      "num_tokens": 40747398.0,
      "step": 18255
    },
    {
      "entropy": 5.380796766281128,
      "epoch": 2.050766552479362,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045865258300980817,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2271013155579567,
      "num_tokens": 40758219.0,
      "step": 18260
    },
    {
      "entropy": 5.5226586818695065,
      "epoch": 2.05132812938732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045862938529366514,
      "loss": 4.9627,
      "mean_token_accuracy": 0.22624574601650238,
      "num_tokens": 40770002.0,
      "step": 18265
    },
    {
      "entropy": 5.449618577957153,
      "epoch": 2.051889706295277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004586061817306913,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22075289934873582,
      "num_tokens": 40780087.0,
      "step": 18270
    },
    {
      "entropy": 5.502636671066284,
      "epoch": 2.0524512832032347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045858297232162536,
      "loss": 5.0924,
      "mean_token_accuracy": 0.20624615848064423,
      "num_tokens": 40792225.0,
      "step": 18275
    },
    {
      "entropy": 5.534949684143067,
      "epoch": 2.0530128601111923,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045855975706720644,
      "loss": 5.1369,
      "mean_token_accuracy": 0.2085693970322609,
      "num_tokens": 40804437.0,
      "step": 18280
    },
    {
      "entropy": 5.444004774093628,
      "epoch": 2.05357443701915,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004585365359681736,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22914239913225173,
      "num_tokens": 40814625.0,
      "step": 18285
    },
    {
      "entropy": 5.4335548877716064,
      "epoch": 2.054136013927107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045851330902526633,
      "loss": 4.9497,
      "mean_token_accuracy": 0.21985217779874802,
      "num_tokens": 40825843.0,
      "step": 18290
    },
    {
      "entropy": 5.490375757217407,
      "epoch": 2.054697590835065,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045849007623922406,
      "loss": 5.1324,
      "mean_token_accuracy": 0.2020561143755913,
      "num_tokens": 40837146.0,
      "step": 18295
    },
    {
      "entropy": 5.614725780487061,
      "epoch": 2.0552591677430225,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004584668376107867,
      "loss": 5.191,
      "mean_token_accuracy": 0.20533615946769715,
      "num_tokens": 40849109.0,
      "step": 18300
    },
    {
      "entropy": 5.50355019569397,
      "epoch": 2.0558207446509797,
      "grad_norm": 1.015625,
      "learning_rate": 0.000458443593140694,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21141783148050308,
      "num_tokens": 40861168.0,
      "step": 18305
    },
    {
      "entropy": 5.544817495346069,
      "epoch": 2.0563823215589374,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045842034282968615,
      "loss": 5.0699,
      "mean_token_accuracy": 0.21235246807336808,
      "num_tokens": 40873403.0,
      "step": 18310
    },
    {
      "entropy": 5.47606840133667,
      "epoch": 2.056943898466895,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004583970866785035,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2140461027622223,
      "num_tokens": 40883434.0,
      "step": 18315
    },
    {
      "entropy": 5.48830018043518,
      "epoch": 2.0575054753748527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045837382468788643,
      "loss": 5.009,
      "mean_token_accuracy": 0.21490588039159775,
      "num_tokens": 40894742.0,
      "step": 18320
    },
    {
      "entropy": 5.4624017715454105,
      "epoch": 2.05806705228281,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004583505568585757,
      "loss": 4.9998,
      "mean_token_accuracy": 0.2196988970041275,
      "num_tokens": 40905233.0,
      "step": 18325
    },
    {
      "entropy": 5.491911602020264,
      "epoch": 2.0586286291907676,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045832728319131205,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22475702911615372,
      "num_tokens": 40915897.0,
      "step": 18330
    },
    {
      "entropy": 5.4700355052948,
      "epoch": 2.0591902060987253,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004583040036868367,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21097755581140518,
      "num_tokens": 40928178.0,
      "step": 18335
    },
    {
      "entropy": 5.457161951065063,
      "epoch": 2.059751783006683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004582807183458907,
      "loss": 5.0645,
      "mean_token_accuracy": 0.2080140382051468,
      "num_tokens": 40938988.0,
      "step": 18340
    },
    {
      "entropy": 5.441991567611694,
      "epoch": 2.06031335991464,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045825742716921553,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22158170491456985,
      "num_tokens": 40949671.0,
      "step": 18345
    },
    {
      "entropy": 5.418078422546387,
      "epoch": 2.060874936822598,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004582341301575529,
      "loss": 4.8986,
      "mean_token_accuracy": 0.21813996285200118,
      "num_tokens": 40960818.0,
      "step": 18350
    },
    {
      "entropy": 5.589268159866333,
      "epoch": 2.0614365137305555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045821082731164445,
      "loss": 5.1508,
      "mean_token_accuracy": 0.20092292726039887,
      "num_tokens": 40973866.0,
      "step": 18355
    },
    {
      "entropy": 5.490549516677857,
      "epoch": 2.061998090638513,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004581875186322322,
      "loss": 5.022,
      "mean_token_accuracy": 0.2121705397963524,
      "num_tokens": 40983355.0,
      "step": 18360
    },
    {
      "entropy": 5.424019241333008,
      "epoch": 2.0625596675464704,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045816420412005826,
      "loss": 4.906,
      "mean_token_accuracy": 0.22412308305501938,
      "num_tokens": 40994045.0,
      "step": 18365
    },
    {
      "entropy": 5.411766433715821,
      "epoch": 2.063121244454428,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045814088377586505,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22670055329799652,
      "num_tokens": 41003974.0,
      "step": 18370
    },
    {
      "entropy": 5.489456272125244,
      "epoch": 2.0636828213623857,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045811755760039513,
      "loss": 5.0824,
      "mean_token_accuracy": 0.21575163006782533,
      "num_tokens": 41015262.0,
      "step": 18375
    },
    {
      "entropy": 5.5584570407867435,
      "epoch": 2.0642443982703433,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045809422559439113,
      "loss": 5.1138,
      "mean_token_accuracy": 0.2107735976576805,
      "num_tokens": 41027103.0,
      "step": 18380
    },
    {
      "entropy": 5.529270076751709,
      "epoch": 2.0648059751783006,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045807088775859595,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21617121398448944,
      "num_tokens": 41037240.0,
      "step": 18385
    },
    {
      "entropy": 5.563993883132935,
      "epoch": 2.065367552086258,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004580475440937527,
      "loss": 5.0004,
      "mean_token_accuracy": 0.2192753165960312,
      "num_tokens": 41048258.0,
      "step": 18390
    },
    {
      "entropy": 5.557229328155517,
      "epoch": 2.065929128994216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045802419460060464,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2110954701900482,
      "num_tokens": 41058204.0,
      "step": 18395
    },
    {
      "entropy": 5.601803636550903,
      "epoch": 2.066490705902173,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004580008392798953,
      "loss": 5.0787,
      "mean_token_accuracy": 0.21253258734941483,
      "num_tokens": 41070019.0,
      "step": 18400
    },
    {
      "entropy": 5.562759447097778,
      "epoch": 2.0670522828101308,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045797747813236815,
      "loss": 5.0389,
      "mean_token_accuracy": 0.22048287689685822,
      "num_tokens": 41081503.0,
      "step": 18405
    },
    {
      "entropy": 5.519818305969238,
      "epoch": 2.0676138597180884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045795411115876725,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2068215250968933,
      "num_tokens": 41091959.0,
      "step": 18410
    },
    {
      "entropy": 5.536817836761474,
      "epoch": 2.068175436626046,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045793073835983644,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21535297185182573,
      "num_tokens": 41103193.0,
      "step": 18415
    },
    {
      "entropy": 5.56896276473999,
      "epoch": 2.0687370135340033,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045790735973631996,
      "loss": 5.108,
      "mean_token_accuracy": 0.2200862780213356,
      "num_tokens": 41115886.0,
      "step": 18420
    },
    {
      "entropy": 5.4808789730072025,
      "epoch": 2.069298590441961,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045788397528896224,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21946857273578643,
      "num_tokens": 41126868.0,
      "step": 18425
    },
    {
      "entropy": 5.557408809661865,
      "epoch": 2.0698601673499186,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045786058501850774,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2221153661608696,
      "num_tokens": 41136417.0,
      "step": 18430
    },
    {
      "entropy": 5.509507036209106,
      "epoch": 2.0704217442578763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045783718892570136,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20932406485080718,
      "num_tokens": 41149515.0,
      "step": 18435
    },
    {
      "entropy": 5.579898834228516,
      "epoch": 2.0709833211658335,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000457813787011288,
      "loss": 5.1039,
      "mean_token_accuracy": 0.21785481721162797,
      "num_tokens": 41160023.0,
      "step": 18440
    },
    {
      "entropy": 5.412167930603028,
      "epoch": 2.071544898073791,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004577903792760127,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22888001203536987,
      "num_tokens": 41170909.0,
      "step": 18445
    },
    {
      "entropy": 5.559626770019531,
      "epoch": 2.072106474981749,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004577669657206209,
      "loss": 5.1047,
      "mean_token_accuracy": 0.20876959711313248,
      "num_tokens": 41181960.0,
      "step": 18450
    },
    {
      "entropy": 5.5878949642181395,
      "epoch": 2.0726680518897065,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004577435463458579,
      "loss": 5.096,
      "mean_token_accuracy": 0.21168456673622132,
      "num_tokens": 41192292.0,
      "step": 18455
    },
    {
      "entropy": 5.5581199645996096,
      "epoch": 2.0732296287976637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004577201211524696,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2213801383972168,
      "num_tokens": 41203826.0,
      "step": 18460
    },
    {
      "entropy": 5.4602149486541744,
      "epoch": 2.0737912057056214,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004576966901412018,
      "loss": 5.0666,
      "mean_token_accuracy": 0.21620500236749648,
      "num_tokens": 41215095.0,
      "step": 18465
    },
    {
      "entropy": 5.5341572761535645,
      "epoch": 2.074352782613579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045767325331280044,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2149103358387947,
      "num_tokens": 41224953.0,
      "step": 18470
    },
    {
      "entropy": 5.561403608322143,
      "epoch": 2.0749143595215367,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004576498106680119,
      "loss": 5.0449,
      "mean_token_accuracy": 0.20596526563167572,
      "num_tokens": 41235479.0,
      "step": 18475
    },
    {
      "entropy": 5.527247953414917,
      "epoch": 2.075475936429494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004576263622075826,
      "loss": 5.0703,
      "mean_token_accuracy": 0.21642846316099168,
      "num_tokens": 41248520.0,
      "step": 18480
    },
    {
      "entropy": 5.576947259902954,
      "epoch": 2.0760375133374516,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000457602907932259,
      "loss": 5.1564,
      "mean_token_accuracy": 0.2077233985066414,
      "num_tokens": 41259112.0,
      "step": 18485
    },
    {
      "entropy": 5.502575874328613,
      "epoch": 2.0765990902454092,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045757944784278803,
      "loss": 4.9915,
      "mean_token_accuracy": 0.219262033700943,
      "num_tokens": 41269554.0,
      "step": 18490
    },
    {
      "entropy": 5.542146062850952,
      "epoch": 2.0771606671533664,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004575559819399166,
      "loss": 5.0239,
      "mean_token_accuracy": 0.21835334748029708,
      "num_tokens": 41281339.0,
      "step": 18495
    },
    {
      "entropy": 5.476851749420166,
      "epoch": 2.077722244061324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045753251022439195,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22250728011131288,
      "num_tokens": 41291629.0,
      "step": 18500
    },
    {
      "entropy": 5.560992813110351,
      "epoch": 2.0782838209692818,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004575090326969614,
      "loss": 5.088,
      "mean_token_accuracy": 0.21118887960910798,
      "num_tokens": 41302162.0,
      "step": 18505
    },
    {
      "entropy": 5.482925939559936,
      "epoch": 2.0788453978772394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004574855493583723,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2173047974705696,
      "num_tokens": 41313667.0,
      "step": 18510
    },
    {
      "entropy": 5.4807651996612545,
      "epoch": 2.0794069747851966,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045746206020937276,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21830037534236907,
      "num_tokens": 41325063.0,
      "step": 18515
    },
    {
      "entropy": 5.512169933319091,
      "epoch": 2.0799685516931543,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045743856525071034,
      "loss": 5.0709,
      "mean_token_accuracy": 0.21254366636276245,
      "num_tokens": 41335974.0,
      "step": 18520
    },
    {
      "entropy": 5.480447435379029,
      "epoch": 2.080530128601112,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045741506448313327,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21852853149175644,
      "num_tokens": 41347469.0,
      "step": 18525
    },
    {
      "entropy": 5.568706083297729,
      "epoch": 2.0810917055090696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045739155790738975,
      "loss": 5.0903,
      "mean_token_accuracy": 0.21995055079460143,
      "num_tokens": 41358418.0,
      "step": 18530
    },
    {
      "entropy": 5.539313411712646,
      "epoch": 2.081653282417027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045736804552422836,
      "loss": 5.0909,
      "mean_token_accuracy": 0.21096516102552415,
      "num_tokens": 41368859.0,
      "step": 18535
    },
    {
      "entropy": 5.5604674339294435,
      "epoch": 2.0822148593249845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004573445273343976,
      "loss": 5.1294,
      "mean_token_accuracy": 0.20696448981761933,
      "num_tokens": 41380929.0,
      "step": 18540
    },
    {
      "entropy": 5.523511362075806,
      "epoch": 2.082776436232942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004573210033386464,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22523413896560668,
      "num_tokens": 41392190.0,
      "step": 18545
    },
    {
      "entropy": 5.605782127380371,
      "epoch": 2.0833380131409,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004572974735377238,
      "loss": 5.0118,
      "mean_token_accuracy": 0.21911531239748,
      "num_tokens": 41402646.0,
      "step": 18550
    },
    {
      "entropy": 5.476578187942505,
      "epoch": 2.083899590048857,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004572739379323788,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2140674591064453,
      "num_tokens": 41415179.0,
      "step": 18555
    },
    {
      "entropy": 5.504357242584229,
      "epoch": 2.0844611669568147,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045725039652336105,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21215118169784547,
      "num_tokens": 41426695.0,
      "step": 18560
    },
    {
      "entropy": 5.499299144744873,
      "epoch": 2.0850227438647724,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004572268493114199,
      "loss": 5.0149,
      "mean_token_accuracy": 0.2138336941599846,
      "num_tokens": 41438078.0,
      "step": 18565
    },
    {
      "entropy": 5.540561866760254,
      "epoch": 2.08558432077273,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004572032962973052,
      "loss": 5.0716,
      "mean_token_accuracy": 0.2127721682190895,
      "num_tokens": 41448862.0,
      "step": 18570
    },
    {
      "entropy": 5.432217264175415,
      "epoch": 2.0861458976806873,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045717973748176695,
      "loss": 4.8698,
      "mean_token_accuracy": 0.22463872134685517,
      "num_tokens": 41458528.0,
      "step": 18575
    },
    {
      "entropy": 5.466809272766113,
      "epoch": 2.086707474588645,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045715617286555516,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2125450000166893,
      "num_tokens": 41469553.0,
      "step": 18580
    },
    {
      "entropy": 5.592465782165528,
      "epoch": 2.0872690514966026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045713260244942017,
      "loss": 5.1517,
      "mean_token_accuracy": 0.21182382106781006,
      "num_tokens": 41482327.0,
      "step": 18585
    },
    {
      "entropy": 5.468498706817627,
      "epoch": 2.08783062840456,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004571090262341125,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22608505487442015,
      "num_tokens": 41493126.0,
      "step": 18590
    },
    {
      "entropy": 5.504410457611084,
      "epoch": 2.0883922053125175,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045708544422038284,
      "loss": 5.0713,
      "mean_token_accuracy": 0.2091891959309578,
      "num_tokens": 41505597.0,
      "step": 18595
    },
    {
      "entropy": 5.584372425079346,
      "epoch": 2.088953782220475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004570618564089819,
      "loss": 5.13,
      "mean_token_accuracy": 0.20768361091613768,
      "num_tokens": 41517296.0,
      "step": 18600
    },
    {
      "entropy": 5.569519138336181,
      "epoch": 2.089515359128433,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045703826280066095,
      "loss": 5.015,
      "mean_token_accuracy": 0.2192860797047615,
      "num_tokens": 41529900.0,
      "step": 18605
    },
    {
      "entropy": 5.5330963134765625,
      "epoch": 2.09007693603639,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004570146633961711,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21736267805099488,
      "num_tokens": 41540641.0,
      "step": 18610
    },
    {
      "entropy": 5.434649801254272,
      "epoch": 2.0906385129443477,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045699105819626374,
      "loss": 5.0404,
      "mean_token_accuracy": 0.2048357054591179,
      "num_tokens": 41551335.0,
      "step": 18615
    },
    {
      "entropy": 5.504931211471558,
      "epoch": 2.0912000898523053,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004569674472016904,
      "loss": 5.0073,
      "mean_token_accuracy": 0.22068156599998473,
      "num_tokens": 41562265.0,
      "step": 18620
    },
    {
      "entropy": 5.5059041500091555,
      "epoch": 2.091761666760263,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004569438304132031,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21933801621198654,
      "num_tokens": 41574758.0,
      "step": 18625
    },
    {
      "entropy": 5.42665376663208,
      "epoch": 2.09232324366822,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045692020783155376,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2153540164232254,
      "num_tokens": 41585007.0,
      "step": 18630
    },
    {
      "entropy": 5.427863502502442,
      "epoch": 2.092884820576178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004568965794574943,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21884457767009735,
      "num_tokens": 41596026.0,
      "step": 18635
    },
    {
      "entropy": 5.566444683074951,
      "epoch": 2.0934463974841355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004568729452917772,
      "loss": 5.0789,
      "mean_token_accuracy": 0.21639298051595687,
      "num_tokens": 41608440.0,
      "step": 18640
    },
    {
      "entropy": 5.499663639068603,
      "epoch": 2.094007974392093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045684930533515506,
      "loss": 4.973,
      "mean_token_accuracy": 0.2177707701921463,
      "num_tokens": 41620467.0,
      "step": 18645
    },
    {
      "entropy": 5.485700035095215,
      "epoch": 2.0945695513000504,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045682565958838045,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21357071101665498,
      "num_tokens": 41630918.0,
      "step": 18650
    },
    {
      "entropy": 5.500881147384644,
      "epoch": 2.095131128208008,
      "grad_norm": 1.0,
      "learning_rate": 0.0004568020080522063,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21469495743513106,
      "num_tokens": 41641148.0,
      "step": 18655
    },
    {
      "entropy": 5.5581659317016605,
      "epoch": 2.0956927051159657,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004567783507273858,
      "loss": 5.0309,
      "mean_token_accuracy": 0.2199338674545288,
      "num_tokens": 41652148.0,
      "step": 18660
    },
    {
      "entropy": 5.506256818771362,
      "epoch": 2.0962542820239234,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045675468761467204,
      "loss": 4.9993,
      "mean_token_accuracy": 0.21786906868219375,
      "num_tokens": 41662275.0,
      "step": 18665
    },
    {
      "entropy": 5.566605854034424,
      "epoch": 2.0968158589318806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045673101871481853,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20855064541101456,
      "num_tokens": 41673141.0,
      "step": 18670
    },
    {
      "entropy": 5.631915950775147,
      "epoch": 2.0973774358398383,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004567073440285789,
      "loss": 5.1686,
      "mean_token_accuracy": 0.20746597796678543,
      "num_tokens": 41683473.0,
      "step": 18675
    },
    {
      "entropy": 5.426448345184326,
      "epoch": 2.097939012747796,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045668366355670697,
      "loss": 4.8098,
      "mean_token_accuracy": 0.22444780468940734,
      "num_tokens": 41693847.0,
      "step": 18680
    },
    {
      "entropy": 5.482524967193603,
      "epoch": 2.098500589655753,
      "grad_norm": 1.0,
      "learning_rate": 0.0004566599772999567,
      "loss": 5.0684,
      "mean_token_accuracy": 0.2175423324108124,
      "num_tokens": 41705810.0,
      "step": 18685
    },
    {
      "entropy": 5.495926713943481,
      "epoch": 2.099062166563711,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045663628525908237,
      "loss": 4.928,
      "mean_token_accuracy": 0.21603805273771287,
      "num_tokens": 41717160.0,
      "step": 18690
    },
    {
      "entropy": 5.541777849197388,
      "epoch": 2.0996237434716685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004566125874348382,
      "loss": 5.0975,
      "mean_token_accuracy": 0.21484368592500686,
      "num_tokens": 41728631.0,
      "step": 18695
    },
    {
      "entropy": 5.584958887100219,
      "epoch": 2.100185320379626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045658888382797886,
      "loss": 5.1469,
      "mean_token_accuracy": 0.2069434106349945,
      "num_tokens": 41738933.0,
      "step": 18700
    },
    {
      "entropy": 5.6201997756958,
      "epoch": 2.1007468972875833,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045656517443925897,
      "loss": 5.1265,
      "mean_token_accuracy": 0.2181014060974121,
      "num_tokens": 41750386.0,
      "step": 18705
    },
    {
      "entropy": 5.498168992996216,
      "epoch": 2.101308474195541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004565414592694336,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2136332720518112,
      "num_tokens": 41760711.0,
      "step": 18710
    },
    {
      "entropy": 5.49470329284668,
      "epoch": 2.1018700511034987,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004565177383192577,
      "loss": 5.0965,
      "mean_token_accuracy": 0.21033832430839539,
      "num_tokens": 41772147.0,
      "step": 18715
    },
    {
      "entropy": 5.536369180679321,
      "epoch": 2.1024316280114563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045649401158948664,
      "loss": 4.9635,
      "mean_token_accuracy": 0.2190065234899521,
      "num_tokens": 41783265.0,
      "step": 18720
    },
    {
      "entropy": 5.529592990875244,
      "epoch": 2.1029932049194136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004564702790808758,
      "loss": 4.9663,
      "mean_token_accuracy": 0.21945076286792756,
      "num_tokens": 41793977.0,
      "step": 18725
    },
    {
      "entropy": 5.523620891571045,
      "epoch": 2.103554781827371,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004564465407941809,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2241202637553215,
      "num_tokens": 41804790.0,
      "step": 18730
    },
    {
      "entropy": 5.4101275444030765,
      "epoch": 2.104116358735329,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045642279673015787,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21147693544626237,
      "num_tokens": 41815611.0,
      "step": 18735
    },
    {
      "entropy": 5.4383580684661865,
      "epoch": 2.1046779356432865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004563990468895625,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21274339258670807,
      "num_tokens": 41826935.0,
      "step": 18740
    },
    {
      "entropy": 5.450624418258667,
      "epoch": 2.1052395125512438,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004563752912731512,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2177711844444275,
      "num_tokens": 41838932.0,
      "step": 18745
    },
    {
      "entropy": 5.433754205703735,
      "epoch": 2.1058010894592014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004563515298816803,
      "loss": 5.0236,
      "mean_token_accuracy": 0.21329770535230635,
      "num_tokens": 41850771.0,
      "step": 18750
    },
    {
      "entropy": 5.459846258163452,
      "epoch": 2.106362666367159,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045632776271590624,
      "loss": 4.9575,
      "mean_token_accuracy": 0.2232426792383194,
      "num_tokens": 41861515.0,
      "step": 18755
    },
    {
      "entropy": 5.486744976043701,
      "epoch": 2.1069242432751167,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004563039897765859,
      "loss": 4.8655,
      "mean_token_accuracy": 0.23283354043960572,
      "num_tokens": 41872195.0,
      "step": 18760
    },
    {
      "entropy": 5.430644226074219,
      "epoch": 2.107485820183074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004562802110644764,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2259806588292122,
      "num_tokens": 41882496.0,
      "step": 18765
    },
    {
      "entropy": 5.56043004989624,
      "epoch": 2.1080473970910316,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045625642658033447,
      "loss": 5.1968,
      "mean_token_accuracy": 0.20316258668899537,
      "num_tokens": 41893078.0,
      "step": 18770
    },
    {
      "entropy": 5.576148271560669,
      "epoch": 2.1086089739989893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045623263632491766,
      "loss": 5.044,
      "mean_token_accuracy": 0.21328755468130112,
      "num_tokens": 41905062.0,
      "step": 18775
    },
    {
      "entropy": 5.465897083282471,
      "epoch": 2.1091705509069465,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045620884029898344,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21370731443166732,
      "num_tokens": 41916086.0,
      "step": 18780
    },
    {
      "entropy": 5.505476188659668,
      "epoch": 2.109732127814904,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045618503850328954,
      "loss": 5.0772,
      "mean_token_accuracy": 0.21464026421308519,
      "num_tokens": 41925466.0,
      "step": 18785
    },
    {
      "entropy": 5.571772527694702,
      "epoch": 2.110293704722862,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004561612309385936,
      "loss": 5.115,
      "mean_token_accuracy": 0.2072063997387886,
      "num_tokens": 41936567.0,
      "step": 18790
    },
    {
      "entropy": 5.492284107208252,
      "epoch": 2.1108552816308195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045613741760565385,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21496164053678513,
      "num_tokens": 41947021.0,
      "step": 18795
    },
    {
      "entropy": 5.512422657012939,
      "epoch": 2.1114168585387767,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004561135985052285,
      "loss": 5.0286,
      "mean_token_accuracy": 0.2150877073407173,
      "num_tokens": 41957929.0,
      "step": 18800
    },
    {
      "entropy": 5.47967381477356,
      "epoch": 2.1119784354467344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000456089773638076,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22358918637037278,
      "num_tokens": 41968677.0,
      "step": 18805
    },
    {
      "entropy": 5.5085124492645265,
      "epoch": 2.112540012354692,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004560659430049548,
      "loss": 5.0971,
      "mean_token_accuracy": 0.2126852124929428,
      "num_tokens": 41980330.0,
      "step": 18810
    },
    {
      "entropy": 5.518780851364136,
      "epoch": 2.1131015892626497,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004560421066066237,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21738462895154953,
      "num_tokens": 41991258.0,
      "step": 18815
    },
    {
      "entropy": 5.539147758483887,
      "epoch": 2.113663166170607,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045601826444384175,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21918980926275253,
      "num_tokens": 42002679.0,
      "step": 18820
    },
    {
      "entropy": 5.490826606750488,
      "epoch": 2.1142247430785646,
      "grad_norm": 1.0,
      "learning_rate": 0.00045599441651736805,
      "loss": 4.9698,
      "mean_token_accuracy": 0.214610455930233,
      "num_tokens": 42012845.0,
      "step": 18825
    },
    {
      "entropy": 5.548410558700562,
      "epoch": 2.1147863199865222,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045597056282796196,
      "loss": 5.0547,
      "mean_token_accuracy": 0.21569840759038925,
      "num_tokens": 42024461.0,
      "step": 18830
    },
    {
      "entropy": 5.499419832229615,
      "epoch": 2.11534789689448,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004559467033763829,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2102286010980606,
      "num_tokens": 42036996.0,
      "step": 18835
    },
    {
      "entropy": 5.581981801986695,
      "epoch": 2.115909473802437,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004559228381633907,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21612170338630676,
      "num_tokens": 42048731.0,
      "step": 18840
    },
    {
      "entropy": 5.587025928497314,
      "epoch": 2.1164710507103948,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045589896718974517,
      "loss": 5.0375,
      "mean_token_accuracy": 0.22045449316501617,
      "num_tokens": 42060396.0,
      "step": 18845
    },
    {
      "entropy": 5.3889978408813475,
      "epoch": 2.1170326276183524,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004558750904562063,
      "loss": 4.9424,
      "mean_token_accuracy": 0.21314848810434342,
      "num_tokens": 42072037.0,
      "step": 18850
    },
    {
      "entropy": 5.4552794933319095,
      "epoch": 2.11759420452631,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004558512079635345,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22514116019010544,
      "num_tokens": 42082577.0,
      "step": 18855
    },
    {
      "entropy": 5.402896308898926,
      "epoch": 2.1181557814342673,
      "grad_norm": 1.0,
      "learning_rate": 0.0004558273197124899,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22545453161001205,
      "num_tokens": 42093978.0,
      "step": 18860
    },
    {
      "entropy": 5.512389707565307,
      "epoch": 2.118717358342225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045580342570383347,
      "loss": 5.1168,
      "mean_token_accuracy": 0.21387167721986772,
      "num_tokens": 42105773.0,
      "step": 18865
    },
    {
      "entropy": 5.513806772232056,
      "epoch": 2.1192789352501826,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004557795259383258,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21928571462631224,
      "num_tokens": 42116232.0,
      "step": 18870
    },
    {
      "entropy": 5.491640377044678,
      "epoch": 2.11984051215814,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045575562041672794,
      "loss": 5.0465,
      "mean_token_accuracy": 0.21324197053909302,
      "num_tokens": 42127813.0,
      "step": 18875
    },
    {
      "entropy": 5.473332929611206,
      "epoch": 2.1204020890660975,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045573170913980094,
      "loss": 4.987,
      "mean_token_accuracy": 0.2134528711438179,
      "num_tokens": 42138054.0,
      "step": 18880
    },
    {
      "entropy": 5.572250032424927,
      "epoch": 2.120963665974055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045570779210830627,
      "loss": 5.0435,
      "mean_token_accuracy": 0.21087926924228667,
      "num_tokens": 42149575.0,
      "step": 18885
    },
    {
      "entropy": 5.518216562271118,
      "epoch": 2.121525242882013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004556838693230054,
      "loss": 5.0478,
      "mean_token_accuracy": 0.2156390815973282,
      "num_tokens": 42160824.0,
      "step": 18890
    },
    {
      "entropy": 5.522684335708618,
      "epoch": 2.12208681978997,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045565994078466005,
      "loss": 5.0881,
      "mean_token_accuracy": 0.21475025564432143,
      "num_tokens": 42171088.0,
      "step": 18895
    },
    {
      "entropy": 5.490110731124878,
      "epoch": 2.1226483966979277,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045563600649403213,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21549665033817292,
      "num_tokens": 42182112.0,
      "step": 18900
    },
    {
      "entropy": 5.543063020706176,
      "epoch": 2.1232099736058854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004556120664518838,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22617857158184052,
      "num_tokens": 42193563.0,
      "step": 18905
    },
    {
      "entropy": 5.523292398452758,
      "epoch": 2.123771550513843,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004555881206589771,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21772734969854354,
      "num_tokens": 42204891.0,
      "step": 18910
    },
    {
      "entropy": 5.535649156570434,
      "epoch": 2.1243331274218002,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004555641691160746,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2203079044818878,
      "num_tokens": 42217226.0,
      "step": 18915
    },
    {
      "entropy": 5.544731140136719,
      "epoch": 2.124894704329758,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000455540211823939,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21794724017381667,
      "num_tokens": 42228134.0,
      "step": 18920
    },
    {
      "entropy": 5.528731870651245,
      "epoch": 2.1254562812377156,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045551624878333296,
      "loss": 5.0669,
      "mean_token_accuracy": 0.21114152818918228,
      "num_tokens": 42239868.0,
      "step": 18925
    },
    {
      "entropy": 5.576525449752808,
      "epoch": 2.1260178581456732,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045549227999501957,
      "loss": 5.0968,
      "mean_token_accuracy": 0.21618064790964125,
      "num_tokens": 42250297.0,
      "step": 18930
    },
    {
      "entropy": 5.451301574707031,
      "epoch": 2.1265794350536305,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045546830545976203,
      "loss": 4.7772,
      "mean_token_accuracy": 0.24056749045848846,
      "num_tokens": 42261263.0,
      "step": 18935
    },
    {
      "entropy": 5.556567001342773,
      "epoch": 2.127141011961588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045544432517832356,
      "loss": 5.0823,
      "mean_token_accuracy": 0.2160559728741646,
      "num_tokens": 42272357.0,
      "step": 18940
    },
    {
      "entropy": 5.560354089736938,
      "epoch": 2.1277025888695458,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045542033915146787,
      "loss": 5.1073,
      "mean_token_accuracy": 0.2153716042637825,
      "num_tokens": 42283630.0,
      "step": 18945
    },
    {
      "entropy": 5.529127168655395,
      "epoch": 2.1282641657775034,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045539634737995857,
      "loss": 5.1096,
      "mean_token_accuracy": 0.19961658865213394,
      "num_tokens": 42294981.0,
      "step": 18950
    },
    {
      "entropy": 5.524041223526001,
      "epoch": 2.1288257426854607,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045537234986455964,
      "loss": 4.9596,
      "mean_token_accuracy": 0.22108154892921447,
      "num_tokens": 42304877.0,
      "step": 18955
    },
    {
      "entropy": 5.505650186538697,
      "epoch": 2.1293873195934183,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004553483466060351,
      "loss": 5.0668,
      "mean_token_accuracy": 0.21097273677587508,
      "num_tokens": 42316325.0,
      "step": 18960
    },
    {
      "entropy": 5.484247922897339,
      "epoch": 2.129948896501376,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045532433760514925,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21446882486343383,
      "num_tokens": 42327437.0,
      "step": 18965
    },
    {
      "entropy": 5.421216535568237,
      "epoch": 2.130510473409333,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004553003228626665,
      "loss": 4.933,
      "mean_token_accuracy": 0.22633096426725388,
      "num_tokens": 42337742.0,
      "step": 18970
    },
    {
      "entropy": 5.501980781555176,
      "epoch": 2.131072050317291,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004552763023793517,
      "loss": 5.0149,
      "mean_token_accuracy": 0.2165967270731926,
      "num_tokens": 42348511.0,
      "step": 18975
    },
    {
      "entropy": 5.494083499908447,
      "epoch": 2.1316336272252485,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004552522761559694,
      "loss": 4.992,
      "mean_token_accuracy": 0.21859514564275742,
      "num_tokens": 42358977.0,
      "step": 18980
    },
    {
      "entropy": 5.562210273742676,
      "epoch": 2.132195204133206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045522824419328475,
      "loss": 5.0601,
      "mean_token_accuracy": 0.212298221886158,
      "num_tokens": 42371609.0,
      "step": 18985
    },
    {
      "entropy": 5.544377565383911,
      "epoch": 2.1327567810411634,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045520420649206283,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21839981079101561,
      "num_tokens": 42383150.0,
      "step": 18990
    },
    {
      "entropy": 5.4789728164672855,
      "epoch": 2.133318357949121,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004551801630530692,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2150705099105835,
      "num_tokens": 42395150.0,
      "step": 18995
    },
    {
      "entropy": 5.522980403900147,
      "epoch": 2.1338799348570787,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004551561138770693,
      "loss": 5.0375,
      "mean_token_accuracy": 0.21387800127267836,
      "num_tokens": 42406348.0,
      "step": 19000
    },
    {
      "entropy": 5.5495411396026615,
      "epoch": 2.1344415117650364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045513205896482875,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20580437928438186,
      "num_tokens": 42417183.0,
      "step": 19005
    },
    {
      "entropy": 5.535960721969604,
      "epoch": 2.1350030886729936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045510799831711366,
      "loss": 4.986,
      "mean_token_accuracy": 0.22771431803703307,
      "num_tokens": 42427805.0,
      "step": 19010
    },
    {
      "entropy": 5.5317810535430905,
      "epoch": 2.1355646655809513,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045508393193469,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2168296217918396,
      "num_tokens": 42439521.0,
      "step": 19015
    },
    {
      "entropy": 5.602089262008667,
      "epoch": 2.136126242488909,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004550598598183241,
      "loss": 5.1065,
      "mean_token_accuracy": 0.21246589869260787,
      "num_tokens": 42451201.0,
      "step": 19020
    },
    {
      "entropy": 5.529308652877807,
      "epoch": 2.1366878193968666,
      "grad_norm": 1.0,
      "learning_rate": 0.0004550357819687825,
      "loss": 4.999,
      "mean_token_accuracy": 0.21780294477939605,
      "num_tokens": 42462315.0,
      "step": 19025
    },
    {
      "entropy": 5.632309818267823,
      "epoch": 2.137249396304824,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004550116983868317,
      "loss": 5.1645,
      "mean_token_accuracy": 0.20546071231365204,
      "num_tokens": 42473993.0,
      "step": 19030
    },
    {
      "entropy": 5.637898254394531,
      "epoch": 2.1378109732127815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004549876090732386,
      "loss": 5.1771,
      "mean_token_accuracy": 0.20699512213468552,
      "num_tokens": 42484307.0,
      "step": 19035
    },
    {
      "entropy": 5.539485931396484,
      "epoch": 2.138372550120739,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004549635140287703,
      "loss": 5.0689,
      "mean_token_accuracy": 0.21453823000192643,
      "num_tokens": 42497305.0,
      "step": 19040
    },
    {
      "entropy": 5.562845182418823,
      "epoch": 2.138934127028697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045493941325419385,
      "loss": 5.0828,
      "mean_token_accuracy": 0.20761047005653382,
      "num_tokens": 42508307.0,
      "step": 19045
    },
    {
      "entropy": 5.551513290405273,
      "epoch": 2.139495703936654,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004549153067502768,
      "loss": 5.1343,
      "mean_token_accuracy": 0.20893438458442687,
      "num_tokens": 42519946.0,
      "step": 19050
    },
    {
      "entropy": 5.484661388397217,
      "epoch": 2.1400572808446117,
      "grad_norm": 1.125,
      "learning_rate": 0.00045489119451778645,
      "loss": 4.91,
      "mean_token_accuracy": 0.2198232963681221,
      "num_tokens": 42530737.0,
      "step": 19055
    },
    {
      "entropy": 5.411928272247314,
      "epoch": 2.1406188577525693,
      "grad_norm": 1.0,
      "learning_rate": 0.0004548670765574908,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22011160850524902,
      "num_tokens": 42542184.0,
      "step": 19060
    },
    {
      "entropy": 5.51154317855835,
      "epoch": 2.1411804346605265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045484295287015757,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21582040190696716,
      "num_tokens": 42553816.0,
      "step": 19065
    },
    {
      "entropy": 5.57795524597168,
      "epoch": 2.141742011568484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004548188234565551,
      "loss": 5.0573,
      "mean_token_accuracy": 0.2195124953985214,
      "num_tokens": 42564685.0,
      "step": 19070
    },
    {
      "entropy": 5.503572797775268,
      "epoch": 2.142303588476442,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045479468831745144,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20874025225639342,
      "num_tokens": 42574864.0,
      "step": 19075
    },
    {
      "entropy": 5.508390474319458,
      "epoch": 2.1428651653843995,
      "grad_norm": 1.125,
      "learning_rate": 0.0004547705474536152,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21250364780426026,
      "num_tokens": 42585214.0,
      "step": 19080
    },
    {
      "entropy": 5.550159645080567,
      "epoch": 2.1434267422923567,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045474640086581506,
      "loss": 5.0957,
      "mean_token_accuracy": 0.21216932386159898,
      "num_tokens": 42596017.0,
      "step": 19085
    },
    {
      "entropy": 5.530079412460327,
      "epoch": 2.1439883192003144,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045472224855481975,
      "loss": 5.0013,
      "mean_token_accuracy": 0.20888324230909347,
      "num_tokens": 42607903.0,
      "step": 19090
    },
    {
      "entropy": 5.54863338470459,
      "epoch": 2.144549896108272,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045469809052139844,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2172306150197983,
      "num_tokens": 42618690.0,
      "step": 19095
    },
    {
      "entropy": 5.486841917037964,
      "epoch": 2.1451114730162297,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045467392676632016,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21515931338071823,
      "num_tokens": 42629655.0,
      "step": 19100
    },
    {
      "entropy": 5.526285457611084,
      "epoch": 2.145673049924187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004546497572903543,
      "loss": 5.0905,
      "mean_token_accuracy": 0.21219980865716934,
      "num_tokens": 42641214.0,
      "step": 19105
    },
    {
      "entropy": 5.580718994140625,
      "epoch": 2.1462346268321446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004546255820942706,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20636202543973922,
      "num_tokens": 42653829.0,
      "step": 19110
    },
    {
      "entropy": 5.581433629989624,
      "epoch": 2.1467962037401023,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004546014011788386,
      "loss": 5.0758,
      "mean_token_accuracy": 0.21148422807455064,
      "num_tokens": 42665245.0,
      "step": 19115
    },
    {
      "entropy": 5.520538568496704,
      "epoch": 2.14735778064806,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004545772145448284,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21245959550142288,
      "num_tokens": 42676409.0,
      "step": 19120
    },
    {
      "entropy": 5.443544673919678,
      "epoch": 2.147919357556017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045455302219301,
      "loss": 4.8613,
      "mean_token_accuracy": 0.2246546819806099,
      "num_tokens": 42686586.0,
      "step": 19125
    },
    {
      "entropy": 5.405819606781006,
      "epoch": 2.148480934463975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004545288241241538,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22098598182201384,
      "num_tokens": 42696597.0,
      "step": 19130
    },
    {
      "entropy": 5.57641191482544,
      "epoch": 2.1490425113719325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004545046203390302,
      "loss": 5.1157,
      "mean_token_accuracy": 0.21626488566398622,
      "num_tokens": 42707447.0,
      "step": 19135
    },
    {
      "entropy": 5.69161958694458,
      "epoch": 2.14960408827989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045448041083840973,
      "loss": 5.1268,
      "mean_token_accuracy": 0.20393733084201812,
      "num_tokens": 42717533.0,
      "step": 19140
    },
    {
      "entropy": 5.600433444976806,
      "epoch": 2.1501656651878474,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00045445619562306354,
      "loss": 5.1225,
      "mean_token_accuracy": 0.2111228510737419,
      "num_tokens": 42730563.0,
      "step": 19145
    },
    {
      "entropy": 5.502301836013794,
      "epoch": 2.150727242095805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045443197469376243,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2174910321831703,
      "num_tokens": 42740301.0,
      "step": 19150
    },
    {
      "entropy": 5.53596510887146,
      "epoch": 2.1512888190037627,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004544077480512776,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2109946295619011,
      "num_tokens": 42752296.0,
      "step": 19155
    },
    {
      "entropy": 5.483240842819214,
      "epoch": 2.15185039591172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045438351569638055,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21888632774353028,
      "num_tokens": 42762530.0,
      "step": 19160
    },
    {
      "entropy": 5.5139813899993895,
      "epoch": 2.1524119728196776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045435927762984277,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2133881226181984,
      "num_tokens": 42773059.0,
      "step": 19165
    },
    {
      "entropy": 5.4664778232574465,
      "epoch": 2.152973549727635,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000454335033852436,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2172984391450882,
      "num_tokens": 42783832.0,
      "step": 19170
    },
    {
      "entropy": 5.519624805450439,
      "epoch": 2.153535126635593,
      "grad_norm": 1.0,
      "learning_rate": 0.0004543107843649322,
      "loss": 5.0948,
      "mean_token_accuracy": 0.21813974231481553,
      "num_tokens": 42794565.0,
      "step": 19175
    },
    {
      "entropy": 5.631042385101319,
      "epoch": 2.15409670354355,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004542865291681035,
      "loss": 5.1274,
      "mean_token_accuracy": 0.20971587300300598,
      "num_tokens": 42805536.0,
      "step": 19180
    },
    {
      "entropy": 5.571031951904297,
      "epoch": 2.1546582804515078,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045426226826272213,
      "loss": 5.064,
      "mean_token_accuracy": 0.20922781378030778,
      "num_tokens": 42816718.0,
      "step": 19185
    },
    {
      "entropy": 5.5425605297088625,
      "epoch": 2.1552198573594654,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004542380016495607,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21452638655900955,
      "num_tokens": 42830183.0,
      "step": 19190
    },
    {
      "entropy": 5.487700891494751,
      "epoch": 2.155781434267423,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045421372932939177,
      "loss": 5.0396,
      "mean_token_accuracy": 0.2164797767996788,
      "num_tokens": 42841241.0,
      "step": 19195
    },
    {
      "entropy": 5.686842346191407,
      "epoch": 2.1563430111753803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004541894513029881,
      "loss": 5.1241,
      "mean_token_accuracy": 0.21199548095464707,
      "num_tokens": 42852775.0,
      "step": 19200
    },
    {
      "entropy": 5.735410356521607,
      "epoch": 2.156904588083338,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045416516757112286,
      "loss": 5.2241,
      "mean_token_accuracy": 0.2041873574256897,
      "num_tokens": 42863947.0,
      "step": 19205
    },
    {
      "entropy": 5.49827013015747,
      "epoch": 2.1574661649912956,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045414087813456925,
      "loss": 5.0702,
      "mean_token_accuracy": 0.22008589953184127,
      "num_tokens": 42874794.0,
      "step": 19210
    },
    {
      "entropy": 5.524099588394165,
      "epoch": 2.1580277418992533,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004541165829941006,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21243919432163239,
      "num_tokens": 42886038.0,
      "step": 19215
    },
    {
      "entropy": 5.558943223953247,
      "epoch": 2.1585893188072105,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004540922821504904,
      "loss": 5.0481,
      "mean_token_accuracy": 0.2110162779688835,
      "num_tokens": 42897243.0,
      "step": 19220
    },
    {
      "entropy": 5.507229900360107,
      "epoch": 2.159150895715168,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004540679756045125,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21356573402881623,
      "num_tokens": 42908802.0,
      "step": 19225
    },
    {
      "entropy": 5.589541864395142,
      "epoch": 2.159712472623126,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004540436633569407,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20550845712423324,
      "num_tokens": 42921156.0,
      "step": 19230
    },
    {
      "entropy": 5.45367693901062,
      "epoch": 2.1602740495310835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045401934540854937,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21830828934907914,
      "num_tokens": 42931336.0,
      "step": 19235
    },
    {
      "entropy": 5.53152871131897,
      "epoch": 2.1608356264390407,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004539950217601127,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21999713629484177,
      "num_tokens": 42942312.0,
      "step": 19240
    },
    {
      "entropy": 5.552735233306885,
      "epoch": 2.1613972033469984,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045397069241240497,
      "loss": 5.0736,
      "mean_token_accuracy": 0.21625786721706391,
      "num_tokens": 42953429.0,
      "step": 19245
    },
    {
      "entropy": 5.467890739440918,
      "epoch": 2.161958780254956,
      "grad_norm": 1.078125,
      "learning_rate": 0.000453946357366201,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21559067368507384,
      "num_tokens": 42963354.0,
      "step": 19250
    },
    {
      "entropy": 5.603626823425293,
      "epoch": 2.1625203571629132,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045392201662227565,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2173087030649185,
      "num_tokens": 42974643.0,
      "step": 19255
    },
    {
      "entropy": 5.550736665725708,
      "epoch": 2.163081934070871,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045389767018140395,
      "loss": 5.085,
      "mean_token_accuracy": 0.21206020712852477,
      "num_tokens": 42985821.0,
      "step": 19260
    },
    {
      "entropy": 5.5998992919921875,
      "epoch": 2.1636435109788286,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045387331804436096,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20934343039989473,
      "num_tokens": 42998258.0,
      "step": 19265
    },
    {
      "entropy": 5.534904289245605,
      "epoch": 2.1642050878867862,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004538489602119222,
      "loss": 5.0076,
      "mean_token_accuracy": 0.2169676750898361,
      "num_tokens": 43009117.0,
      "step": 19270
    },
    {
      "entropy": 5.47719292640686,
      "epoch": 2.1647666647947434,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004538245966848632,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21428176313638686,
      "num_tokens": 43020437.0,
      "step": 19275
    },
    {
      "entropy": 5.530072832107544,
      "epoch": 2.165328241702701,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045380022746395953,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21465242952108382,
      "num_tokens": 43031722.0,
      "step": 19280
    },
    {
      "entropy": 5.519904232025146,
      "epoch": 2.1658898186106588,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045377585254998743,
      "loss": 4.9926,
      "mean_token_accuracy": 0.22443864792585372,
      "num_tokens": 43042957.0,
      "step": 19285
    },
    {
      "entropy": 5.513291072845459,
      "epoch": 2.1664513955186164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004537514719437228,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21128850877285005,
      "num_tokens": 43054086.0,
      "step": 19290
    },
    {
      "entropy": 5.608696413040161,
      "epoch": 2.1670129724265736,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045372708564594196,
      "loss": 5.1572,
      "mean_token_accuracy": 0.20577065497636796,
      "num_tokens": 43066259.0,
      "step": 19295
    },
    {
      "entropy": 5.464995527267456,
      "epoch": 2.1675745493345313,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045370269365742143,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22134891301393508,
      "num_tokens": 43076837.0,
      "step": 19300
    },
    {
      "entropy": 5.496037435531616,
      "epoch": 2.168136126242489,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045367829597893776,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22162239402532577,
      "num_tokens": 43087802.0,
      "step": 19305
    },
    {
      "entropy": 5.505429935455322,
      "epoch": 2.1686977031504466,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045365389261126793,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21637981086969377,
      "num_tokens": 43097562.0,
      "step": 19310
    },
    {
      "entropy": 5.454412126541138,
      "epoch": 2.169259280058404,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045362948355518883,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21911730319261552,
      "num_tokens": 43107973.0,
      "step": 19315
    },
    {
      "entropy": 5.451364326477051,
      "epoch": 2.1698208569663615,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004536050688114777,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22173949480056762,
      "num_tokens": 43118923.0,
      "step": 19320
    },
    {
      "entropy": 5.492047691345215,
      "epoch": 2.170382433874319,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045358064838091187,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2170213744044304,
      "num_tokens": 43130904.0,
      "step": 19325
    },
    {
      "entropy": 5.528943061828613,
      "epoch": 2.170944010782277,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045355622226426903,
      "loss": 4.9934,
      "mean_token_accuracy": 0.2131267160177231,
      "num_tokens": 43142293.0,
      "step": 19330
    },
    {
      "entropy": 5.552287960052491,
      "epoch": 2.171505587690234,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045353179046232674,
      "loss": 5.0463,
      "mean_token_accuracy": 0.22204008549451829,
      "num_tokens": 43154340.0,
      "step": 19335
    },
    {
      "entropy": 5.519361162185669,
      "epoch": 2.1720671645981917,
      "grad_norm": 1.0,
      "learning_rate": 0.000453507352975863,
      "loss": 5.0071,
      "mean_token_accuracy": 0.22217732667922974,
      "num_tokens": 43165924.0,
      "step": 19340
    },
    {
      "entropy": 5.468971204757691,
      "epoch": 2.1726287415061494,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045348290980565595,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21756821870803833,
      "num_tokens": 43177394.0,
      "step": 19345
    },
    {
      "entropy": 5.5049591064453125,
      "epoch": 2.1731903184141066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045345846095248376,
      "loss": 4.9464,
      "mean_token_accuracy": 0.21580301225185394,
      "num_tokens": 43188043.0,
      "step": 19350
    },
    {
      "entropy": 5.520325374603272,
      "epoch": 2.1737518953220643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045343400641712503,
      "loss": 5.1498,
      "mean_token_accuracy": 0.20727440118789672,
      "num_tokens": 43199466.0,
      "step": 19355
    },
    {
      "entropy": 5.437262487411499,
      "epoch": 2.174313472230022,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045340954620035827,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21733541935682296,
      "num_tokens": 43210673.0,
      "step": 19360
    },
    {
      "entropy": 5.536609745025634,
      "epoch": 2.1748750491379796,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045338508030296246,
      "loss": 5.059,
      "mean_token_accuracy": 0.214582721889019,
      "num_tokens": 43221263.0,
      "step": 19365
    },
    {
      "entropy": 5.548491907119751,
      "epoch": 2.175436626045937,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045336060872571646,
      "loss": 5.065,
      "mean_token_accuracy": 0.21778450161218643,
      "num_tokens": 43232590.0,
      "step": 19370
    },
    {
      "entropy": 5.604807901382446,
      "epoch": 2.1759982029538945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004533361314693994,
      "loss": 5.0924,
      "mean_token_accuracy": 0.21733055412769317,
      "num_tokens": 43243693.0,
      "step": 19375
    },
    {
      "entropy": 5.541410493850708,
      "epoch": 2.176559779861852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045331164853479084,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21512263119220734,
      "num_tokens": 43254526.0,
      "step": 19380
    },
    {
      "entropy": 5.607054567337036,
      "epoch": 2.17712135676981,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045328715992267024,
      "loss": 5.1062,
      "mean_token_accuracy": 0.21070781350135803,
      "num_tokens": 43264490.0,
      "step": 19385
    },
    {
      "entropy": 5.511202239990235,
      "epoch": 2.177682933677767,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045326266563381734,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21723948717117308,
      "num_tokens": 43275012.0,
      "step": 19390
    },
    {
      "entropy": 5.492418098449707,
      "epoch": 2.1782445105857247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045323816566901205,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21589162498712539,
      "num_tokens": 43285194.0,
      "step": 19395
    },
    {
      "entropy": 5.527652072906494,
      "epoch": 2.1788060874936823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004532136600290343,
      "loss": 5.062,
      "mean_token_accuracy": 0.21967443376779555,
      "num_tokens": 43298159.0,
      "step": 19400
    },
    {
      "entropy": 5.570392704010009,
      "epoch": 2.17936766440164,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004531891487146646,
      "loss": 5.1036,
      "mean_token_accuracy": 0.20800223648548127,
      "num_tokens": 43309664.0,
      "step": 19405
    },
    {
      "entropy": 5.514455366134643,
      "epoch": 2.179929241309597,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004531646317266833,
      "loss": 5.032,
      "mean_token_accuracy": 0.21771198511123657,
      "num_tokens": 43320399.0,
      "step": 19410
    },
    {
      "entropy": 5.465872097015381,
      "epoch": 2.180490818217555,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045314010906587096,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21656649559736252,
      "num_tokens": 43331510.0,
      "step": 19415
    },
    {
      "entropy": 5.597373008728027,
      "epoch": 2.1810523951255125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045311558073300853,
      "loss": 5.0953,
      "mean_token_accuracy": 0.21858106553554535,
      "num_tokens": 43342844.0,
      "step": 19420
    },
    {
      "entropy": 5.47257513999939,
      "epoch": 2.18161397203347,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004530910467288769,
      "loss": 4.9545,
      "mean_token_accuracy": 0.2133133366703987,
      "num_tokens": 43352556.0,
      "step": 19425
    },
    {
      "entropy": 5.419732332229614,
      "epoch": 2.1821755489414274,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045306650705425727,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22044442296028138,
      "num_tokens": 43363357.0,
      "step": 19430
    },
    {
      "entropy": 5.568076324462891,
      "epoch": 2.182737125849385,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045304196170993093,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21361021995544432,
      "num_tokens": 43374697.0,
      "step": 19435
    },
    {
      "entropy": 5.649209833145141,
      "epoch": 2.1832987027573427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004530174106966795,
      "loss": 5.1087,
      "mean_token_accuracy": 0.20899368971586227,
      "num_tokens": 43386826.0,
      "step": 19440
    },
    {
      "entropy": 5.501683473587036,
      "epoch": 2.1838602796653,
      "grad_norm": 1.0,
      "learning_rate": 0.00045299285401528475,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21592700481414795,
      "num_tokens": 43398352.0,
      "step": 19445
    },
    {
      "entropy": 5.53292293548584,
      "epoch": 2.1844218565732576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004529682916665285,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22032152265310287,
      "num_tokens": 43409509.0,
      "step": 19450
    },
    {
      "entropy": 5.507028532028198,
      "epoch": 2.1849834334812153,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045294372365119276,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21619485318660736,
      "num_tokens": 43420195.0,
      "step": 19455
    },
    {
      "entropy": 5.533158302307129,
      "epoch": 2.185545010389173,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004529191499700598,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21609335988759995,
      "num_tokens": 43430599.0,
      "step": 19460
    },
    {
      "entropy": 5.442693471908569,
      "epoch": 2.18610658729713,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004528945706239122,
      "loss": 4.9188,
      "mean_token_accuracy": 0.2183141827583313,
      "num_tokens": 43442212.0,
      "step": 19465
    },
    {
      "entropy": 5.478957939147949,
      "epoch": 2.186668164205088,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045286998561353236,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21876858025789261,
      "num_tokens": 43452631.0,
      "step": 19470
    },
    {
      "entropy": 5.461137580871582,
      "epoch": 2.1872297411130455,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004528453949397033,
      "loss": 5.0195,
      "mean_token_accuracy": 0.21476077884435654,
      "num_tokens": 43463323.0,
      "step": 19475
    },
    {
      "entropy": 5.491626024246216,
      "epoch": 2.187791318021003,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045282079860320786,
      "loss": 5.0666,
      "mean_token_accuracy": 0.20933491438627244,
      "num_tokens": 43474451.0,
      "step": 19480
    },
    {
      "entropy": 5.475252437591553,
      "epoch": 2.1883528949289603,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045279619660482925,
      "loss": 4.927,
      "mean_token_accuracy": 0.22053967416286469,
      "num_tokens": 43485948.0,
      "step": 19485
    },
    {
      "entropy": 5.5854511737823485,
      "epoch": 2.188914471836918,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045277158894535075,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21937210112810135,
      "num_tokens": 43498955.0,
      "step": 19490
    },
    {
      "entropy": 5.527094602584839,
      "epoch": 2.1894760487448757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045274697562555584,
      "loss": 5.0003,
      "mean_token_accuracy": 0.22366307973861693,
      "num_tokens": 43509416.0,
      "step": 19495
    },
    {
      "entropy": 5.487946462631226,
      "epoch": 2.1900376256528333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004527223566462284,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21759290993213654,
      "num_tokens": 43520387.0,
      "step": 19500
    },
    {
      "entropy": 5.45314302444458,
      "epoch": 2.1905992025607905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004526977320081522,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21430032700300217,
      "num_tokens": 43532006.0,
      "step": 19505
    },
    {
      "entropy": 5.454686021804809,
      "epoch": 2.191160779468748,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004526731017121111,
      "loss": 4.9482,
      "mean_token_accuracy": 0.21870541721582412,
      "num_tokens": 43542303.0,
      "step": 19510
    },
    {
      "entropy": 5.57761664390564,
      "epoch": 2.191722356376706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045264846575888973,
      "loss": 5.0942,
      "mean_token_accuracy": 0.21262185722589494,
      "num_tokens": 43553711.0,
      "step": 19515
    },
    {
      "entropy": 5.44252233505249,
      "epoch": 2.1922839332846635,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045262382414927235,
      "loss": 4.9724,
      "mean_token_accuracy": 0.2136426106095314,
      "num_tokens": 43564488.0,
      "step": 19520
    },
    {
      "entropy": 5.476780605316162,
      "epoch": 2.1928455101926208,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045259917688404336,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21852104216814042,
      "num_tokens": 43575414.0,
      "step": 19525
    },
    {
      "entropy": 5.461527681350708,
      "epoch": 2.1934070871005784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004525745239639877,
      "loss": 4.9679,
      "mean_token_accuracy": 0.21877976804971694,
      "num_tokens": 43586789.0,
      "step": 19530
    },
    {
      "entropy": 5.484581422805786,
      "epoch": 2.193968664008536,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045254986538989045,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22550459802150727,
      "num_tokens": 43597423.0,
      "step": 19535
    },
    {
      "entropy": 5.510950517654419,
      "epoch": 2.1945302409164933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004525252011625366,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2122252494096756,
      "num_tokens": 43606703.0,
      "step": 19540
    },
    {
      "entropy": 5.589262819290161,
      "epoch": 2.195091817824451,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004525005312827114,
      "loss": 5.0935,
      "mean_token_accuracy": 0.21470154374837874,
      "num_tokens": 43617869.0,
      "step": 19545
    },
    {
      "entropy": 5.58360652923584,
      "epoch": 2.1956533947324086,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004524758557512005,
      "loss": 5.1733,
      "mean_token_accuracy": 0.2136438637971878,
      "num_tokens": 43629102.0,
      "step": 19550
    },
    {
      "entropy": 5.542883443832397,
      "epoch": 2.1962149716403663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004524511745687895,
      "loss": 5.0668,
      "mean_token_accuracy": 0.2179759442806244,
      "num_tokens": 43640421.0,
      "step": 19555
    },
    {
      "entropy": 5.5498213291168215,
      "epoch": 2.1967765485483235,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004524264877362643,
      "loss": 4.972,
      "mean_token_accuracy": 0.22218890488147736,
      "num_tokens": 43650516.0,
      "step": 19560
    },
    {
      "entropy": 5.515789604187011,
      "epoch": 2.197338125456281,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045240179525441094,
      "loss": 4.9934,
      "mean_token_accuracy": 0.21932897269725798,
      "num_tokens": 43661439.0,
      "step": 19565
    },
    {
      "entropy": 5.520218133926392,
      "epoch": 2.197899702364239,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004523770971240156,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21577452272176742,
      "num_tokens": 43672137.0,
      "step": 19570
    },
    {
      "entropy": 5.535954761505127,
      "epoch": 2.1984612792721965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004523523933458647,
      "loss": 5.0347,
      "mean_token_accuracy": 0.21461211144924164,
      "num_tokens": 43683612.0,
      "step": 19575
    },
    {
      "entropy": 5.536217164993286,
      "epoch": 2.1990228561801537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004523276839207448,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2166285455226898,
      "num_tokens": 43694331.0,
      "step": 19580
    },
    {
      "entropy": 5.55322756767273,
      "epoch": 2.1995844330881114,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045230296884944264,
      "loss": 5.0796,
      "mean_token_accuracy": 0.2109744966030121,
      "num_tokens": 43706572.0,
      "step": 19585
    },
    {
      "entropy": 5.477943086624146,
      "epoch": 2.200146009996069,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045227824813274517,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22842492312192916,
      "num_tokens": 43718472.0,
      "step": 19590
    },
    {
      "entropy": 5.56979660987854,
      "epoch": 2.2007075869040267,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004522535217714396,
      "loss": 5.0701,
      "mean_token_accuracy": 0.216226464509964,
      "num_tokens": 43729642.0,
      "step": 19595
    },
    {
      "entropy": 5.534376382827759,
      "epoch": 2.201269163811984,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004522287897663131,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20783924907445908,
      "num_tokens": 43739846.0,
      "step": 19600
    },
    {
      "entropy": 5.482331895828247,
      "epoch": 2.2018307407199416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045220405211815327,
      "loss": 4.9256,
      "mean_token_accuracy": 0.21548637747764587,
      "num_tokens": 43750265.0,
      "step": 19605
    },
    {
      "entropy": 5.557156610488891,
      "epoch": 2.202392317627899,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045217930882774754,
      "loss": 5.0322,
      "mean_token_accuracy": 0.2115569144487381,
      "num_tokens": 43761245.0,
      "step": 19610
    },
    {
      "entropy": 5.545237302780151,
      "epoch": 2.202953894535857,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045215455989588396,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2166493773460388,
      "num_tokens": 43772677.0,
      "step": 19615
    },
    {
      "entropy": 5.551075410842896,
      "epoch": 2.203515471443814,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045212980532335055,
      "loss": 5.093,
      "mean_token_accuracy": 0.2074023574590683,
      "num_tokens": 43783128.0,
      "step": 19620
    },
    {
      "entropy": 5.4757905960083,
      "epoch": 2.2040770483517718,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045210504511093534,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21937112808227538,
      "num_tokens": 43795323.0,
      "step": 19625
    },
    {
      "entropy": 5.618589115142822,
      "epoch": 2.2046386252597294,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004520802792594269,
      "loss": 5.1022,
      "mean_token_accuracy": 0.21376935243606568,
      "num_tokens": 43806917.0,
      "step": 19630
    },
    {
      "entropy": 5.644349241256714,
      "epoch": 2.2052002021676866,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045205550776961366,
      "loss": 5.0859,
      "mean_token_accuracy": 0.21520653516054153,
      "num_tokens": 43818963.0,
      "step": 19635
    },
    {
      "entropy": 5.464762830734253,
      "epoch": 2.2057617790756443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004520307306422844,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20970046669244766,
      "num_tokens": 43830163.0,
      "step": 19640
    },
    {
      "entropy": 5.514451122283935,
      "epoch": 2.206323355983602,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045200594787822795,
      "loss": 4.97,
      "mean_token_accuracy": 0.22079598754644394,
      "num_tokens": 43840434.0,
      "step": 19645
    },
    {
      "entropy": 5.401845264434814,
      "epoch": 2.2068849328915596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004519811594782335,
      "loss": 4.8375,
      "mean_token_accuracy": 0.22526646703481673,
      "num_tokens": 43851086.0,
      "step": 19650
    },
    {
      "entropy": 5.492336988449097,
      "epoch": 2.207446509799517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045195636544309033,
      "loss": 5.0499,
      "mean_token_accuracy": 0.21033890098333358,
      "num_tokens": 43862834.0,
      "step": 19655
    },
    {
      "entropy": 5.5211890697479244,
      "epoch": 2.2080080867074745,
      "grad_norm": 1.0,
      "learning_rate": 0.00045193156577358784,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20864915996789932,
      "num_tokens": 43874868.0,
      "step": 19660
    },
    {
      "entropy": 5.549234008789062,
      "epoch": 2.208569663615432,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004519067604705157,
      "loss": 5.0302,
      "mean_token_accuracy": 0.22280402034521102,
      "num_tokens": 43886175.0,
      "step": 19665
    },
    {
      "entropy": 5.496849155426025,
      "epoch": 2.20913124052339,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045188194953466363,
      "loss": 4.9989,
      "mean_token_accuracy": 0.22176545709371567,
      "num_tokens": 43896489.0,
      "step": 19670
    },
    {
      "entropy": 5.4805910110473635,
      "epoch": 2.209692817431347,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045185713296682175,
      "loss": 5.0,
      "mean_token_accuracy": 0.21901033371686934,
      "num_tokens": 43907759.0,
      "step": 19675
    },
    {
      "entropy": 5.498042726516724,
      "epoch": 2.2102543943393047,
      "grad_norm": 1.0,
      "learning_rate": 0.00045183231076778005,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21929456740617753,
      "num_tokens": 43918136.0,
      "step": 19680
    },
    {
      "entropy": 5.5516180992126465,
      "epoch": 2.2108159712472624,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045180748293832906,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21394529938697815,
      "num_tokens": 43929662.0,
      "step": 19685
    },
    {
      "entropy": 5.522339153289795,
      "epoch": 2.21137754815522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045178264947925924,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2238750174641609,
      "num_tokens": 43939599.0,
      "step": 19690
    },
    {
      "entropy": 5.50408673286438,
      "epoch": 2.2119391250631772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045175781039136133,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22214894294738768,
      "num_tokens": 43949728.0,
      "step": 19695
    },
    {
      "entropy": 5.580263423919678,
      "epoch": 2.212500701971135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004517329656754261,
      "loss": 5.0698,
      "mean_token_accuracy": 0.21721083223819732,
      "num_tokens": 43961147.0,
      "step": 19700
    },
    {
      "entropy": 5.580533885955811,
      "epoch": 2.2130622788790926,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004517081153322448,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2218971237540245,
      "num_tokens": 43972541.0,
      "step": 19705
    },
    {
      "entropy": 5.512256240844726,
      "epoch": 2.2136238557870502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004516832593626085,
      "loss": 5.0643,
      "mean_token_accuracy": 0.21487784683704375,
      "num_tokens": 43983540.0,
      "step": 19710
    },
    {
      "entropy": 5.476344633102417,
      "epoch": 2.2141854326950074,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004516583977673087,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22504897713661193,
      "num_tokens": 43994255.0,
      "step": 19715
    },
    {
      "entropy": 5.614608240127564,
      "epoch": 2.214747009602965,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045163353054713704,
      "loss": 5.1261,
      "mean_token_accuracy": 0.21330010890960693,
      "num_tokens": 44004364.0,
      "step": 19720
    },
    {
      "entropy": 5.569414901733398,
      "epoch": 2.2153085865109228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004516086577028852,
      "loss": 5.104,
      "mean_token_accuracy": 0.2123065322637558,
      "num_tokens": 44015487.0,
      "step": 19725
    },
    {
      "entropy": 5.60509934425354,
      "epoch": 2.21587016341888,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004515837792353453,
      "loss": 5.0749,
      "mean_token_accuracy": 0.21828112602233887,
      "num_tokens": 44026974.0,
      "step": 19730
    },
    {
      "entropy": 5.560262537002563,
      "epoch": 2.2164317403268377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004515588951453093,
      "loss": 5.0123,
      "mean_token_accuracy": 0.21647901386022567,
      "num_tokens": 44038718.0,
      "step": 19735
    },
    {
      "entropy": 5.4815867900848385,
      "epoch": 2.2169933172347953,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004515340054335696,
      "loss": 4.9396,
      "mean_token_accuracy": 0.22086912840604783,
      "num_tokens": 44048821.0,
      "step": 19740
    },
    {
      "entropy": 5.544683313369751,
      "epoch": 2.217554894142753,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004515091101009188,
      "loss": 5.1085,
      "mean_token_accuracy": 0.2082324281334877,
      "num_tokens": 44060973.0,
      "step": 19745
    },
    {
      "entropy": 5.653628826141357,
      "epoch": 2.21811647105071,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004514842091481493,
      "loss": 5.0808,
      "mean_token_accuracy": 0.21973627656698227,
      "num_tokens": 44073595.0,
      "step": 19750
    },
    {
      "entropy": 5.477323246002197,
      "epoch": 2.218678047958668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004514593025760543,
      "loss": 5.0063,
      "mean_token_accuracy": 0.2132320448756218,
      "num_tokens": 44084860.0,
      "step": 19755
    },
    {
      "entropy": 5.522267055511475,
      "epoch": 2.2192396248666255,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045143439038542653,
      "loss": 5.094,
      "mean_token_accuracy": 0.20872837901115418,
      "num_tokens": 44096026.0,
      "step": 19760
    },
    {
      "entropy": 5.55786247253418,
      "epoch": 2.219801201774583,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004514094725770594,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21249134987592697,
      "num_tokens": 44107981.0,
      "step": 19765
    },
    {
      "entropy": 5.548605632781983,
      "epoch": 2.2203627786825404,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045138454915174625,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21582875400781631,
      "num_tokens": 44119589.0,
      "step": 19770
    },
    {
      "entropy": 5.496036195755005,
      "epoch": 2.220924355590498,
      "grad_norm": 1.0,
      "learning_rate": 0.00045135962011028064,
      "loss": 5.0718,
      "mean_token_accuracy": 0.21270845234394073,
      "num_tokens": 44131094.0,
      "step": 19775
    },
    {
      "entropy": 5.541290235519409,
      "epoch": 2.2214859324984557,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004513346854534564,
      "loss": 5.0551,
      "mean_token_accuracy": 0.2077616348862648,
      "num_tokens": 44141348.0,
      "step": 19780
    },
    {
      "entropy": 5.641642665863037,
      "epoch": 2.2220475094064134,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004513097451820673,
      "loss": 5.1642,
      "mean_token_accuracy": 0.21126456409692765,
      "num_tokens": 44152085.0,
      "step": 19785
    },
    {
      "entropy": 5.537990522384644,
      "epoch": 2.2226090863143706,
      "grad_norm": 1.0,
      "learning_rate": 0.00045128479929690756,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22720523476600646,
      "num_tokens": 44163518.0,
      "step": 19790
    },
    {
      "entropy": 5.46861572265625,
      "epoch": 2.2231706632223283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004512598477987714,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20946355760097504,
      "num_tokens": 44175513.0,
      "step": 19795
    },
    {
      "entropy": 5.574788045883179,
      "epoch": 2.223732240130286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045123489068845334,
      "loss": 5.0243,
      "mean_token_accuracy": 0.2166624680161476,
      "num_tokens": 44186660.0,
      "step": 19800
    },
    {
      "entropy": 5.541599130630493,
      "epoch": 2.2242938170382436,
      "grad_norm": 1.0,
      "learning_rate": 0.00045120992796674804,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20813322812318802,
      "num_tokens": 44197613.0,
      "step": 19805
    },
    {
      "entropy": 5.471748161315918,
      "epoch": 2.224855393946201,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004511849596344503,
      "loss": 5.0073,
      "mean_token_accuracy": 0.20724640190601348,
      "num_tokens": 44209469.0,
      "step": 19810
    },
    {
      "entropy": 5.537791633605957,
      "epoch": 2.2254169708541585,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045115998569235507,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2121313765645027,
      "num_tokens": 44220672.0,
      "step": 19815
    },
    {
      "entropy": 5.571954822540283,
      "epoch": 2.225978547762116,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045113500614125756,
      "loss": 5.0662,
      "mean_token_accuracy": 0.20908642262220384,
      "num_tokens": 44232867.0,
      "step": 19820
    },
    {
      "entropy": 5.557937288284302,
      "epoch": 2.2265401246700733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004511100209819532,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21383792161941528,
      "num_tokens": 44245567.0,
      "step": 19825
    },
    {
      "entropy": 5.53202714920044,
      "epoch": 2.227101701578031,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045108503021523737,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22519684731960296,
      "num_tokens": 44257449.0,
      "step": 19830
    },
    {
      "entropy": 5.527451992034912,
      "epoch": 2.2276632784859887,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045106003384190604,
      "loss": 5.095,
      "mean_token_accuracy": 0.21207186728715896,
      "num_tokens": 44268890.0,
      "step": 19835
    },
    {
      "entropy": 5.5825995922088625,
      "epoch": 2.2282248553939463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045103503186275486,
      "loss": 5.0491,
      "mean_token_accuracy": 0.20933411717414857,
      "num_tokens": 44280719.0,
      "step": 19840
    },
    {
      "entropy": 5.617470407485962,
      "epoch": 2.2287864323019035,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045101002427857997,
      "loss": 5.1412,
      "mean_token_accuracy": 0.20498011857271195,
      "num_tokens": 44292567.0,
      "step": 19845
    },
    {
      "entropy": 5.576122188568116,
      "epoch": 2.229348009209861,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045098501109017765,
      "loss": 5.0723,
      "mean_token_accuracy": 0.2124025121331215,
      "num_tokens": 44304177.0,
      "step": 19850
    },
    {
      "entropy": 5.5016676902771,
      "epoch": 2.229909586117819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045095999229834437,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2205975517630577,
      "num_tokens": 44315142.0,
      "step": 19855
    },
    {
      "entropy": 5.487029409408569,
      "epoch": 2.2304711630257765,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004509349679038767,
      "loss": 5.1335,
      "mean_token_accuracy": 0.2149762272834778,
      "num_tokens": 44327457.0,
      "step": 19860
    },
    {
      "entropy": 5.617055034637451,
      "epoch": 2.2310327399337337,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045090993790757137,
      "loss": 5.1292,
      "mean_token_accuracy": 0.20898691266775132,
      "num_tokens": 44338978.0,
      "step": 19865
    },
    {
      "entropy": 5.61577033996582,
      "epoch": 2.2315943168416914,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004508849023102253,
      "loss": 5.114,
      "mean_token_accuracy": 0.21508942693471908,
      "num_tokens": 44349121.0,
      "step": 19870
    },
    {
      "entropy": 5.503602170944214,
      "epoch": 2.232155893749649,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004508598611126359,
      "loss": 4.99,
      "mean_token_accuracy": 0.22268788665533065,
      "num_tokens": 44360753.0,
      "step": 19875
    },
    {
      "entropy": 5.5981050491333,
      "epoch": 2.2327174706576067,
      "grad_norm": 1.0,
      "learning_rate": 0.0004508348143156002,
      "loss": 5.0315,
      "mean_token_accuracy": 0.22305696308612824,
      "num_tokens": 44371977.0,
      "step": 19880
    },
    {
      "entropy": 5.505199670791626,
      "epoch": 2.233279047565564,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045080976191991587,
      "loss": 4.966,
      "mean_token_accuracy": 0.2189818173646927,
      "num_tokens": 44383077.0,
      "step": 19885
    },
    {
      "entropy": 5.50550217628479,
      "epoch": 2.2338406244735216,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004507847039263805,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2122347518801689,
      "num_tokens": 44394827.0,
      "step": 19890
    },
    {
      "entropy": 5.4938231945037845,
      "epoch": 2.2344022013814793,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000450759640335792,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2212918609380722,
      "num_tokens": 44406594.0,
      "step": 19895
    },
    {
      "entropy": 5.478568744659424,
      "epoch": 2.234963778289437,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004507345711489484,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22228638231754302,
      "num_tokens": 44417842.0,
      "step": 19900
    },
    {
      "entropy": 5.533528900146484,
      "epoch": 2.235525355197394,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045070949636664785,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2144067719578743,
      "num_tokens": 44428234.0,
      "step": 19905
    },
    {
      "entropy": 5.443822908401489,
      "epoch": 2.236086932105352,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004506844159896888,
      "loss": 4.9712,
      "mean_token_accuracy": 0.2197059229016304,
      "num_tokens": 44437504.0,
      "step": 19910
    },
    {
      "entropy": 5.538375663757324,
      "epoch": 2.2366485090133095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004506593300188698,
      "loss": 5.0336,
      "mean_token_accuracy": 0.22090145498514174,
      "num_tokens": 44447873.0,
      "step": 19915
    },
    {
      "entropy": 5.532779502868652,
      "epoch": 2.2372100859212667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004506342384549897,
      "loss": 5.0847,
      "mean_token_accuracy": 0.21877450346946717,
      "num_tokens": 44458990.0,
      "step": 19920
    },
    {
      "entropy": 5.505418348312378,
      "epoch": 2.2377716628292244,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004506091412988472,
      "loss": 5.0027,
      "mean_token_accuracy": 0.21689185500144958,
      "num_tokens": 44469374.0,
      "step": 19925
    },
    {
      "entropy": 5.502139234542847,
      "epoch": 2.238333239737182,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045058403855124165,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21908223628997803,
      "num_tokens": 44479629.0,
      "step": 19930
    },
    {
      "entropy": 5.562967014312744,
      "epoch": 2.2388948166451397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004505589302129721,
      "loss": 5.0676,
      "mean_token_accuracy": 0.2127097502350807,
      "num_tokens": 44490603.0,
      "step": 19935
    },
    {
      "entropy": 5.463040685653686,
      "epoch": 2.239456393553097,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004505338162848382,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22412658035755156,
      "num_tokens": 44501527.0,
      "step": 19940
    },
    {
      "entropy": 5.580593299865723,
      "epoch": 2.2400179704610546,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004505086967676395,
      "loss": 5.1512,
      "mean_token_accuracy": 0.21155949532985688,
      "num_tokens": 44512619.0,
      "step": 19945
    },
    {
      "entropy": 5.623563814163208,
      "epoch": 2.240579547369012,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004504835716621757,
      "loss": 5.163,
      "mean_token_accuracy": 0.20709873884916305,
      "num_tokens": 44524639.0,
      "step": 19950
    },
    {
      "entropy": 5.5480992794036865,
      "epoch": 2.24114112427697,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004504584409692471,
      "loss": 5.0225,
      "mean_token_accuracy": 0.2204157218337059,
      "num_tokens": 44535013.0,
      "step": 19955
    },
    {
      "entropy": 5.6503355503082275,
      "epoch": 2.241702701184927,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045043330468965364,
      "loss": 5.1184,
      "mean_token_accuracy": 0.21190449148416518,
      "num_tokens": 44546838.0,
      "step": 19960
    },
    {
      "entropy": 5.5199073314666744,
      "epoch": 2.2422642780928848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045040816282419566,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22054285556077957,
      "num_tokens": 44557984.0,
      "step": 19965
    },
    {
      "entropy": 5.466346263885498,
      "epoch": 2.2428258550008424,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045038301537367374,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2222136825323105,
      "num_tokens": 44569029.0,
      "step": 19970
    },
    {
      "entropy": 5.506089448928833,
      "epoch": 2.2433874319088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045035786233888864,
      "loss": 5.0764,
      "mean_token_accuracy": 0.2161836951971054,
      "num_tokens": 44578804.0,
      "step": 19975
    },
    {
      "entropy": 5.563601732254028,
      "epoch": 2.2439490088167573,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045033270372064123,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21877280622720718,
      "num_tokens": 44589362.0,
      "step": 19980
    },
    {
      "entropy": 5.558225488662719,
      "epoch": 2.244510585724715,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045030753951973243,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2126692146062851,
      "num_tokens": 44600137.0,
      "step": 19985
    },
    {
      "entropy": 5.509299087524414,
      "epoch": 2.2450721626326726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045028236973696363,
      "loss": 5.0272,
      "mean_token_accuracy": 0.22364579141139984,
      "num_tokens": 44611707.0,
      "step": 19990
    },
    {
      "entropy": 5.537660694122314,
      "epoch": 2.2456337395406303,
      "grad_norm": 1.0,
      "learning_rate": 0.0004502571943731362,
      "loss": 5.0344,
      "mean_token_accuracy": 0.2163463979959488,
      "num_tokens": 44621141.0,
      "step": 19995
    },
    {
      "entropy": 5.551248931884766,
      "epoch": 2.2461953164485875,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004502320134290517,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21621730029582978,
      "num_tokens": 44632043.0,
      "step": 20000
    },
    {
      "entropy": 5.546774578094483,
      "epoch": 2.246756893356545,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045020682690551203,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21187424510717393,
      "num_tokens": 44643905.0,
      "step": 20005
    },
    {
      "entropy": 5.532029104232788,
      "epoch": 2.247318470264503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045018163480331887,
      "loss": 5.0382,
      "mean_token_accuracy": 0.20630647242069244,
      "num_tokens": 44655105.0,
      "step": 20010
    },
    {
      "entropy": 5.516870403289795,
      "epoch": 2.24788004717246,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045015643712327465,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2180454507470131,
      "num_tokens": 44667096.0,
      "step": 20015
    },
    {
      "entropy": 5.592005062103271,
      "epoch": 2.2484416240804177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004501312338661815,
      "loss": 5.0394,
      "mean_token_accuracy": 0.21560201197862625,
      "num_tokens": 44678206.0,
      "step": 20020
    },
    {
      "entropy": 5.504658174514771,
      "epoch": 2.2490032009883754,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004501060250328419,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21928033381700515,
      "num_tokens": 44688619.0,
      "step": 20025
    },
    {
      "entropy": 5.5198901176452635,
      "epoch": 2.249564777896333,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004500808106240586,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21135976016521454,
      "num_tokens": 44700757.0,
      "step": 20030
    },
    {
      "entropy": 5.427563095092774,
      "epoch": 2.2501263548042907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004500555906406343,
      "loss": 4.8952,
      "mean_token_accuracy": 0.225138384103775,
      "num_tokens": 44710822.0,
      "step": 20035
    },
    {
      "entropy": 5.4066441535949705,
      "epoch": 2.250687931712248,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004500303650833722,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22407299131155015,
      "num_tokens": 44722163.0,
      "step": 20040
    },
    {
      "entropy": 5.588882875442505,
      "epoch": 2.2512495086202056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004500051339530754,
      "loss": 5.0743,
      "mean_token_accuracy": 0.21150657385587693,
      "num_tokens": 44734271.0,
      "step": 20045
    },
    {
      "entropy": 5.623574876785279,
      "epoch": 2.2518110855281632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004499798972505472,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2150291845202446,
      "num_tokens": 44745345.0,
      "step": 20050
    },
    {
      "entropy": 5.481700658798218,
      "epoch": 2.2523726624361204,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044995465497659127,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21378520429134368,
      "num_tokens": 44756986.0,
      "step": 20055
    },
    {
      "entropy": 5.467776203155518,
      "epoch": 2.252934239344078,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004499294071320112,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22258745282888412,
      "num_tokens": 44767259.0,
      "step": 20060
    },
    {
      "entropy": 5.47691707611084,
      "epoch": 2.2534958162520358,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000449904153717611,
      "loss": 5.05,
      "mean_token_accuracy": 0.21129870265722275,
      "num_tokens": 44778884.0,
      "step": 20065
    },
    {
      "entropy": 5.526091575622559,
      "epoch": 2.2540573931599934,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004498788947341947,
      "loss": 4.9611,
      "mean_token_accuracy": 0.221601140499115,
      "num_tokens": 44789216.0,
      "step": 20070
    },
    {
      "entropy": 5.441502857208252,
      "epoch": 2.2546189700679506,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004498536301825665,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22332427501678467,
      "num_tokens": 44799357.0,
      "step": 20075
    },
    {
      "entropy": 5.477306222915649,
      "epoch": 2.2551805469759083,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000449828360063531,
      "loss": 5.0555,
      "mean_token_accuracy": 0.2129678398370743,
      "num_tokens": 44811040.0,
      "step": 20080
    },
    {
      "entropy": 5.574104928970337,
      "epoch": 2.255742123883866,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044980308437789267,
      "loss": 5.1153,
      "mean_token_accuracy": 0.20677785724401473,
      "num_tokens": 44822103.0,
      "step": 20085
    },
    {
      "entropy": 5.537933683395385,
      "epoch": 2.2563037007918236,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004497778031264564,
      "loss": 5.017,
      "mean_token_accuracy": 0.21280057579278946,
      "num_tokens": 44834712.0,
      "step": 20090
    },
    {
      "entropy": 5.5099669933319095,
      "epoch": 2.256865277699781,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044975251631002694,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2177702933549881,
      "num_tokens": 44845412.0,
      "step": 20095
    },
    {
      "entropy": 5.492259979248047,
      "epoch": 2.2574268546077385,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004497272239294097,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21486400365829467,
      "num_tokens": 44857720.0,
      "step": 20100
    },
    {
      "entropy": 5.526381921768189,
      "epoch": 2.257988431515696,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004497019259854099,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21998067051172257,
      "num_tokens": 44868253.0,
      "step": 20105
    },
    {
      "entropy": 5.540175342559815,
      "epoch": 2.2585500084236534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044967662247883295,
      "loss": 4.9949,
      "mean_token_accuracy": 0.2128733828663826,
      "num_tokens": 44878525.0,
      "step": 20110
    },
    {
      "entropy": 5.415162229537964,
      "epoch": 2.259111585331611,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044965131341048454,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22038732916116716,
      "num_tokens": 44889565.0,
      "step": 20115
    },
    {
      "entropy": 5.434309768676758,
      "epoch": 2.2596731622395687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004496259987811707,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2187286213040352,
      "num_tokens": 44901191.0,
      "step": 20120
    },
    {
      "entropy": 5.497772550582885,
      "epoch": 2.2602347391475264,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004496006785916973,
      "loss": 4.9301,
      "mean_token_accuracy": 0.21891519278287888,
      "num_tokens": 44911483.0,
      "step": 20125
    },
    {
      "entropy": 5.6200696468353275,
      "epoch": 2.260796316055484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044957535284287055,
      "loss": 5.1022,
      "mean_token_accuracy": 0.21530003398656844,
      "num_tokens": 44923537.0,
      "step": 20130
    },
    {
      "entropy": 5.568510961532593,
      "epoch": 2.2613578929634413,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044955002153549686,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2140822157263756,
      "num_tokens": 44935780.0,
      "step": 20135
    },
    {
      "entropy": 5.566320133209229,
      "epoch": 2.261919469871399,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004495246846703828,
      "loss": 5.0518,
      "mean_token_accuracy": 0.2156530052423477,
      "num_tokens": 44946879.0,
      "step": 20140
    },
    {
      "entropy": 5.566345834732056,
      "epoch": 2.2624810467793566,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044949934224833505,
      "loss": 5.0458,
      "mean_token_accuracy": 0.21842523217201232,
      "num_tokens": 44958385.0,
      "step": 20145
    },
    {
      "entropy": 5.548521614074707,
      "epoch": 2.263042623687314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004494739942701606,
      "loss": 5.1198,
      "mean_token_accuracy": 0.21350047141313552,
      "num_tokens": 44970597.0,
      "step": 20150
    },
    {
      "entropy": 5.520525503158569,
      "epoch": 2.2636042005952715,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044944864073666655,
      "loss": 5.0507,
      "mean_token_accuracy": 0.22157226949930192,
      "num_tokens": 44983123.0,
      "step": 20155
    },
    {
      "entropy": 5.5499523162841795,
      "epoch": 2.264165777503229,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044942328164866,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21604295670986176,
      "num_tokens": 44993378.0,
      "step": 20160
    },
    {
      "entropy": 5.49625883102417,
      "epoch": 2.2647273544111868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004493979170069486,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21094827800989152,
      "num_tokens": 45004905.0,
      "step": 20165
    },
    {
      "entropy": 5.494584465026856,
      "epoch": 2.265288931319144,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044937254681233984,
      "loss": 5.0203,
      "mean_token_accuracy": 0.21530242264270782,
      "num_tokens": 45016640.0,
      "step": 20170
    },
    {
      "entropy": 5.5954962253570555,
      "epoch": 2.2658505082271017,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044934717106564164,
      "loss": 5.1529,
      "mean_token_accuracy": 0.2048359125852585,
      "num_tokens": 45028285.0,
      "step": 20175
    },
    {
      "entropy": 5.58119044303894,
      "epoch": 2.2664120851350593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044932178976766183,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21442173719406127,
      "num_tokens": 45038924.0,
      "step": 20180
    },
    {
      "entropy": 5.5329478740692135,
      "epoch": 2.266973662043017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044929640291920865,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2222355931997299,
      "num_tokens": 45050483.0,
      "step": 20185
    },
    {
      "entropy": 5.593491268157959,
      "epoch": 2.267535238950974,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004492710105210904,
      "loss": 5.1788,
      "mean_token_accuracy": 0.20496722012758256,
      "num_tokens": 45063242.0,
      "step": 20190
    },
    {
      "entropy": 5.5104012966156,
      "epoch": 2.268096815858932,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044924561257411554,
      "loss": 5.0321,
      "mean_token_accuracy": 0.2103570967912674,
      "num_tokens": 45074057.0,
      "step": 20195
    },
    {
      "entropy": 5.527031517028808,
      "epoch": 2.2686583927668895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004492202090790928,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21220585852861404,
      "num_tokens": 45085488.0,
      "step": 20200
    },
    {
      "entropy": 5.608630704879761,
      "epoch": 2.2692199696748467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004491948000368311,
      "loss": 5.1257,
      "mean_token_accuracy": 0.21333254873752594,
      "num_tokens": 45096303.0,
      "step": 20205
    },
    {
      "entropy": 5.514669942855835,
      "epoch": 2.2697815465828044,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044916938544813936,
      "loss": 4.9096,
      "mean_token_accuracy": 0.22026229053735732,
      "num_tokens": 45106411.0,
      "step": 20210
    },
    {
      "entropy": 5.539509248733521,
      "epoch": 2.270343123490762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044914396531382686,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2146800473332405,
      "num_tokens": 45117355.0,
      "step": 20215
    },
    {
      "entropy": 5.4909828186035154,
      "epoch": 2.2709047003987197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044911853963470295,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22228717505931855,
      "num_tokens": 45127612.0,
      "step": 20220
    },
    {
      "entropy": 5.614795780181884,
      "epoch": 2.2714662773066774,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004490931084115772,
      "loss": 5.127,
      "mean_token_accuracy": 0.21472587585449218,
      "num_tokens": 45138495.0,
      "step": 20225
    },
    {
      "entropy": 5.566579532623291,
      "epoch": 2.2720278542146346,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044906767164525936,
      "loss": 4.9726,
      "mean_token_accuracy": 0.223576095700264,
      "num_tokens": 45149727.0,
      "step": 20230
    },
    {
      "entropy": 5.5417181015014645,
      "epoch": 2.2725894311225923,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004490422293365594,
      "loss": 5.0437,
      "mean_token_accuracy": 0.21483254581689834,
      "num_tokens": 45160921.0,
      "step": 20235
    },
    {
      "entropy": 5.447387075424194,
      "epoch": 2.27315100803055,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004490167814862872,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22045573145151137,
      "num_tokens": 45172607.0,
      "step": 20240
    },
    {
      "entropy": 5.603069305419922,
      "epoch": 2.273712584938507,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004489913280952533,
      "loss": 5.1374,
      "mean_token_accuracy": 0.2154537096619606,
      "num_tokens": 45186330.0,
      "step": 20245
    },
    {
      "entropy": 5.426331520080566,
      "epoch": 2.274274161846465,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004489658691642681,
      "loss": 4.89,
      "mean_token_accuracy": 0.2276003196835518,
      "num_tokens": 45198239.0,
      "step": 20250
    },
    {
      "entropy": 5.430828857421875,
      "epoch": 2.2748357387544225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044894040469414207,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22258893847465516,
      "num_tokens": 45209295.0,
      "step": 20255
    },
    {
      "entropy": 5.541683959960937,
      "epoch": 2.27539731566238,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044891493468568607,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21564511954784393,
      "num_tokens": 45220210.0,
      "step": 20260
    },
    {
      "entropy": 5.556733751296997,
      "epoch": 2.2759588925703373,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044888945913971106,
      "loss": 5.05,
      "mean_token_accuracy": 0.2158108815550804,
      "num_tokens": 45231809.0,
      "step": 20265
    },
    {
      "entropy": 5.552735328674316,
      "epoch": 2.276520469478295,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044886397805702835,
      "loss": 5.0284,
      "mean_token_accuracy": 0.22377929836511612,
      "num_tokens": 45242847.0,
      "step": 20270
    },
    {
      "entropy": 5.485146713256836,
      "epoch": 2.2770820463862527,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044883849143844914,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22531650066375733,
      "num_tokens": 45253357.0,
      "step": 20275
    },
    {
      "entropy": 5.444186067581176,
      "epoch": 2.2776436232942103,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004488129992847848,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22022038400173188,
      "num_tokens": 45264607.0,
      "step": 20280
    },
    {
      "entropy": 5.578175592422485,
      "epoch": 2.2782052002021675,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004487875015968473,
      "loss": 5.0539,
      "mean_token_accuracy": 0.22028457075357438,
      "num_tokens": 45275211.0,
      "step": 20285
    },
    {
      "entropy": 5.480272912979126,
      "epoch": 2.278766777110125,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004487619983754483,
      "loss": 4.9251,
      "mean_token_accuracy": 0.23015227317810058,
      "num_tokens": 45286665.0,
      "step": 20290
    },
    {
      "entropy": 5.481760406494141,
      "epoch": 2.279328354018083,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044873648962139983,
      "loss": 5.0647,
      "mean_token_accuracy": 0.2142808198928833,
      "num_tokens": 45298066.0,
      "step": 20295
    },
    {
      "entropy": 5.502311706542969,
      "epoch": 2.27988993092604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004487109753355142,
      "loss": 5.0238,
      "mean_token_accuracy": 0.2139996349811554,
      "num_tokens": 45308799.0,
      "step": 20300
    },
    {
      "entropy": 5.501436710357666,
      "epoch": 2.2804515078339977,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004486854555186037,
      "loss": 4.8958,
      "mean_token_accuracy": 0.2249002143740654,
      "num_tokens": 45319932.0,
      "step": 20305
    },
    {
      "entropy": 5.565549612045288,
      "epoch": 2.2810130847419554,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004486599301714809,
      "loss": 5.0509,
      "mean_token_accuracy": 0.2173139289021492,
      "num_tokens": 45330008.0,
      "step": 20310
    },
    {
      "entropy": 5.547131776809692,
      "epoch": 2.281574661649913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004486343992949587,
      "loss": 5.0631,
      "mean_token_accuracy": 0.2076545238494873,
      "num_tokens": 45340877.0,
      "step": 20315
    },
    {
      "entropy": 5.551836919784546,
      "epoch": 2.2821362385578707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044860886288984977,
      "loss": 5.012,
      "mean_token_accuracy": 0.21866750717163086,
      "num_tokens": 45351956.0,
      "step": 20320
    },
    {
      "entropy": 5.578943872451783,
      "epoch": 2.282697815465828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004485833209569674,
      "loss": 5.106,
      "mean_token_accuracy": 0.20931802242994307,
      "num_tokens": 45363133.0,
      "step": 20325
    },
    {
      "entropy": 5.515356349945068,
      "epoch": 2.2832593923737856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004485577734971247,
      "loss": 4.9626,
      "mean_token_accuracy": 0.21819316893815993,
      "num_tokens": 45373917.0,
      "step": 20330
    },
    {
      "entropy": 5.5602789402008055,
      "epoch": 2.2838209692817433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044853222051113525,
      "loss": 5.07,
      "mean_token_accuracy": 0.21720924973487854,
      "num_tokens": 45385575.0,
      "step": 20335
    },
    {
      "entropy": 5.537344694137573,
      "epoch": 2.2843825461897005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044850666199981244,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2195739209651947,
      "num_tokens": 45396615.0,
      "step": 20340
    },
    {
      "entropy": 5.53484697341919,
      "epoch": 2.284944123097658,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044848109796397035,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21506604850292205,
      "num_tokens": 45408931.0,
      "step": 20345
    },
    {
      "entropy": 5.595519781112671,
      "epoch": 2.285505700005616,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044845552840442276,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21388394832611085,
      "num_tokens": 45419993.0,
      "step": 20350
    },
    {
      "entropy": 5.547989559173584,
      "epoch": 2.2860672769135735,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044842995332198393,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20773522853851317,
      "num_tokens": 45431889.0,
      "step": 20355
    },
    {
      "entropy": 5.4633935451507565,
      "epoch": 2.2866288538215307,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044840437271746807,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21630053520202636,
      "num_tokens": 45444233.0,
      "step": 20360
    },
    {
      "entropy": 5.500809812545777,
      "epoch": 2.2871904307294884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004483787865916898,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21912205666303636,
      "num_tokens": 45455743.0,
      "step": 20365
    },
    {
      "entropy": 5.4943009376525875,
      "epoch": 2.287752007637446,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044835319494546366,
      "loss": 4.931,
      "mean_token_accuracy": 0.21772704273462296,
      "num_tokens": 45466272.0,
      "step": 20370
    },
    {
      "entropy": 5.564087772369385,
      "epoch": 2.2883135845454037,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044832759777960456,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21926030218601228,
      "num_tokens": 45477828.0,
      "step": 20375
    },
    {
      "entropy": 5.5381693840026855,
      "epoch": 2.288875161453361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004483019950949276,
      "loss": 4.9995,
      "mean_token_accuracy": 0.21388251334428787,
      "num_tokens": 45489315.0,
      "step": 20380
    },
    {
      "entropy": 5.441487073898315,
      "epoch": 2.2894367383613186,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004482763868922478,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22285035252571106,
      "num_tokens": 45499836.0,
      "step": 20385
    },
    {
      "entropy": 5.532894277572632,
      "epoch": 2.289998315269276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004482507731723807,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21364971697330476,
      "num_tokens": 45511106.0,
      "step": 20390
    },
    {
      "entropy": 5.542057037353516,
      "epoch": 2.2905598921772334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004482251539361418,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21543824225664138,
      "num_tokens": 45521314.0,
      "step": 20395
    },
    {
      "entropy": 5.583531475067138,
      "epoch": 2.291121469085191,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004481995291843467,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21962677389383317,
      "num_tokens": 45531675.0,
      "step": 20400
    },
    {
      "entropy": 5.555460739135742,
      "epoch": 2.2916830459931488,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481738989178115,
      "loss": 5.0699,
      "mean_token_accuracy": 0.21404756754636764,
      "num_tokens": 45542064.0,
      "step": 20405
    },
    {
      "entropy": 5.522829246520996,
      "epoch": 2.2922446229011064,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004481482631373522,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2153431609272957,
      "num_tokens": 45553764.0,
      "step": 20410
    },
    {
      "entropy": 5.621661758422851,
      "epoch": 2.292806199809064,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044812262184378503,
      "loss": 5.0759,
      "mean_token_accuracy": 0.21525806188583374,
      "num_tokens": 45565302.0,
      "step": 20415
    },
    {
      "entropy": 5.457597875595093,
      "epoch": 2.2933677767170213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044809697503792647,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22829592376947402,
      "num_tokens": 45577165.0,
      "step": 20420
    },
    {
      "entropy": 5.505373811721801,
      "epoch": 2.293929353624979,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044807132272059306,
      "loss": 4.979,
      "mean_token_accuracy": 0.22215070575475693,
      "num_tokens": 45587771.0,
      "step": 20425
    },
    {
      "entropy": 5.496624279022217,
      "epoch": 2.2944909305329366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004480456648926016,
      "loss": 4.998,
      "mean_token_accuracy": 0.22650871127843858,
      "num_tokens": 45598815.0,
      "step": 20430
    },
    {
      "entropy": 5.497972726821899,
      "epoch": 2.295052507440894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004480200015547691,
      "loss": 4.9786,
      "mean_token_accuracy": 0.2154836744070053,
      "num_tokens": 45609007.0,
      "step": 20435
    },
    {
      "entropy": 5.537816047668457,
      "epoch": 2.2956140843488515,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044799433270791255,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21754280179739,
      "num_tokens": 45621323.0,
      "step": 20440
    },
    {
      "entropy": 5.433668947219848,
      "epoch": 2.296175661256809,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004479686583528494,
      "loss": 4.8565,
      "mean_token_accuracy": 0.22207009345293044,
      "num_tokens": 45631895.0,
      "step": 20445
    },
    {
      "entropy": 5.535416841506958,
      "epoch": 2.296737238164767,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004479429784903971,
      "loss": 5.0232,
      "mean_token_accuracy": 0.2190642848610878,
      "num_tokens": 45643204.0,
      "step": 20450
    },
    {
      "entropy": 5.58393611907959,
      "epoch": 2.297298815072724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004479172931213733,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2178732767701149,
      "num_tokens": 45654607.0,
      "step": 20455
    },
    {
      "entropy": 5.4296633243560795,
      "epoch": 2.2978603919806817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004478916022465958,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21905323266983032,
      "num_tokens": 45666746.0,
      "step": 20460
    },
    {
      "entropy": 5.499394035339355,
      "epoch": 2.2984219688886394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004478659058668827,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2198905184864998,
      "num_tokens": 45676663.0,
      "step": 20465
    },
    {
      "entropy": 5.49765567779541,
      "epoch": 2.298983545796597,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044784020398305204,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21482738852500916,
      "num_tokens": 45688746.0,
      "step": 20470
    },
    {
      "entropy": 5.6275248527526855,
      "epoch": 2.2995451227045542,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004478144965959222,
      "loss": 5.1369,
      "mean_token_accuracy": 0.2159050449728966,
      "num_tokens": 45700199.0,
      "step": 20475
    },
    {
      "entropy": 5.536463403701783,
      "epoch": 2.300106699612512,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044778878370631183,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21027853041887284,
      "num_tokens": 45710771.0,
      "step": 20480
    },
    {
      "entropy": 5.565132904052734,
      "epoch": 2.3006682765204696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044776306531503957,
      "loss": 5.1629,
      "mean_token_accuracy": 0.21185825020074844,
      "num_tokens": 45721957.0,
      "step": 20485
    },
    {
      "entropy": 5.546128129959106,
      "epoch": 2.301229853428427,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044773734142292425,
      "loss": 5.0313,
      "mean_token_accuracy": 0.21309620141983032,
      "num_tokens": 45733904.0,
      "step": 20490
    },
    {
      "entropy": 5.607325220108033,
      "epoch": 2.3017914303363844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004477116120307851,
      "loss": 5.0279,
      "mean_token_accuracy": 0.21692472249269484,
      "num_tokens": 45744004.0,
      "step": 20495
    },
    {
      "entropy": 5.541652584075928,
      "epoch": 2.302353007244342,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004476858771394412,
      "loss": 4.9478,
      "mean_token_accuracy": 0.23010701090097427,
      "num_tokens": 45755130.0,
      "step": 20500
    },
    {
      "entropy": 5.524637842178345,
      "epoch": 2.3029145841522998,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004476601367497119,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20721037536859513,
      "num_tokens": 45767353.0,
      "step": 20505
    },
    {
      "entropy": 5.578741788864136,
      "epoch": 2.3034761610602574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000447634390862417,
      "loss": 5.1895,
      "mean_token_accuracy": 0.1998126119375229,
      "num_tokens": 45779266.0,
      "step": 20510
    },
    {
      "entropy": 5.6353506565094,
      "epoch": 2.3040377379682146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044760863947837607,
      "loss": 5.1154,
      "mean_token_accuracy": 0.21289069801568986,
      "num_tokens": 45790601.0,
      "step": 20515
    },
    {
      "entropy": 5.537068700790405,
      "epoch": 2.3045993148761723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044758288259840904,
      "loss": 4.9998,
      "mean_token_accuracy": 0.21581446826457978,
      "num_tokens": 45801449.0,
      "step": 20520
    },
    {
      "entropy": 5.440796613693237,
      "epoch": 2.30516089178413,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004475571202233363,
      "loss": 4.9479,
      "mean_token_accuracy": 0.21947215646505355,
      "num_tokens": 45811737.0,
      "step": 20525
    },
    {
      "entropy": 5.532358646392822,
      "epoch": 2.305722468692087,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004475313523539778,
      "loss": 5.0034,
      "mean_token_accuracy": 0.22088104784488677,
      "num_tokens": 45822955.0,
      "step": 20530
    },
    {
      "entropy": 5.53133864402771,
      "epoch": 2.306284045600045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044750557899115407,
      "loss": 5.0799,
      "mean_token_accuracy": 0.22381386756896973,
      "num_tokens": 45834697.0,
      "step": 20535
    },
    {
      "entropy": 5.476288843154907,
      "epoch": 2.3068456225080025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004474798001356859,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21661789566278458,
      "num_tokens": 45846869.0,
      "step": 20540
    },
    {
      "entropy": 5.47454252243042,
      "epoch": 2.30740719941596,
      "grad_norm": 1.03125,
      "learning_rate": 0.000447454015788394,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2198227360844612,
      "num_tokens": 45858171.0,
      "step": 20545
    },
    {
      "entropy": 5.523162937164306,
      "epoch": 2.3079687763239174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004474282259500992,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2129409983754158,
      "num_tokens": 45868715.0,
      "step": 20550
    },
    {
      "entropy": 5.545602226257325,
      "epoch": 2.308530353231875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000447402430621623,
      "loss": 4.9858,
      "mean_token_accuracy": 0.22567598670721054,
      "num_tokens": 45880569.0,
      "step": 20555
    },
    {
      "entropy": 5.541715049743653,
      "epoch": 2.3090919301398327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004473766298037865,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21914239674806596,
      "num_tokens": 45890452.0,
      "step": 20560
    },
    {
      "entropy": 5.592472410202026,
      "epoch": 2.3096535070477904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044735082349741116,
      "loss": 5.0968,
      "mean_token_accuracy": 0.213210129737854,
      "num_tokens": 45903125.0,
      "step": 20565
    },
    {
      "entropy": 5.487708902359008,
      "epoch": 2.3102150839557476,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044732501170331886,
      "loss": 4.9632,
      "mean_token_accuracy": 0.22159618884325027,
      "num_tokens": 45914961.0,
      "step": 20570
    },
    {
      "entropy": 5.5966212272644045,
      "epoch": 2.3107766608637053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004472991944223313,
      "loss": 5.1335,
      "mean_token_accuracy": 0.20942577719688416,
      "num_tokens": 45927198.0,
      "step": 20575
    },
    {
      "entropy": 5.562869834899902,
      "epoch": 2.311338237771663,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044727337165527056,
      "loss": 5.0911,
      "mean_token_accuracy": 0.2143895223736763,
      "num_tokens": 45937937.0,
      "step": 20580
    },
    {
      "entropy": 5.617971563339234,
      "epoch": 2.31189981467962,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044724754340295887,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20797871500253678,
      "num_tokens": 45948983.0,
      "step": 20585
    },
    {
      "entropy": 5.520430374145508,
      "epoch": 2.312461391587578,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004472217096662186,
      "loss": 5.0753,
      "mean_token_accuracy": 0.2134031668305397,
      "num_tokens": 45960674.0,
      "step": 20590
    },
    {
      "entropy": 5.5350316047668455,
      "epoch": 2.3130229684955355,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004471958704458723,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21467837542295456,
      "num_tokens": 45971260.0,
      "step": 20595
    },
    {
      "entropy": 5.624927568435669,
      "epoch": 2.313584545403493,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044717002574274267,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22433886528015137,
      "num_tokens": 45983942.0,
      "step": 20600
    },
    {
      "entropy": 5.528057956695557,
      "epoch": 2.314146122311451,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004471441755576527,
      "loss": 5.0681,
      "mean_token_accuracy": 0.20806693881750107,
      "num_tokens": 45995983.0,
      "step": 20605
    },
    {
      "entropy": 5.543956327438354,
      "epoch": 2.314707699219408,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044711831989142534,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2166326329112053,
      "num_tokens": 46007074.0,
      "step": 20610
    },
    {
      "entropy": 5.6087061882019045,
      "epoch": 2.3152692761273657,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044709245874488405,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21721668243408204,
      "num_tokens": 46018376.0,
      "step": 20615
    },
    {
      "entropy": 5.526835107803345,
      "epoch": 2.3158308530353233,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004470665921188519,
      "loss": 4.9915,
      "mean_token_accuracy": 0.217928509414196,
      "num_tokens": 46029192.0,
      "step": 20620
    },
    {
      "entropy": 5.502330112457275,
      "epoch": 2.3163924299432805,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044704072001415285,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21590762138366698,
      "num_tokens": 46040735.0,
      "step": 20625
    },
    {
      "entropy": 5.484842109680176,
      "epoch": 2.316954006851238,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044701484243161055,
      "loss": 4.9189,
      "mean_token_accuracy": 0.2250294104218483,
      "num_tokens": 46050859.0,
      "step": 20630
    },
    {
      "entropy": 5.443659448623658,
      "epoch": 2.317515583759196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000446988959372049,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2189323663711548,
      "num_tokens": 46060704.0,
      "step": 20635
    },
    {
      "entropy": 5.464203882217407,
      "epoch": 2.3180771606671535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004469630708362922,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21272530555725097,
      "num_tokens": 46072317.0,
      "step": 20640
    },
    {
      "entropy": 5.59088659286499,
      "epoch": 2.3186387375751107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004469371768251646,
      "loss": 5.1212,
      "mean_token_accuracy": 0.21275742799043657,
      "num_tokens": 46083823.0,
      "step": 20645
    },
    {
      "entropy": 5.627917003631592,
      "epoch": 2.3192003144830684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004469112773394906,
      "loss": 5.0757,
      "mean_token_accuracy": 0.21398030519485473,
      "num_tokens": 46093920.0,
      "step": 20650
    },
    {
      "entropy": 5.49130277633667,
      "epoch": 2.319761891391026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004468853723800948,
      "loss": 4.9974,
      "mean_token_accuracy": 0.22291107773780822,
      "num_tokens": 46104618.0,
      "step": 20655
    },
    {
      "entropy": 5.487282657623291,
      "epoch": 2.3203234682989837,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004468594619478021,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2188535898923874,
      "num_tokens": 46114748.0,
      "step": 20660
    },
    {
      "entropy": 5.539809608459473,
      "epoch": 2.320885045206941,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468335460434375,
      "loss": 5.0588,
      "mean_token_accuracy": 0.2204669326543808,
      "num_tokens": 46125125.0,
      "step": 20665
    },
    {
      "entropy": 5.437728023529052,
      "epoch": 2.3214466221148986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004468076246678261,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21401802003383635,
      "num_tokens": 46135877.0,
      "step": 20670
    },
    {
      "entropy": 5.57085075378418,
      "epoch": 2.3220081990228563,
      "grad_norm": 1.0,
      "learning_rate": 0.00044678169782179334,
      "loss": 5.1457,
      "mean_token_accuracy": 0.20827373415231704,
      "num_tokens": 46146769.0,
      "step": 20675
    },
    {
      "entropy": 5.514590120315551,
      "epoch": 2.3225697759308135,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044675576550616466,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2178739294409752,
      "num_tokens": 46158408.0,
      "step": 20680
    },
    {
      "entropy": 5.567808008193969,
      "epoch": 2.323131352838771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044672982772176584,
      "loss": 5.126,
      "mean_token_accuracy": 0.2080349236726761,
      "num_tokens": 46169961.0,
      "step": 20685
    },
    {
      "entropy": 5.598749399185181,
      "epoch": 2.323692929746729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004467038844694227,
      "loss": 5.0834,
      "mean_token_accuracy": 0.21343600749969482,
      "num_tokens": 46181750.0,
      "step": 20690
    },
    {
      "entropy": 5.554172515869141,
      "epoch": 2.3242545066546865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004466779357499612,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2210423991084099,
      "num_tokens": 46192745.0,
      "step": 20695
    },
    {
      "entropy": 5.537807416915894,
      "epoch": 2.324816083562644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044665198156420774,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21833985149860383,
      "num_tokens": 46203267.0,
      "step": 20700
    },
    {
      "entropy": 5.502267599105835,
      "epoch": 2.3253776604706013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044662602191298856,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21790873557329177,
      "num_tokens": 46214966.0,
      "step": 20705
    },
    {
      "entropy": 5.507996273040772,
      "epoch": 2.325939237378559,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044660005679713027,
      "loss": 4.9702,
      "mean_token_accuracy": 0.22030646204948426,
      "num_tokens": 46225755.0,
      "step": 20710
    },
    {
      "entropy": 5.550733947753907,
      "epoch": 2.3265008142865167,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044657408621745964,
      "loss": 5.0399,
      "mean_token_accuracy": 0.21778851449489595,
      "num_tokens": 46237091.0,
      "step": 20715
    },
    {
      "entropy": 5.5589512348175045,
      "epoch": 2.327062391194474,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004465481101748036,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2133980244398117,
      "num_tokens": 46249250.0,
      "step": 20720
    },
    {
      "entropy": 5.53260064125061,
      "epoch": 2.3276239681024316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044652212866998914,
      "loss": 4.9762,
      "mean_token_accuracy": 0.21771362274885178,
      "num_tokens": 46259866.0,
      "step": 20725
    },
    {
      "entropy": 5.488447284698486,
      "epoch": 2.328185545010389,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044649614170384354,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22697565853595733,
      "num_tokens": 46270888.0,
      "step": 20730
    },
    {
      "entropy": 5.512794351577758,
      "epoch": 2.328747121918347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044647014927719436,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2127450093626976,
      "num_tokens": 46281109.0,
      "step": 20735
    },
    {
      "entropy": 5.519795846939087,
      "epoch": 2.329308698826304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044644415139086913,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2285095661878586,
      "num_tokens": 46292171.0,
      "step": 20740
    },
    {
      "entropy": 5.52105746269226,
      "epoch": 2.3298702757342618,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004464181480456955,
      "loss": 4.9768,
      "mean_token_accuracy": 0.21827558279037476,
      "num_tokens": 46302919.0,
      "step": 20745
    },
    {
      "entropy": 5.428065061569214,
      "epoch": 2.3304318526422194,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004463921392425017,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22453514188528062,
      "num_tokens": 46313199.0,
      "step": 20750
    },
    {
      "entropy": 5.598998975753784,
      "epoch": 2.330993429550177,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044636612498211566,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2141270235180855,
      "num_tokens": 46324316.0,
      "step": 20755
    },
    {
      "entropy": 5.510714435577393,
      "epoch": 2.3315550064581343,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044634010526536575,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21354566365480424,
      "num_tokens": 46335767.0,
      "step": 20760
    },
    {
      "entropy": 5.46871657371521,
      "epoch": 2.332116583366092,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004463140800930804,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21326034367084504,
      "num_tokens": 46347840.0,
      "step": 20765
    },
    {
      "entropy": 5.532648086547852,
      "epoch": 2.3326781602740496,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004462880494660883,
      "loss": 5.1086,
      "mean_token_accuracy": 0.20897155702114106,
      "num_tokens": 46359766.0,
      "step": 20770
    },
    {
      "entropy": 5.570405960083008,
      "epoch": 2.333239737182007,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044626201338521834,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22979419380426408,
      "num_tokens": 46371426.0,
      "step": 20775
    },
    {
      "entropy": 5.541094827651977,
      "epoch": 2.3338013140899645,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004462359718512993,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21225306391716003,
      "num_tokens": 46382257.0,
      "step": 20780
    },
    {
      "entropy": 5.555874681472778,
      "epoch": 2.334362890997922,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044620992486516073,
      "loss": 5.0307,
      "mean_token_accuracy": 0.2126808762550354,
      "num_tokens": 46393689.0,
      "step": 20785
    },
    {
      "entropy": 5.570913362503052,
      "epoch": 2.33492446790588,
      "grad_norm": 1.125,
      "learning_rate": 0.0004461838724276316,
      "loss": 5.0714,
      "mean_token_accuracy": 0.21686942428350447,
      "num_tokens": 46404856.0,
      "step": 20790
    },
    {
      "entropy": 5.635024976730347,
      "epoch": 2.3354860448138375,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044615781453954164,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20711927115917206,
      "num_tokens": 46415395.0,
      "step": 20795
    },
    {
      "entropy": 5.621561050415039,
      "epoch": 2.3360476217217947,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044613175120172035,
      "loss": 5.0787,
      "mean_token_accuracy": 0.2147354081273079,
      "num_tokens": 46426773.0,
      "step": 20800
    },
    {
      "entropy": 5.470870447158814,
      "epoch": 2.3366091986297524,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004461056824149979,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22235175520181655,
      "num_tokens": 46438206.0,
      "step": 20805
    },
    {
      "entropy": 5.523656940460205,
      "epoch": 2.33717077553771,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000446079608180204,
      "loss": 5.047,
      "mean_token_accuracy": 0.21878981292247773,
      "num_tokens": 46449922.0,
      "step": 20810
    },
    {
      "entropy": 5.491311359405517,
      "epoch": 2.3377323524456672,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004460535284981692,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2231728881597519,
      "num_tokens": 46461715.0,
      "step": 20815
    },
    {
      "entropy": 5.4939055919647215,
      "epoch": 2.338293929353625,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004460274433697235,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21216827929019927,
      "num_tokens": 46473136.0,
      "step": 20820
    },
    {
      "entropy": 5.469852018356323,
      "epoch": 2.3388555062615826,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044600135279569784,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22352322638034822,
      "num_tokens": 46483644.0,
      "step": 20825
    },
    {
      "entropy": 5.581731176376342,
      "epoch": 2.3394170831695402,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044597525677692273,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2184366136789322,
      "num_tokens": 46494896.0,
      "step": 20830
    },
    {
      "entropy": 5.602496242523193,
      "epoch": 2.3399786600774974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004459491553142291,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21489936262369155,
      "num_tokens": 46505415.0,
      "step": 20835
    },
    {
      "entropy": 5.5247228145599365,
      "epoch": 2.340540236985455,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044592304840844813,
      "loss": 5.0306,
      "mean_token_accuracy": 0.21800798028707505,
      "num_tokens": 46516280.0,
      "step": 20840
    },
    {
      "entropy": 5.443897438049317,
      "epoch": 2.3411018138934128,
      "grad_norm": 1.03125,
      "learning_rate": 0.000445896936060411,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22214221954345703,
      "num_tokens": 46526456.0,
      "step": 20845
    },
    {
      "entropy": 5.5771317958831785,
      "epoch": 2.3416633908013704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044587081827094906,
      "loss": 5.0357,
      "mean_token_accuracy": 0.217498542368412,
      "num_tokens": 46537547.0,
      "step": 20850
    },
    {
      "entropy": 5.461738729476929,
      "epoch": 2.3422249677093276,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044584469504089407,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22364979684352876,
      "num_tokens": 46549424.0,
      "step": 20855
    },
    {
      "entropy": 5.521225166320801,
      "epoch": 2.3427865446172853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004458185663710777,
      "loss": 4.972,
      "mean_token_accuracy": 0.21903700828552247,
      "num_tokens": 46560234.0,
      "step": 20860
    },
    {
      "entropy": 5.531472778320312,
      "epoch": 2.343348121525243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004457924322623319,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2161193311214447,
      "num_tokens": 46570847.0,
      "step": 20865
    },
    {
      "entropy": 5.452326679229737,
      "epoch": 2.3439096984332,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044576629271548884,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2191176414489746,
      "num_tokens": 46582379.0,
      "step": 20870
    },
    {
      "entropy": 5.506544303894043,
      "epoch": 2.344471275341158,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004457401477313808,
      "loss": 5.008,
      "mean_token_accuracy": 0.21314579695463182,
      "num_tokens": 46594669.0,
      "step": 20875
    },
    {
      "entropy": 5.433322191238403,
      "epoch": 2.3450328522491155,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004457139973108403,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2272897869348526,
      "num_tokens": 46605834.0,
      "step": 20880
    },
    {
      "entropy": 5.513880062103271,
      "epoch": 2.345594429157073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004456878414546998,
      "loss": 5.0054,
      "mean_token_accuracy": 0.22167778313159942,
      "num_tokens": 46616684.0,
      "step": 20885
    },
    {
      "entropy": 5.523144149780274,
      "epoch": 2.346156006065031,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044566168016379234,
      "loss": 5.037,
      "mean_token_accuracy": 0.2205515906214714,
      "num_tokens": 46628045.0,
      "step": 20890
    },
    {
      "entropy": 5.49727873802185,
      "epoch": 2.346717582972988,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004456355134389507,
      "loss": 5.0074,
      "mean_token_accuracy": 0.21491773575544357,
      "num_tokens": 46639455.0,
      "step": 20895
    },
    {
      "entropy": 5.639538383483886,
      "epoch": 2.3472791598809457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044560934128100814,
      "loss": 5.2038,
      "mean_token_accuracy": 0.2077769637107849,
      "num_tokens": 46651025.0,
      "step": 20900
    },
    {
      "entropy": 5.589868640899658,
      "epoch": 2.3478407367889034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044558316369079805,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21430882811546326,
      "num_tokens": 46661859.0,
      "step": 20905
    },
    {
      "entropy": 5.57333197593689,
      "epoch": 2.3484023136968606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044555698066915385,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21654639691114425,
      "num_tokens": 46673441.0,
      "step": 20910
    },
    {
      "entropy": 5.584176397323608,
      "epoch": 2.3489638906048182,
      "grad_norm": 1.0,
      "learning_rate": 0.0004455307922169092,
      "loss": 5.0659,
      "mean_token_accuracy": 0.21244944483041764,
      "num_tokens": 46684959.0,
      "step": 20915
    },
    {
      "entropy": 5.559808158874512,
      "epoch": 2.349525467512776,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004455045983348979,
      "loss": 5.1045,
      "mean_token_accuracy": 0.22493693828582764,
      "num_tokens": 46694940.0,
      "step": 20920
    },
    {
      "entropy": 5.5084264278411865,
      "epoch": 2.3500870444207336,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004454783990239542,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22067876011133195,
      "num_tokens": 46706184.0,
      "step": 20925
    },
    {
      "entropy": 5.5443846702575685,
      "epoch": 2.350648621328691,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044545219428491215,
      "loss": 5.0053,
      "mean_token_accuracy": 0.2253313645720482,
      "num_tokens": 46718267.0,
      "step": 20930
    },
    {
      "entropy": 5.520178318023682,
      "epoch": 2.3512101982366485,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044542598411860605,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21547778993844985,
      "num_tokens": 46729293.0,
      "step": 20935
    },
    {
      "entropy": 5.495725631713867,
      "epoch": 2.351771775144606,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004453997685258706,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21692253798246383,
      "num_tokens": 46740214.0,
      "step": 20940
    },
    {
      "entropy": 5.580843353271485,
      "epoch": 2.3523333520525638,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044537354750754033,
      "loss": 5.0847,
      "mean_token_accuracy": 0.21256499290466307,
      "num_tokens": 46751269.0,
      "step": 20945
    },
    {
      "entropy": 5.495484685897827,
      "epoch": 2.352894928960521,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004453473210644503,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23038448095321656,
      "num_tokens": 46761851.0,
      "step": 20950
    },
    {
      "entropy": 5.478664588928223,
      "epoch": 2.3534565058684787,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004453210891974355,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21962607353925706,
      "num_tokens": 46773441.0,
      "step": 20955
    },
    {
      "entropy": 5.432811880111695,
      "epoch": 2.3540180827764363,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044529485190733113,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21836940795183182,
      "num_tokens": 46785016.0,
      "step": 20960
    },
    {
      "entropy": 5.551907348632812,
      "epoch": 2.3545796596843935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004452686091949727,
      "loss": 5.0248,
      "mean_token_accuracy": 0.21214471459388734,
      "num_tokens": 46796059.0,
      "step": 20965
    },
    {
      "entropy": 5.496445226669311,
      "epoch": 2.355141236592351,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044524236106119563,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22076206803321838,
      "num_tokens": 46805725.0,
      "step": 20970
    },
    {
      "entropy": 5.534666919708252,
      "epoch": 2.355702813500309,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044521610750683586,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22328146994113923,
      "num_tokens": 46816300.0,
      "step": 20975
    },
    {
      "entropy": 5.479611015319824,
      "epoch": 2.3562643904082665,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004451898485327291,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21872564703226088,
      "num_tokens": 46827498.0,
      "step": 20980
    },
    {
      "entropy": 5.521636533737182,
      "epoch": 2.356825967316224,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004451635841397116,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21629632264375687,
      "num_tokens": 46838416.0,
      "step": 20985
    },
    {
      "entropy": 5.556480884552002,
      "epoch": 2.3573875442241814,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004451373143286195,
      "loss": 5.086,
      "mean_token_accuracy": 0.21370928585529328,
      "num_tokens": 46849371.0,
      "step": 20990
    },
    {
      "entropy": 5.4913653373718265,
      "epoch": 2.357949121132139,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044511103910028945,
      "loss": 5.0106,
      "mean_token_accuracy": 0.21929299533367158,
      "num_tokens": 46860095.0,
      "step": 20995
    },
    {
      "entropy": 5.60245099067688,
      "epoch": 2.3585106980400967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004450847584555579,
      "loss": 5.1015,
      "mean_token_accuracy": 0.20974529385566712,
      "num_tokens": 46871217.0,
      "step": 21000
    },
    {
      "epoch": 2.3585106980400967,
      "eval_entropy": 5.375346313286664,
      "eval_loss": 5.1246514320373535,
      "eval_mean_token_accuracy": 0.21969848376383386,
      "eval_num_tokens": 46871217.0,
      "eval_runtime": 18.5299,
      "eval_samples_per_second": 1673.676,
      "eval_steps_per_second": 209.23,
      "step": 21000
    },
    {
      "entropy": 5.54589319229126,
      "epoch": 2.359072274948054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004450584723952617,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2173622339963913,
      "num_tokens": 46881224.0,
      "step": 21005
    },
    {
      "entropy": 5.50583119392395,
      "epoch": 2.3596338518560116,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004450321809202377,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21326765269041062,
      "num_tokens": 46892673.0,
      "step": 21010
    },
    {
      "entropy": 5.671557331085205,
      "epoch": 2.3601954287639693,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004450058840313232,
      "loss": 5.1074,
      "mean_token_accuracy": 0.21133753508329392,
      "num_tokens": 46902753.0,
      "step": 21015
    },
    {
      "entropy": 5.463589334487915,
      "epoch": 2.360757005671927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044497958172935536,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22649319916963578,
      "num_tokens": 46914093.0,
      "step": 21020
    },
    {
      "entropy": 5.497450685501098,
      "epoch": 2.361318582579884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004449532740151717,
      "loss": 5.021,
      "mean_token_accuracy": 0.22476583421230317,
      "num_tokens": 46924946.0,
      "step": 21025
    },
    {
      "entropy": 5.552854776382446,
      "epoch": 2.361880159487842,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044492696088960985,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21217524856328965,
      "num_tokens": 46936756.0,
      "step": 21030
    },
    {
      "entropy": 5.616649580001831,
      "epoch": 2.3624417363957995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004449006423535078,
      "loss": 5.0923,
      "mean_token_accuracy": 0.21316900849342346,
      "num_tokens": 46948552.0,
      "step": 21035
    },
    {
      "entropy": 5.60118932723999,
      "epoch": 2.363003313303757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044487431840770325,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20847312957048417,
      "num_tokens": 46960926.0,
      "step": 21040
    },
    {
      "entropy": 5.586189937591553,
      "epoch": 2.3635648902117143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004448479890530347,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20977888405323028,
      "num_tokens": 46972657.0,
      "step": 21045
    },
    {
      "entropy": 5.571494483947754,
      "epoch": 2.364126467119672,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044482165429034014,
      "loss": 5.039,
      "mean_token_accuracy": 0.21684652864933013,
      "num_tokens": 46984585.0,
      "step": 21050
    },
    {
      "entropy": 5.499054098129273,
      "epoch": 2.3646880440276297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004447953141204583,
      "loss": 5.0673,
      "mean_token_accuracy": 0.21634261161088944,
      "num_tokens": 46996052.0,
      "step": 21055
    },
    {
      "entropy": 5.586583757400513,
      "epoch": 2.365249620935587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044476896854422785,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22112126052379608,
      "num_tokens": 47006185.0,
      "step": 21060
    },
    {
      "entropy": 5.494100618362427,
      "epoch": 2.3658111978435445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044474261756248745,
      "loss": 5.0333,
      "mean_token_accuracy": 0.22035948783159257,
      "num_tokens": 47016647.0,
      "step": 21065
    },
    {
      "entropy": 5.582767963409424,
      "epoch": 2.366372774751502,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004447162611760764,
      "loss": 5.0463,
      "mean_token_accuracy": 0.214677894115448,
      "num_tokens": 47026847.0,
      "step": 21070
    },
    {
      "entropy": 5.5505763530731205,
      "epoch": 2.36693435165946,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004446898993858338,
      "loss": 5.0927,
      "mean_token_accuracy": 0.2119578719139099,
      "num_tokens": 47037948.0,
      "step": 21075
    },
    {
      "entropy": 5.5542572975158695,
      "epoch": 2.3674959285674175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044466353219259884,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21751994490623475,
      "num_tokens": 47047751.0,
      "step": 21080
    },
    {
      "entropy": 5.54599609375,
      "epoch": 2.3680575054753747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004446371595972114,
      "loss": 5.014,
      "mean_token_accuracy": 0.21468401104211807,
      "num_tokens": 47059589.0,
      "step": 21085
    },
    {
      "entropy": 5.542170476913452,
      "epoch": 2.3686190823833324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004446107816005109,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21395278126001357,
      "num_tokens": 47070060.0,
      "step": 21090
    },
    {
      "entropy": 5.621419668197632,
      "epoch": 2.36918065929129,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004445843982033373,
      "loss": 5.0709,
      "mean_token_accuracy": 0.21810976266860962,
      "num_tokens": 47081419.0,
      "step": 21095
    },
    {
      "entropy": 5.514518690109253,
      "epoch": 2.3697422361992473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004445580094065307,
      "loss": 5.0411,
      "mean_token_accuracy": 0.2074613392353058,
      "num_tokens": 47093701.0,
      "step": 21100
    },
    {
      "entropy": 5.526240015029908,
      "epoch": 2.370303813107205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004445316152109313,
      "loss": 5.0015,
      "mean_token_accuracy": 0.2178979143500328,
      "num_tokens": 47104930.0,
      "step": 21105
    },
    {
      "entropy": 5.522853040695191,
      "epoch": 2.3708653900151626,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004445052156173796,
      "loss": 4.9792,
      "mean_token_accuracy": 0.21464157700538636,
      "num_tokens": 47115109.0,
      "step": 21110
    },
    {
      "entropy": 5.494487380981445,
      "epoch": 2.3714269669231203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004444788106267161,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22040158957242967,
      "num_tokens": 47125386.0,
      "step": 21115
    },
    {
      "entropy": 5.544689321517945,
      "epoch": 2.3719885438310775,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004444524002397815,
      "loss": 5.042,
      "mean_token_accuracy": 0.2158852830529213,
      "num_tokens": 47136066.0,
      "step": 21120
    },
    {
      "entropy": 5.582882881164551,
      "epoch": 2.372550120739035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004444259844574167,
      "loss": 5.0717,
      "mean_token_accuracy": 0.21625349521636963,
      "num_tokens": 47147551.0,
      "step": 21125
    },
    {
      "entropy": 5.575290775299072,
      "epoch": 2.373111697646993,
      "grad_norm": 1.0,
      "learning_rate": 0.00044439956328046296,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21402848660945892,
      "num_tokens": 47158798.0,
      "step": 21130
    },
    {
      "entropy": 5.60425763130188,
      "epoch": 2.3736732745549505,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004443731367097613,
      "loss": 5.0925,
      "mean_token_accuracy": 0.21053385585546494,
      "num_tokens": 47169519.0,
      "step": 21135
    },
    {
      "entropy": 5.571637153625488,
      "epoch": 2.3742348514629077,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044434670474615334,
      "loss": 5.0414,
      "mean_token_accuracy": 0.21531961858272552,
      "num_tokens": 47180147.0,
      "step": 21140
    },
    {
      "entropy": 5.576913166046142,
      "epoch": 2.3747964283708654,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044432026739048064,
      "loss": 5.0887,
      "mean_token_accuracy": 0.2163708120584488,
      "num_tokens": 47191832.0,
      "step": 21145
    },
    {
      "entropy": 5.484866809844971,
      "epoch": 2.375358005278823,
      "grad_norm": 1.0,
      "learning_rate": 0.000444293824643585,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22523394227027893,
      "num_tokens": 47201979.0,
      "step": 21150
    },
    {
      "entropy": 5.570370244979858,
      "epoch": 2.3759195821867802,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004442673765063082,
      "loss": 5.1144,
      "mean_token_accuracy": 0.21057225316762923,
      "num_tokens": 47212271.0,
      "step": 21155
    },
    {
      "entropy": 5.540214729309082,
      "epoch": 2.376481159094738,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004442409229794925,
      "loss": 4.9754,
      "mean_token_accuracy": 0.2144185185432434,
      "num_tokens": 47223412.0,
      "step": 21160
    },
    {
      "entropy": 5.607240819931031,
      "epoch": 2.3770427360026956,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004442144640639803,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20397671610116958,
      "num_tokens": 47235919.0,
      "step": 21165
    },
    {
      "entropy": 5.552993440628052,
      "epoch": 2.377604312910653,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004441879997606138,
      "loss": 5.0163,
      "mean_token_accuracy": 0.2180008575320244,
      "num_tokens": 47246568.0,
      "step": 21170
    },
    {
      "entropy": 5.607348966598511,
      "epoch": 2.378165889818611,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004441615300702358,
      "loss": 5.1008,
      "mean_token_accuracy": 0.21429647654294967,
      "num_tokens": 47259474.0,
      "step": 21175
    },
    {
      "entropy": 5.54945821762085,
      "epoch": 2.378727466726568,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004441350549936891,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21769962459802628,
      "num_tokens": 47270906.0,
      "step": 21180
    },
    {
      "entropy": 5.49964747428894,
      "epoch": 2.3792890436345258,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004441085745318167,
      "loss": 5.0976,
      "mean_token_accuracy": 0.2016710877418518,
      "num_tokens": 47283341.0,
      "step": 21185
    },
    {
      "entropy": 5.606412506103515,
      "epoch": 2.3798506205424834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044408208868546164,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21590970903635026,
      "num_tokens": 47294590.0,
      "step": 21190
    },
    {
      "entropy": 5.4556268692016605,
      "epoch": 2.3804121974504406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004440555974554673,
      "loss": 4.875,
      "mean_token_accuracy": 0.2220437839627266,
      "num_tokens": 47305111.0,
      "step": 21195
    },
    {
      "entropy": 5.523229742050171,
      "epoch": 2.3809737743583983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044402910084267717,
      "loss": 5.0012,
      "mean_token_accuracy": 0.22105286419391632,
      "num_tokens": 47316145.0,
      "step": 21200
    },
    {
      "entropy": 5.556243276596069,
      "epoch": 2.381535351266356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004440025988479349,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22023187726736068,
      "num_tokens": 47327085.0,
      "step": 21205
    },
    {
      "entropy": 5.576622533798218,
      "epoch": 2.3820969281743136,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044397609147208436,
      "loss": 5.1543,
      "mean_token_accuracy": 0.2071654126048088,
      "num_tokens": 47338308.0,
      "step": 21210
    },
    {
      "entropy": 5.58474645614624,
      "epoch": 2.382658505082271,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004439495787159695,
      "loss": 5.1114,
      "mean_token_accuracy": 0.21173688322305678,
      "num_tokens": 47348717.0,
      "step": 21215
    },
    {
      "entropy": 5.55818510055542,
      "epoch": 2.3832200819902285,
      "grad_norm": 1.0,
      "learning_rate": 0.00044392306058043453,
      "loss": 4.9654,
      "mean_token_accuracy": 0.22702925503253937,
      "num_tokens": 47359804.0,
      "step": 21220
    },
    {
      "entropy": 5.527502012252808,
      "epoch": 2.383781658898186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004438965370663238,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21127522736787796,
      "num_tokens": 47370589.0,
      "step": 21225
    },
    {
      "entropy": 5.507923555374146,
      "epoch": 2.384343235806144,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004438700081744819,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21896038204431534,
      "num_tokens": 47381586.0,
      "step": 21230
    },
    {
      "entropy": 5.4773296356201175,
      "epoch": 2.384904812714101,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044384347390575337,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22752731591463088,
      "num_tokens": 47394064.0,
      "step": 21235
    },
    {
      "entropy": 5.515050554275513,
      "epoch": 2.3854663896220587,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004438169342609831,
      "loss": 5.0267,
      "mean_token_accuracy": 0.21912865042686464,
      "num_tokens": 47405371.0,
      "step": 21240
    },
    {
      "entropy": 5.436237907409668,
      "epoch": 2.3860279665300164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044379038924101627,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21847208887338637,
      "num_tokens": 47416446.0,
      "step": 21245
    },
    {
      "entropy": 5.512809896469117,
      "epoch": 2.3865895434379736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004437638388466979,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22685326486825944,
      "num_tokens": 47426588.0,
      "step": 21250
    },
    {
      "entropy": 5.562418365478516,
      "epoch": 2.3871511203459312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044373728307887353,
      "loss": 5.134,
      "mean_token_accuracy": 0.20546475201845169,
      "num_tokens": 47438015.0,
      "step": 21255
    },
    {
      "entropy": 5.657154703140259,
      "epoch": 2.387712697253889,
      "grad_norm": 1.0,
      "learning_rate": 0.00044371072193838857,
      "loss": 5.1575,
      "mean_token_accuracy": 0.20751930326223372,
      "num_tokens": 47448953.0,
      "step": 21260
    },
    {
      "entropy": 5.589694356918335,
      "epoch": 2.3882742741618466,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044368415542608874,
      "loss": 5.0176,
      "mean_token_accuracy": 0.2232425943017006,
      "num_tokens": 47459669.0,
      "step": 21265
    },
    {
      "entropy": 5.63691668510437,
      "epoch": 2.3888358510698042,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044365758354282005,
      "loss": 5.0789,
      "mean_token_accuracy": 0.21518020927906037,
      "num_tokens": 47471018.0,
      "step": 21270
    },
    {
      "entropy": 5.5892712593078615,
      "epoch": 2.3893974279777614,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004436310062894285,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21414951384067535,
      "num_tokens": 47481613.0,
      "step": 21275
    },
    {
      "entropy": 5.464930725097656,
      "epoch": 2.389959004885719,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004436044236667603,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21735286265611647,
      "num_tokens": 47491245.0,
      "step": 21280
    },
    {
      "entropy": 5.5089075565338135,
      "epoch": 2.3905205817936768,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004435778356756618,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2192789539694786,
      "num_tokens": 47502308.0,
      "step": 21285
    },
    {
      "entropy": 5.439525556564331,
      "epoch": 2.391082158701634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044355124231697967,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22227999418973923,
      "num_tokens": 47512799.0,
      "step": 21290
    },
    {
      "entropy": 5.5816398620605465,
      "epoch": 2.3916437356095916,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004435246435915607,
      "loss": 5.0501,
      "mean_token_accuracy": 0.22246460914611815,
      "num_tokens": 47524740.0,
      "step": 21295
    },
    {
      "entropy": 5.633766794204712,
      "epoch": 2.3922053125175493,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004434980395002516,
      "loss": 5.1025,
      "mean_token_accuracy": 0.215407395362854,
      "num_tokens": 47535785.0,
      "step": 21300
    },
    {
      "entropy": 5.521541357040405,
      "epoch": 2.392766889425507,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044347143004389955,
      "loss": 4.909,
      "mean_token_accuracy": 0.22849756181240083,
      "num_tokens": 47545698.0,
      "step": 21305
    },
    {
      "entropy": 5.461128377914429,
      "epoch": 2.393328466333464,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004434448152233519,
      "loss": 5.0563,
      "mean_token_accuracy": 0.22308383882045746,
      "num_tokens": 47559786.0,
      "step": 21310
    },
    {
      "entropy": 5.535771417617798,
      "epoch": 2.393890043241422,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044341819503945594,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21891767978668214,
      "num_tokens": 47570267.0,
      "step": 21315
    },
    {
      "entropy": 5.632030296325683,
      "epoch": 2.3944516201493795,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433915694930594,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20712898820638656,
      "num_tokens": 47580462.0,
      "step": 21320
    },
    {
      "entropy": 5.5144202709198,
      "epoch": 2.395013197057337,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004433649385850099,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2163635790348053,
      "num_tokens": 47590651.0,
      "step": 21325
    },
    {
      "entropy": 5.625806188583374,
      "epoch": 2.3955747739652944,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004433383023161555,
      "loss": 5.1651,
      "mean_token_accuracy": 0.2140408545732498,
      "num_tokens": 47602263.0,
      "step": 21330
    },
    {
      "entropy": 5.572895669937134,
      "epoch": 2.396136350873252,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004433116606873443,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22223273515701295,
      "num_tokens": 47613570.0,
      "step": 21335
    },
    {
      "entropy": 5.568210220336914,
      "epoch": 2.3966979277812097,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044328501369942454,
      "loss": 5.0832,
      "mean_token_accuracy": 0.21449922770261765,
      "num_tokens": 47624915.0,
      "step": 21340
    },
    {
      "entropy": 5.535996913909912,
      "epoch": 2.3972595046891674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004432583613532446,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21386823654174805,
      "num_tokens": 47635974.0,
      "step": 21345
    },
    {
      "entropy": 5.542981958389282,
      "epoch": 2.3978210815971246,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004432317036496533,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22272330820560454,
      "num_tokens": 47647312.0,
      "step": 21350
    },
    {
      "entropy": 5.488211727142334,
      "epoch": 2.3983826585050823,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044320504058949916,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22205788642168045,
      "num_tokens": 47659224.0,
      "step": 21355
    },
    {
      "entropy": 5.526345872879029,
      "epoch": 2.39894423541304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004431783721736314,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21710985898971558,
      "num_tokens": 47670720.0,
      "step": 21360
    },
    {
      "entropy": 5.527979373931885,
      "epoch": 2.3995058123209976,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044315169840289896,
      "loss": 5.0373,
      "mean_token_accuracy": 0.2167396977543831,
      "num_tokens": 47680646.0,
      "step": 21365
    },
    {
      "entropy": 5.60579662322998,
      "epoch": 2.400067389228955,
      "grad_norm": 1.0,
      "learning_rate": 0.0004431250192781513,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21877805888652802,
      "num_tokens": 47692155.0,
      "step": 21370
    },
    {
      "entropy": 5.633114433288574,
      "epoch": 2.4006289661369125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004430983348002378,
      "loss": 5.1441,
      "mean_token_accuracy": 0.21032606810331345,
      "num_tokens": 47703662.0,
      "step": 21375
    },
    {
      "entropy": 5.534679269790649,
      "epoch": 2.40119054304487,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044307164497000813,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21744474172592163,
      "num_tokens": 47713617.0,
      "step": 21380
    },
    {
      "entropy": 5.5374634742736815,
      "epoch": 2.4017521199528273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044304494978831204,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22166139483451844,
      "num_tokens": 47725051.0,
      "step": 21385
    },
    {
      "entropy": 5.4525754928588865,
      "epoch": 2.402313696860785,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004430182492559997,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22397128045558928,
      "num_tokens": 47735023.0,
      "step": 21390
    },
    {
      "entropy": 5.516439485549927,
      "epoch": 2.4028752737687427,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000442991543373921,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21986833661794664,
      "num_tokens": 47746252.0,
      "step": 21395
    },
    {
      "entropy": 5.541053056716919,
      "epoch": 2.4034368506767003,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004429648321429264,
      "loss": 5.0974,
      "mean_token_accuracy": 0.2094160199165344,
      "num_tokens": 47756980.0,
      "step": 21400
    },
    {
      "entropy": 5.565073490142822,
      "epoch": 2.4039984275846575,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044293811556386656,
      "loss": 5.0412,
      "mean_token_accuracy": 0.22232271432876588,
      "num_tokens": 47767633.0,
      "step": 21405
    },
    {
      "entropy": 5.495944213867188,
      "epoch": 2.404560004492615,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004429113936375918,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21777793467044831,
      "num_tokens": 47779346.0,
      "step": 21410
    },
    {
      "entropy": 5.655402565002442,
      "epoch": 2.405121581400573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044288466636495323,
      "loss": 5.0994,
      "mean_token_accuracy": 0.21635898649692537,
      "num_tokens": 47791039.0,
      "step": 21415
    },
    {
      "entropy": 5.598259925842285,
      "epoch": 2.4056831583085305,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044285793374680184,
      "loss": 5.125,
      "mean_token_accuracy": 0.21230610609054565,
      "num_tokens": 47802096.0,
      "step": 21420
    },
    {
      "entropy": 5.54836220741272,
      "epoch": 2.4062447352164877,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044283119578398866,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21002331525087356,
      "num_tokens": 47814087.0,
      "step": 21425
    },
    {
      "entropy": 5.49899377822876,
      "epoch": 2.4068063121244454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004428044524773651,
      "loss": 4.9778,
      "mean_token_accuracy": 0.21474130004644393,
      "num_tokens": 47824656.0,
      "step": 21430
    },
    {
      "entropy": 5.517704582214355,
      "epoch": 2.407367889032403,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004427777038277827,
      "loss": 5.0018,
      "mean_token_accuracy": 0.22438367158174516,
      "num_tokens": 47836045.0,
      "step": 21435
    },
    {
      "entropy": 5.561577463150025,
      "epoch": 2.4079294659403607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044275094983609313,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21972071677446364,
      "num_tokens": 47846494.0,
      "step": 21440
    },
    {
      "entropy": 5.60315728187561,
      "epoch": 2.408491042848318,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044272419050314833,
      "loss": 5.0934,
      "mean_token_accuracy": 0.21812470108270646,
      "num_tokens": 47858689.0,
      "step": 21445
    },
    {
      "entropy": 5.4736391544342045,
      "epoch": 2.4090526197562756,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044269742582980017,
      "loss": 4.9771,
      "mean_token_accuracy": 0.2168256014585495,
      "num_tokens": 47870534.0,
      "step": 21450
    },
    {
      "entropy": 5.523947763442993,
      "epoch": 2.4096141966642333,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000442670655816901,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21528241336345671,
      "num_tokens": 47880354.0,
      "step": 21455
    },
    {
      "entropy": 5.581337118148804,
      "epoch": 2.410175773572191,
      "grad_norm": 0.890625,
      "learning_rate": 0.00044264388046530304,
      "loss": 5.0548,
      "mean_token_accuracy": 0.21407393664121627,
      "num_tokens": 47892528.0,
      "step": 21460
    },
    {
      "entropy": 5.596236610412598,
      "epoch": 2.410737350480148,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044261709977585893,
      "loss": 5.0682,
      "mean_token_accuracy": 0.22339522391557692,
      "num_tokens": 47904653.0,
      "step": 21465
    },
    {
      "entropy": 5.610722494125366,
      "epoch": 2.411298927388106,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004425903137494214,
      "loss": 5.136,
      "mean_token_accuracy": 0.2092830702662468,
      "num_tokens": 47916452.0,
      "step": 21470
    },
    {
      "entropy": 5.585277795791626,
      "epoch": 2.4118605042960635,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004425635223868433,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21341212838888168,
      "num_tokens": 47927746.0,
      "step": 21475
    },
    {
      "entropy": 5.574788284301758,
      "epoch": 2.4124220812040207,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044253672568897754,
      "loss": 5.0452,
      "mean_token_accuracy": 0.20959641188383102,
      "num_tokens": 47941057.0,
      "step": 21480
    },
    {
      "entropy": 5.555737543106079,
      "epoch": 2.4129836581119783,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004425099236566775,
      "loss": 5.08,
      "mean_token_accuracy": 0.21327551156282426,
      "num_tokens": 47951912.0,
      "step": 21485
    },
    {
      "entropy": 5.553208732604981,
      "epoch": 2.413545235019936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044248311629079657,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21758248955011367,
      "num_tokens": 47963483.0,
      "step": 21490
    },
    {
      "entropy": 5.634589529037475,
      "epoch": 2.4141068119278937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004424563035921882,
      "loss": 5.1145,
      "mean_token_accuracy": 0.21504732966423035,
      "num_tokens": 47975047.0,
      "step": 21495
    },
    {
      "entropy": 5.593435001373291,
      "epoch": 2.414668388835851,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044242948556170634,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21129971146583557,
      "num_tokens": 47986432.0,
      "step": 21500
    },
    {
      "entropy": 5.517426300048828,
      "epoch": 2.4152299657438085,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044240266220020455,
      "loss": 4.9843,
      "mean_token_accuracy": 0.22007084339857103,
      "num_tokens": 47997283.0,
      "step": 21505
    },
    {
      "entropy": 5.5323097705841064,
      "epoch": 2.415791542651766,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004423758335085371,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21180306673049926,
      "num_tokens": 48008353.0,
      "step": 21510
    },
    {
      "entropy": 5.532192039489746,
      "epoch": 2.416353119559724,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044234899948755827,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21177671402692794,
      "num_tokens": 48020065.0,
      "step": 21515
    },
    {
      "entropy": 5.5409832954406735,
      "epoch": 2.416914696467681,
      "grad_norm": 1.125,
      "learning_rate": 0.00044232216013812226,
      "loss": 5.049,
      "mean_token_accuracy": 0.21223694533109666,
      "num_tokens": 48030062.0,
      "step": 21520
    },
    {
      "entropy": 5.492615747451782,
      "epoch": 2.4174762733756388,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004422953154610838,
      "loss": 4.9934,
      "mean_token_accuracy": 0.20923204869031906,
      "num_tokens": 48039639.0,
      "step": 21525
    },
    {
      "entropy": 5.555186986923218,
      "epoch": 2.4180378502835964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004422684654572977,
      "loss": 5.0709,
      "mean_token_accuracy": 0.2114972487092018,
      "num_tokens": 48050428.0,
      "step": 21530
    },
    {
      "entropy": 5.565184307098389,
      "epoch": 2.418599427191554,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044224161012761874,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21541703194379808,
      "num_tokens": 48061214.0,
      "step": 21535
    },
    {
      "entropy": 5.485684061050415,
      "epoch": 2.4191610040995113,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044221474947290205,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22256745547056198,
      "num_tokens": 48071983.0,
      "step": 21540
    },
    {
      "entropy": 5.503530693054199,
      "epoch": 2.419722581007469,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004421878834940029,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22453695237636567,
      "num_tokens": 48083411.0,
      "step": 21545
    },
    {
      "entropy": 5.560049533843994,
      "epoch": 2.4202841579154266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044216101219177667,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21257443875074386,
      "num_tokens": 48094718.0,
      "step": 21550
    },
    {
      "entropy": 5.566584873199463,
      "epoch": 2.4208457348233843,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044213413556707905,
      "loss": 5.033,
      "mean_token_accuracy": 0.21447941958904265,
      "num_tokens": 48106663.0,
      "step": 21555
    },
    {
      "entropy": 5.565326881408692,
      "epoch": 2.4214073117313415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004421072536207656,
      "loss": 5.023,
      "mean_token_accuracy": 0.21907533556222916,
      "num_tokens": 48117232.0,
      "step": 21560
    },
    {
      "entropy": 5.56593804359436,
      "epoch": 2.421968888639299,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044208036635369247,
      "loss": 5.0395,
      "mean_token_accuracy": 0.224062579870224,
      "num_tokens": 48128817.0,
      "step": 21565
    },
    {
      "entropy": 5.546529817581177,
      "epoch": 2.422530465547257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004420534737667157,
      "loss": 4.9991,
      "mean_token_accuracy": 0.22189513891935347,
      "num_tokens": 48140893.0,
      "step": 21570
    },
    {
      "entropy": 5.501858186721802,
      "epoch": 2.423092042455214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044202657586069153,
      "loss": 4.8989,
      "mean_token_accuracy": 0.21900242567062378,
      "num_tokens": 48152661.0,
      "step": 21575
    },
    {
      "entropy": 5.540249967575074,
      "epoch": 2.4236536193631717,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044199967263647635,
      "loss": 5.047,
      "mean_token_accuracy": 0.21719842106103898,
      "num_tokens": 48163342.0,
      "step": 21580
    },
    {
      "entropy": 5.559485912322998,
      "epoch": 2.4242151962711294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004419727640949268,
      "loss": 5.0423,
      "mean_token_accuracy": 0.21401796787977218,
      "num_tokens": 48175321.0,
      "step": 21585
    },
    {
      "entropy": 5.60004768371582,
      "epoch": 2.424776773179087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004419458502368998,
      "loss": 5.0747,
      "mean_token_accuracy": 0.2131578132510185,
      "num_tokens": 48186847.0,
      "step": 21590
    },
    {
      "entropy": 5.597808933258056,
      "epoch": 2.4253383500870442,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044191893106325207,
      "loss": 5.1329,
      "mean_token_accuracy": 0.21879195123910905,
      "num_tokens": 48198840.0,
      "step": 21595
    },
    {
      "entropy": 5.621781635284424,
      "epoch": 2.425899926995002,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004418920065748409,
      "loss": 5.1224,
      "mean_token_accuracy": 0.20851902514696122,
      "num_tokens": 48209948.0,
      "step": 21600
    },
    {
      "entropy": 5.514352798461914,
      "epoch": 2.4264615039029596,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044186507677252345,
      "loss": 4.9944,
      "mean_token_accuracy": 0.2232287585735321,
      "num_tokens": 48220125.0,
      "step": 21605
    },
    {
      "entropy": 5.549229001998901,
      "epoch": 2.427023080810917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004418381416571573,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21846702098846435,
      "num_tokens": 48230736.0,
      "step": 21610
    },
    {
      "entropy": 5.4540637016296385,
      "epoch": 2.4275846577188744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004418112012296,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22695451378822326,
      "num_tokens": 48241387.0,
      "step": 21615
    },
    {
      "entropy": 5.560689687728882,
      "epoch": 2.428146234626832,
      "grad_norm": 1.0,
      "learning_rate": 0.00044178425549070934,
      "loss": 4.9885,
      "mean_token_accuracy": 0.22091599851846694,
      "num_tokens": 48252279.0,
      "step": 21620
    },
    {
      "entropy": 5.610567331314087,
      "epoch": 2.4287078115347898,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044175730444134326,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21970888823270798,
      "num_tokens": 48264235.0,
      "step": 21625
    },
    {
      "entropy": 5.509802961349488,
      "epoch": 2.4292693884427474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044173034808236,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22488942593336106,
      "num_tokens": 48276194.0,
      "step": 21630
    },
    {
      "entropy": 5.610980558395386,
      "epoch": 2.4298309653507046,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004417033864146177,
      "loss": 5.0671,
      "mean_token_accuracy": 0.22037188410758973,
      "num_tokens": 48287129.0,
      "step": 21635
    },
    {
      "entropy": 5.498099279403687,
      "epoch": 2.4303925422586623,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004416764194389751,
      "loss": 5.0086,
      "mean_token_accuracy": 0.21537086516618728,
      "num_tokens": 48298850.0,
      "step": 21640
    },
    {
      "entropy": 5.579710245132446,
      "epoch": 2.43095411916662,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004416494471562904,
      "loss": 5.0582,
      "mean_token_accuracy": 0.21311228573322297,
      "num_tokens": 48309968.0,
      "step": 21645
    },
    {
      "entropy": 5.558413743972778,
      "epoch": 2.4315156960745776,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004416224695674229,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21595038175582887,
      "num_tokens": 48320419.0,
      "step": 21650
    },
    {
      "entropy": 5.558744335174561,
      "epoch": 2.432077272982535,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044159548667323126,
      "loss": 5.0756,
      "mean_token_accuracy": 0.21511569172143935,
      "num_tokens": 48332611.0,
      "step": 21655
    },
    {
      "entropy": 5.536865282058716,
      "epoch": 2.4326388498904925,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004415684984745747,
      "loss": 5.0509,
      "mean_token_accuracy": 0.21908067762851716,
      "num_tokens": 48344551.0,
      "step": 21660
    },
    {
      "entropy": 5.6408796310424805,
      "epoch": 2.43320042679845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004415415049723126,
      "loss": 5.0569,
      "mean_token_accuracy": 0.212600739300251,
      "num_tokens": 48355142.0,
      "step": 21665
    },
    {
      "entropy": 5.527834939956665,
      "epoch": 2.4337620037064074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004415145061673043,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21315344274044037,
      "num_tokens": 48367342.0,
      "step": 21670
    },
    {
      "entropy": 5.616356325149536,
      "epoch": 2.434323580614365,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004414875020604096,
      "loss": 5.0734,
      "mean_token_accuracy": 0.2147837370634079,
      "num_tokens": 48378608.0,
      "step": 21675
    },
    {
      "entropy": 5.60773286819458,
      "epoch": 2.4348851575223227,
      "grad_norm": 1.0,
      "learning_rate": 0.0004414604926524882,
      "loss": 5.0697,
      "mean_token_accuracy": 0.21642567664384843,
      "num_tokens": 48390455.0,
      "step": 21680
    },
    {
      "entropy": 5.568949890136719,
      "epoch": 2.4354467344302804,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414334779444002,
      "loss": 5.0259,
      "mean_token_accuracy": 0.2175319790840149,
      "num_tokens": 48401723.0,
      "step": 21685
    },
    {
      "entropy": 5.551620674133301,
      "epoch": 2.4360083113382376,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044140645793700573,
      "loss": 5.0495,
      "mean_token_accuracy": 0.2114250436425209,
      "num_tokens": 48414208.0,
      "step": 21690
    },
    {
      "entropy": 5.545775413513184,
      "epoch": 2.4365698882461952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044137943263116515,
      "loss": 5.0249,
      "mean_token_accuracy": 0.22128980308771135,
      "num_tokens": 48425978.0,
      "step": 21695
    },
    {
      "entropy": 5.520871114730835,
      "epoch": 2.437131465154153,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004413524020277388,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22787272036075593,
      "num_tokens": 48435988.0,
      "step": 21700
    },
    {
      "entropy": 5.52367525100708,
      "epoch": 2.4376930420621106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004413253661275875,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21544582396745682,
      "num_tokens": 48447217.0,
      "step": 21705
    },
    {
      "entropy": 5.561659431457519,
      "epoch": 2.438254618970068,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044129832493157195,
      "loss": 5.0814,
      "mean_token_accuracy": 0.2130025178194046,
      "num_tokens": 48458597.0,
      "step": 21710
    },
    {
      "entropy": 5.529704856872558,
      "epoch": 2.4388161958780255,
      "grad_norm": 1.0,
      "learning_rate": 0.00044127127844055335,
      "loss": 4.9988,
      "mean_token_accuracy": 0.2169817179441452,
      "num_tokens": 48470601.0,
      "step": 21715
    },
    {
      "entropy": 5.540048742294312,
      "epoch": 2.439377772785983,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044124422665539273,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2163601115345955,
      "num_tokens": 48481570.0,
      "step": 21720
    },
    {
      "entropy": 5.63019700050354,
      "epoch": 2.4399393496939408,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004412171695769515,
      "loss": 5.1493,
      "mean_token_accuracy": 0.20787879675626755,
      "num_tokens": 48492146.0,
      "step": 21725
    },
    {
      "entropy": 5.515824794769287,
      "epoch": 2.440500926601898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044119010720609106,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21134107410907746,
      "num_tokens": 48504644.0,
      "step": 21730
    },
    {
      "entropy": 5.494596529006958,
      "epoch": 2.4410625035098557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044116303954367314,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21949250251054764,
      "num_tokens": 48516405.0,
      "step": 21735
    },
    {
      "entropy": 5.521606492996216,
      "epoch": 2.4416240804178133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004411359665905596,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21290522068738937,
      "num_tokens": 48528759.0,
      "step": 21740
    },
    {
      "entropy": 5.655643463134766,
      "epoch": 2.442185657325771,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004411088883476125,
      "loss": 5.1461,
      "mean_token_accuracy": 0.21223575621843338,
      "num_tokens": 48540731.0,
      "step": 21745
    },
    {
      "entropy": 5.595367765426635,
      "epoch": 2.442747234233728,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044108180481569387,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21804226636886598,
      "num_tokens": 48553642.0,
      "step": 21750
    },
    {
      "entropy": 5.650684928894043,
      "epoch": 2.443308811141686,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044105471599566626,
      "loss": 5.0801,
      "mean_token_accuracy": 0.2166365936398506,
      "num_tokens": 48564596.0,
      "step": 21755
    },
    {
      "entropy": 5.538341474533081,
      "epoch": 2.4438703880496435,
      "grad_norm": 0.96875,
      "learning_rate": 0.000441027621888392,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2120039775967598,
      "num_tokens": 48574521.0,
      "step": 21760
    },
    {
      "entropy": 5.506499814987182,
      "epoch": 2.4444319649576007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044100052249473393,
      "loss": 4.9343,
      "mean_token_accuracy": 0.21970731914043426,
      "num_tokens": 48586299.0,
      "step": 21765
    },
    {
      "entropy": 5.60390043258667,
      "epoch": 2.4449935418655584,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004409734178155548,
      "loss": 5.0907,
      "mean_token_accuracy": 0.21654146164655685,
      "num_tokens": 48598366.0,
      "step": 21770
    },
    {
      "entropy": 5.571668720245361,
      "epoch": 2.445555118773516,
      "grad_norm": 1.125,
      "learning_rate": 0.0004409463078517178,
      "loss": 5.015,
      "mean_token_accuracy": 0.2153596729040146,
      "num_tokens": 48609659.0,
      "step": 21775
    },
    {
      "entropy": 5.531099510192871,
      "epoch": 2.4461166956814737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044091919260408585,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2255837246775627,
      "num_tokens": 48621623.0,
      "step": 21780
    },
    {
      "entropy": 5.510299158096314,
      "epoch": 2.446678272589431,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004408920720735225,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21605013757944108,
      "num_tokens": 48632711.0,
      "step": 21785
    },
    {
      "entropy": 5.531552982330322,
      "epoch": 2.4472398494973886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044086494626089127,
      "loss": 5.0509,
      "mean_token_accuracy": 0.22462499588727952,
      "num_tokens": 48644518.0,
      "step": 21790
    },
    {
      "entropy": 5.539181327819824,
      "epoch": 2.4478014264053463,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004408378151670558,
      "loss": 5.0538,
      "mean_token_accuracy": 0.2135454922914505,
      "num_tokens": 48655261.0,
      "step": 21795
    },
    {
      "entropy": 5.513379955291748,
      "epoch": 2.448363003313304,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044081067879288,
      "loss": 4.9934,
      "mean_token_accuracy": 0.22263031154870988,
      "num_tokens": 48666701.0,
      "step": 21800
    },
    {
      "entropy": 5.469851064682007,
      "epoch": 2.448924580221261,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004407835371392279,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22177508771419524,
      "num_tokens": 48677919.0,
      "step": 21805
    },
    {
      "entropy": 5.538128137588501,
      "epoch": 2.449486157129219,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004407563902069636,
      "loss": 4.9994,
      "mean_token_accuracy": 0.23091389536857604,
      "num_tokens": 48688680.0,
      "step": 21810
    },
    {
      "entropy": 5.610378408432007,
      "epoch": 2.4500477340371765,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004407292379969515,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21744329929351808,
      "num_tokens": 48701149.0,
      "step": 21815
    },
    {
      "entropy": 5.529277515411377,
      "epoch": 2.450609310945134,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004407020805100563,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2127997621893883,
      "num_tokens": 48712236.0,
      "step": 21820
    },
    {
      "entropy": 5.46480073928833,
      "epoch": 2.4511708878530913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004406749177471425,
      "loss": 4.96,
      "mean_token_accuracy": 0.2235446661710739,
      "num_tokens": 48723923.0,
      "step": 21825
    },
    {
      "entropy": 5.484795761108399,
      "epoch": 2.451732464761049,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044064774970907516,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21914336681365967,
      "num_tokens": 48734925.0,
      "step": 21830
    },
    {
      "entropy": 5.5674731731414795,
      "epoch": 2.4522940416690067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004406205763967193,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2195023074746132,
      "num_tokens": 48745758.0,
      "step": 21835
    },
    {
      "entropy": 5.482881116867065,
      "epoch": 2.4528556185769643,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004405933978109398,
      "loss": 4.9715,
      "mean_token_accuracy": 0.21867056488990783,
      "num_tokens": 48756899.0,
      "step": 21840
    },
    {
      "entropy": 5.608753442764282,
      "epoch": 2.4534171954849215,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044056621395260247,
      "loss": 5.0174,
      "mean_token_accuracy": 0.2154842883348465,
      "num_tokens": 48766805.0,
      "step": 21845
    },
    {
      "entropy": 5.585347700119018,
      "epoch": 2.453978772392879,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044053902482257265,
      "loss": 5.042,
      "mean_token_accuracy": 0.22036779075860977,
      "num_tokens": 48779360.0,
      "step": 21850
    },
    {
      "entropy": 5.522671556472778,
      "epoch": 2.454540349300837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000440511830421716,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21524800807237626,
      "num_tokens": 48789836.0,
      "step": 21855
    },
    {
      "entropy": 5.448854398727417,
      "epoch": 2.455101926208794,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004404846307508985,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2141500234603882,
      "num_tokens": 48801327.0,
      "step": 21860
    },
    {
      "entropy": 5.508322477340698,
      "epoch": 2.4556635031167517,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044045742581098624,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2231780081987381,
      "num_tokens": 48812521.0,
      "step": 21865
    },
    {
      "entropy": 5.5456009864807125,
      "epoch": 2.4562250800247094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004404302156028453,
      "loss": 4.9944,
      "mean_token_accuracy": 0.21515218317508697,
      "num_tokens": 48824068.0,
      "step": 21870
    },
    {
      "entropy": 5.537049865722656,
      "epoch": 2.456786656932667,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044040300012734217,
      "loss": 5.0112,
      "mean_token_accuracy": 0.21742115467786788,
      "num_tokens": 48835183.0,
      "step": 21875
    },
    {
      "entropy": 5.551444864273071,
      "epoch": 2.4573482338406243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004403757793853434,
      "loss": 5.0504,
      "mean_token_accuracy": 0.21745655685663223,
      "num_tokens": 48846616.0,
      "step": 21880
    },
    {
      "entropy": 5.501677417755127,
      "epoch": 2.457909810748582,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004403485533777157,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21701752096414567,
      "num_tokens": 48858159.0,
      "step": 21885
    },
    {
      "entropy": 5.4930181980133055,
      "epoch": 2.4584713876565396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004403213221053258,
      "loss": 4.9516,
      "mean_token_accuracy": 0.2211826488375664,
      "num_tokens": 48870363.0,
      "step": 21890
    },
    {
      "entropy": 5.574721479415894,
      "epoch": 2.4590329645644973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044029408556904093,
      "loss": 5.071,
      "mean_token_accuracy": 0.21850603967905044,
      "num_tokens": 48882036.0,
      "step": 21895
    },
    {
      "entropy": 5.573212909698486,
      "epoch": 2.4595945414724545,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004402668437697283,
      "loss": 5.0303,
      "mean_token_accuracy": 0.2178829252719879,
      "num_tokens": 48893741.0,
      "step": 21900
    },
    {
      "entropy": 5.621312093734741,
      "epoch": 2.460156118380412,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044023959670825533,
      "loss": 5.1132,
      "mean_token_accuracy": 0.2123256206512451,
      "num_tokens": 48906260.0,
      "step": 21905
    },
    {
      "entropy": 5.598132801055908,
      "epoch": 2.46071769528837,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004402123443854894,
      "loss": 5.0887,
      "mean_token_accuracy": 0.20818998217582702,
      "num_tokens": 48918078.0,
      "step": 21910
    },
    {
      "entropy": 5.541524076461792,
      "epoch": 2.4612792721963275,
      "grad_norm": 1.0,
      "learning_rate": 0.0004401850868022984,
      "loss": 5.026,
      "mean_token_accuracy": 0.2142586663365364,
      "num_tokens": 48928999.0,
      "step": 21915
    },
    {
      "entropy": 5.492013311386108,
      "epoch": 2.4618408491042847,
      "grad_norm": 1.0,
      "learning_rate": 0.0004401578239595501,
      "loss": 4.9841,
      "mean_token_accuracy": 0.21293566524982452,
      "num_tokens": 48940587.0,
      "step": 21920
    },
    {
      "entropy": 5.571208381652832,
      "epoch": 2.4624024260122424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044013055585811274,
      "loss": 5.1299,
      "mean_token_accuracy": 0.21308628469705582,
      "num_tokens": 48951632.0,
      "step": 21925
    },
    {
      "entropy": 5.665954303741455,
      "epoch": 2.4629640029202,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044010328249885434,
      "loss": 5.1672,
      "mean_token_accuracy": 0.21363916397094726,
      "num_tokens": 48963578.0,
      "step": 21930
    },
    {
      "entropy": 5.572709321975708,
      "epoch": 2.4635255798281577,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004400760038826434,
      "loss": 5.0181,
      "mean_token_accuracy": 0.2178805112838745,
      "num_tokens": 48974690.0,
      "step": 21935
    },
    {
      "entropy": 5.614700698852539,
      "epoch": 2.464087156736115,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044004872001034856,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21723814755678178,
      "num_tokens": 48985871.0,
      "step": 21940
    },
    {
      "entropy": 5.475772428512573,
      "epoch": 2.4646487336440726,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004400214308828384,
      "loss": 4.9233,
      "mean_token_accuracy": 0.228187857568264,
      "num_tokens": 48996475.0,
      "step": 21945
    },
    {
      "entropy": 5.646323585510254,
      "epoch": 2.46521031055203,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043999413650098193,
      "loss": 5.1296,
      "mean_token_accuracy": 0.21441911309957504,
      "num_tokens": 49008900.0,
      "step": 21950
    },
    {
      "entropy": 5.455271530151367,
      "epoch": 2.4657718874599874,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004399668368656481,
      "loss": 4.971,
      "mean_token_accuracy": 0.21826575547456742,
      "num_tokens": 49020583.0,
      "step": 21955
    },
    {
      "entropy": 5.61666693687439,
      "epoch": 2.466333464367945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004399395319777062,
      "loss": 5.1039,
      "mean_token_accuracy": 0.20938464999198914,
      "num_tokens": 49031363.0,
      "step": 21960
    },
    {
      "entropy": 5.550467681884766,
      "epoch": 2.4668950412759028,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043991222183802564,
      "loss": 5.0712,
      "mean_token_accuracy": 0.21578730046749114,
      "num_tokens": 49043404.0,
      "step": 21965
    },
    {
      "entropy": 5.615428781509399,
      "epoch": 2.4674566181838604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000439884906447476,
      "loss": 5.0849,
      "mean_token_accuracy": 0.21261667758226394,
      "num_tokens": 49054240.0,
      "step": 21970
    },
    {
      "entropy": 5.506282186508178,
      "epoch": 2.4680181950918176,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000439857585806927,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22931258231401444,
      "num_tokens": 49064730.0,
      "step": 21975
    },
    {
      "entropy": 5.5686187744140625,
      "epoch": 2.4685797719997753,
      "grad_norm": 1.0,
      "learning_rate": 0.00043983025991724847,
      "loss": 5.0576,
      "mean_token_accuracy": 0.2182838350534439,
      "num_tokens": 49075218.0,
      "step": 21980
    },
    {
      "entropy": 5.524595069885254,
      "epoch": 2.469141348907733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004398029287793106,
      "loss": 5.0014,
      "mean_token_accuracy": 0.22363802790641785,
      "num_tokens": 49086182.0,
      "step": 21985
    },
    {
      "entropy": 5.598664951324463,
      "epoch": 2.4697029258156906,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00043977559239398353,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20227162688970565,
      "num_tokens": 49099054.0,
      "step": 21990
    },
    {
      "entropy": 5.51516580581665,
      "epoch": 2.470264502723648,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004397482507621376,
      "loss": 5.0276,
      "mean_token_accuracy": 0.22136638015508653,
      "num_tokens": 49110234.0,
      "step": 21995
    },
    {
      "entropy": 5.497951793670654,
      "epoch": 2.4708260796316055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397209038846436,
      "loss": 4.9958,
      "mean_token_accuracy": 0.21943045109510423,
      "num_tokens": 49121884.0,
      "step": 22000
    },
    {
      "entropy": 5.508752679824829,
      "epoch": 2.471387656539563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043969355176237207,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21629427671432494,
      "num_tokens": 49133491.0,
      "step": 22005
    },
    {
      "entropy": 5.475440216064453,
      "epoch": 2.471949233447521,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004396661943961941,
      "loss": 4.984,
      "mean_token_accuracy": 0.20998448133468628,
      "num_tokens": 49145895.0,
      "step": 22010
    },
    {
      "entropy": 5.625890588760376,
      "epoch": 2.472510810355478,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004396388317869805,
      "loss": 5.0702,
      "mean_token_accuracy": 0.21601071804761887,
      "num_tokens": 49157840.0,
      "step": 22015
    },
    {
      "entropy": 5.549044036865235,
      "epoch": 2.4730723872634357,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004396114639356027,
      "loss": 5.0236,
      "mean_token_accuracy": 0.22526059299707413,
      "num_tokens": 49170391.0,
      "step": 22020
    },
    {
      "entropy": 5.524961519241333,
      "epoch": 2.4736339641713934,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000439584090842932,
      "loss": 4.9848,
      "mean_token_accuracy": 0.21579521596431733,
      "num_tokens": 49181721.0,
      "step": 22025
    },
    {
      "entropy": 5.574447059631348,
      "epoch": 2.474195541079351,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043955671250984005,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21901206523180008,
      "num_tokens": 49192137.0,
      "step": 22030
    },
    {
      "entropy": 5.651259565353394,
      "epoch": 2.4747571179873082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043952932893719847,
      "loss": 5.0672,
      "mean_token_accuracy": 0.21795873641967772,
      "num_tokens": 49203260.0,
      "step": 22035
    },
    {
      "entropy": 5.5037641525268555,
      "epoch": 2.475318694895266,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004395019401258794,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2217392221093178,
      "num_tokens": 49214799.0,
      "step": 22040
    },
    {
      "entropy": 5.518495321273804,
      "epoch": 2.4758802718032236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043947454607675464,
      "loss": 4.964,
      "mean_token_accuracy": 0.22416856437921523,
      "num_tokens": 49224595.0,
      "step": 22045
    },
    {
      "entropy": 5.559446334838867,
      "epoch": 2.476441848711181,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004394471467906966,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21967774033546447,
      "num_tokens": 49235190.0,
      "step": 22050
    },
    {
      "entropy": 5.504490184783935,
      "epoch": 2.4770034256191384,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043941974226857763,
      "loss": 4.987,
      "mean_token_accuracy": 0.21906189620494843,
      "num_tokens": 49247103.0,
      "step": 22055
    },
    {
      "entropy": 5.6427021503448485,
      "epoch": 2.477565002527096,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004393923325112702,
      "loss": 5.082,
      "mean_token_accuracy": 0.21247027963399887,
      "num_tokens": 49259874.0,
      "step": 22060
    },
    {
      "entropy": 5.435581684112549,
      "epoch": 2.4781265794350538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004393649175196472,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2234013631939888,
      "num_tokens": 49271059.0,
      "step": 22065
    },
    {
      "entropy": 5.4833092212677,
      "epoch": 2.478688156343011,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004393374972945815,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22380977720022202,
      "num_tokens": 49283142.0,
      "step": 22070
    },
    {
      "entropy": 5.5668168544769285,
      "epoch": 2.4792497332509686,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043931007183694616,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21959057301282883,
      "num_tokens": 49294022.0,
      "step": 22075
    },
    {
      "entropy": 5.54131555557251,
      "epoch": 2.4798113101589263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043928264114761437,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21574143022298814,
      "num_tokens": 49305849.0,
      "step": 22080
    },
    {
      "entropy": 5.530166053771973,
      "epoch": 2.480372887066884,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043925520522745953,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21938502192497253,
      "num_tokens": 49316957.0,
      "step": 22085
    },
    {
      "entropy": 5.53228530883789,
      "epoch": 2.480934463974841,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004392277640773553,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2239670932292938,
      "num_tokens": 49329040.0,
      "step": 22090
    },
    {
      "entropy": 5.527100992202759,
      "epoch": 2.481496040882799,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043920031769817543,
      "loss": 5.0117,
      "mean_token_accuracy": 0.22559649497270584,
      "num_tokens": 49340037.0,
      "step": 22095
    },
    {
      "entropy": 5.501721477508545,
      "epoch": 2.4820576177907565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043917286609079364,
      "loss": 5.0085,
      "mean_token_accuracy": 0.20904731303453444,
      "num_tokens": 49350059.0,
      "step": 22100
    },
    {
      "entropy": 5.494209718704224,
      "epoch": 2.482619194698714,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043914540925608426,
      "loss": 4.9962,
      "mean_token_accuracy": 0.22121666520833969,
      "num_tokens": 49360880.0,
      "step": 22105
    },
    {
      "entropy": 5.602012586593628,
      "epoch": 2.4831807716066714,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004391179471949213,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21866675615310668,
      "num_tokens": 49370725.0,
      "step": 22110
    },
    {
      "entropy": 5.567810106277466,
      "epoch": 2.483742348514629,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043909047990817917,
      "loss": 5.0102,
      "mean_token_accuracy": 0.21709024757146836,
      "num_tokens": 49380991.0,
      "step": 22115
    },
    {
      "entropy": 5.511302089691162,
      "epoch": 2.4843039254225867,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043906300739673265,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21965509802103042,
      "num_tokens": 49390802.0,
      "step": 22120
    },
    {
      "entropy": 5.551313972473144,
      "epoch": 2.4848655023305444,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043903552966145624,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21891269385814666,
      "num_tokens": 49401507.0,
      "step": 22125
    },
    {
      "entropy": 5.5666718006134035,
      "epoch": 2.4854270792385016,
      "grad_norm": 1.03125,
      "learning_rate": 0.000439008046703225,
      "loss": 5.0877,
      "mean_token_accuracy": 0.21074884980916977,
      "num_tokens": 49412022.0,
      "step": 22130
    },
    {
      "entropy": 5.56012601852417,
      "epoch": 2.4859886561464593,
      "grad_norm": 1.0625,
      "learning_rate": 0.000438980558522914,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22303961962461472,
      "num_tokens": 49423294.0,
      "step": 22135
    },
    {
      "entropy": 5.513470840454102,
      "epoch": 2.486550233054417,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004389530651213983,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21629400849342345,
      "num_tokens": 49433915.0,
      "step": 22140
    },
    {
      "entropy": 5.462425756454468,
      "epoch": 2.487111809962374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043892556649955354,
      "loss": 5.0556,
      "mean_token_accuracy": 0.21247505843639375,
      "num_tokens": 49445122.0,
      "step": 22145
    },
    {
      "entropy": 5.67967119216919,
      "epoch": 2.487673386870332,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004388980626582551,
      "loss": 5.1633,
      "mean_token_accuracy": 0.21492376625537873,
      "num_tokens": 49456276.0,
      "step": 22150
    },
    {
      "entropy": 5.597844791412354,
      "epoch": 2.4882349637782895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004388705535983788,
      "loss": 5.0631,
      "mean_token_accuracy": 0.213740798830986,
      "num_tokens": 49467346.0,
      "step": 22155
    },
    {
      "entropy": 5.535602331161499,
      "epoch": 2.488796540686247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004388430393208005,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22819118350744247,
      "num_tokens": 49477635.0,
      "step": 22160
    },
    {
      "entropy": 5.5113404273986815,
      "epoch": 2.4893581175942043,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004388155198263963,
      "loss": 5.0358,
      "mean_token_accuracy": 0.2150046706199646,
      "num_tokens": 49491060.0,
      "step": 22165
    },
    {
      "entropy": 5.568464136123657,
      "epoch": 2.489919694502162,
      "grad_norm": 1.0,
      "learning_rate": 0.0004387879951160424,
      "loss": 5.0042,
      "mean_token_accuracy": 0.22114009857177735,
      "num_tokens": 49502551.0,
      "step": 22170
    },
    {
      "entropy": 5.5355149269104,
      "epoch": 2.4904812714101197,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004387604651906152,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21626703143119813,
      "num_tokens": 49514505.0,
      "step": 22175
    },
    {
      "entropy": 5.484716558456421,
      "epoch": 2.4910428483180773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004387329300509913,
      "loss": 4.9465,
      "mean_token_accuracy": 0.2183131083846092,
      "num_tokens": 49526301.0,
      "step": 22180
    },
    {
      "entropy": 5.481555032730102,
      "epoch": 2.4916044252260345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043870538969804733,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22085696309804917,
      "num_tokens": 49536546.0,
      "step": 22185
    },
    {
      "entropy": 5.607558298110962,
      "epoch": 2.492166002133992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004386778441326603,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20982214957475662,
      "num_tokens": 49548046.0,
      "step": 22190
    },
    {
      "entropy": 5.507656908035278,
      "epoch": 2.49272757904195,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00043865029335570725,
      "loss": 4.959,
      "mean_token_accuracy": 0.22057689279317855,
      "num_tokens": 49560001.0,
      "step": 22195
    },
    {
      "entropy": 5.5165609359741214,
      "epoch": 2.4932891559499075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004386227373680655,
      "loss": 4.9638,
      "mean_token_accuracy": 0.22523604929447175,
      "num_tokens": 49571111.0,
      "step": 22200
    },
    {
      "entropy": 5.478017997741699,
      "epoch": 2.4938507328578647,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004385951761706122,
      "loss": 4.956,
      "mean_token_accuracy": 0.22220365554094315,
      "num_tokens": 49582888.0,
      "step": 22205
    },
    {
      "entropy": 5.469198322296142,
      "epoch": 2.4944123097658224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043856760976422506,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21900759488344193,
      "num_tokens": 49593710.0,
      "step": 22210
    },
    {
      "entropy": 5.570020961761474,
      "epoch": 2.49497388667378,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043854003814978174,
      "loss": 5.03,
      "mean_token_accuracy": 0.21788311302661895,
      "num_tokens": 49604263.0,
      "step": 22215
    },
    {
      "entropy": 5.679350852966309,
      "epoch": 2.4955354635817377,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043851246132816025,
      "loss": 5.1078,
      "mean_token_accuracy": 0.21397348195314408,
      "num_tokens": 49615152.0,
      "step": 22220
    },
    {
      "entropy": 5.610162305831909,
      "epoch": 2.496097040489695,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004384848793002385,
      "loss": 5.1058,
      "mean_token_accuracy": 0.211672081053257,
      "num_tokens": 49627778.0,
      "step": 22225
    },
    {
      "entropy": 5.542225646972656,
      "epoch": 2.4966586173976526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004384572920668949,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21819426268339157,
      "num_tokens": 49637701.0,
      "step": 22230
    },
    {
      "entropy": 5.5013261318206785,
      "epoch": 2.4972201943056103,
      "grad_norm": 1.125,
      "learning_rate": 0.0004384296996290076,
      "loss": 4.99,
      "mean_token_accuracy": 0.22697882503271102,
      "num_tokens": 49647680.0,
      "step": 22235
    },
    {
      "entropy": 5.620667409896851,
      "epoch": 2.4977817712135675,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043840210198745533,
      "loss": 5.1323,
      "mean_token_accuracy": 0.21319353431463242,
      "num_tokens": 49659419.0,
      "step": 22240
    },
    {
      "entropy": 5.5593671798706055,
      "epoch": 2.498343348121525,
      "grad_norm": 1.0,
      "learning_rate": 0.00043837449914311676,
      "loss": 5.0171,
      "mean_token_accuracy": 0.2224773049354553,
      "num_tokens": 49671131.0,
      "step": 22245
    },
    {
      "entropy": 5.543187475204467,
      "epoch": 2.498904925029483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043834689109687077,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22007375210523605,
      "num_tokens": 49681748.0,
      "step": 22250
    },
    {
      "entropy": 5.61725115776062,
      "epoch": 2.4994665019374405,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004383192778495964,
      "loss": 5.0163,
      "mean_token_accuracy": 0.22673007547855378,
      "num_tokens": 49693321.0,
      "step": 22255
    },
    {
      "entropy": 5.537936687469482,
      "epoch": 2.500028078845398,
      "grad_norm": 1.078125,
      "learning_rate": 0.000438291659402173,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21937822848558425,
      "num_tokens": 49704419.0,
      "step": 22260
    },
    {
      "entropy": 5.546604108810425,
      "epoch": 2.5005896557533553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043826403575547974,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21784052699804307,
      "num_tokens": 49715692.0,
      "step": 22265
    },
    {
      "entropy": 5.627329635620117,
      "epoch": 2.501151232661313,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004382364069103962,
      "loss": 5.0542,
      "mean_token_accuracy": 0.21116976737976073,
      "num_tokens": 49726324.0,
      "step": 22270
    },
    {
      "entropy": 5.504190683364868,
      "epoch": 2.5017128095692707,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043820877286780224,
      "loss": 4.9931,
      "mean_token_accuracy": 0.22282054275274277,
      "num_tokens": 49737069.0,
      "step": 22275
    },
    {
      "entropy": 5.541795253753662,
      "epoch": 2.502274386477228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004381811336285776,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21762404292821885,
      "num_tokens": 49748507.0,
      "step": 22280
    },
    {
      "entropy": 5.5483287334442135,
      "epoch": 2.5028359633851855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004381534891936024,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21375676691532136,
      "num_tokens": 49759017.0,
      "step": 22285
    },
    {
      "entropy": 5.60446605682373,
      "epoch": 2.503397540293143,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004381258395637568,
      "loss": 5.049,
      "mean_token_accuracy": 0.217348712682724,
      "num_tokens": 49771353.0,
      "step": 22290
    },
    {
      "entropy": 5.660924577713013,
      "epoch": 2.5039591172011004,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004380981847399212,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20720814019441605,
      "num_tokens": 49782337.0,
      "step": 22295
    },
    {
      "entropy": 5.579286670684814,
      "epoch": 2.504520694109058,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043807052472297626,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22454477548599244,
      "num_tokens": 49794212.0,
      "step": 22300
    },
    {
      "entropy": 5.484839820861817,
      "epoch": 2.5050822710170157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004380428595138024,
      "loss": 4.9866,
      "mean_token_accuracy": 0.22578197866678237,
      "num_tokens": 49804602.0,
      "step": 22305
    },
    {
      "entropy": 5.573648118972779,
      "epoch": 2.5056438479249734,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043801518911328074,
      "loss": 5.1475,
      "mean_token_accuracy": 0.20684706419706345,
      "num_tokens": 49816277.0,
      "step": 22310
    },
    {
      "entropy": 5.587388181686402,
      "epoch": 2.506205424832931,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043798751352229227,
      "loss": 5.0091,
      "mean_token_accuracy": 0.22245265990495683,
      "num_tokens": 49827851.0,
      "step": 22315
    },
    {
      "entropy": 5.5123227596282955,
      "epoch": 2.5067670017408883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043795983274171806,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2094222292304039,
      "num_tokens": 49840492.0,
      "step": 22320
    },
    {
      "entropy": 5.443208074569702,
      "epoch": 2.507328578648846,
      "grad_norm": 1.0,
      "learning_rate": 0.00043793214677243967,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22526382207870482,
      "num_tokens": 49851427.0,
      "step": 22325
    },
    {
      "entropy": 5.541797876358032,
      "epoch": 2.5078901555568036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379044556153384,
      "loss": 4.984,
      "mean_token_accuracy": 0.21604935228824615,
      "num_tokens": 49862883.0,
      "step": 22330
    },
    {
      "entropy": 5.533827638626098,
      "epoch": 2.508451732464761,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043787675927129625,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2135650858283043,
      "num_tokens": 49874120.0,
      "step": 22335
    },
    {
      "entropy": 5.655553913116455,
      "epoch": 2.5090133093727185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043784905774119483,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22358110696077346,
      "num_tokens": 49885961.0,
      "step": 22340
    },
    {
      "entropy": 5.501085805892944,
      "epoch": 2.509574886280676,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043782135102591626,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22271086424589157,
      "num_tokens": 49896821.0,
      "step": 22345
    },
    {
      "entropy": 5.42793345451355,
      "epoch": 2.510136463188634,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004377936391263427,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2195675939321518,
      "num_tokens": 49908148.0,
      "step": 22350
    },
    {
      "entropy": 5.538199329376221,
      "epoch": 2.5106980400965915,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043776592204335655,
      "loss": 5.0259,
      "mean_token_accuracy": 0.2198207199573517,
      "num_tokens": 49919487.0,
      "step": 22355
    },
    {
      "entropy": 5.523135423660278,
      "epoch": 2.5112596170045487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004377381997778403,
      "loss": 4.998,
      "mean_token_accuracy": 0.22348760515451432,
      "num_tokens": 49930722.0,
      "step": 22360
    },
    {
      "entropy": 5.604264163970948,
      "epoch": 2.5118211939125064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004377104723306767,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20871011018753052,
      "num_tokens": 49941917.0,
      "step": 22365
    },
    {
      "entropy": 5.611111259460449,
      "epoch": 2.512382770820464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004376827397027485,
      "loss": 5.0531,
      "mean_token_accuracy": 0.21181179583072662,
      "num_tokens": 49952681.0,
      "step": 22370
    },
    {
      "entropy": 5.5823136329650875,
      "epoch": 2.5129443477284212,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004376550018949388,
      "loss": 5.0374,
      "mean_token_accuracy": 0.22057308852672577,
      "num_tokens": 49965542.0,
      "step": 22375
    },
    {
      "entropy": 5.59934024810791,
      "epoch": 2.513505924636379,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043762725890813083,
      "loss": 5.0996,
      "mean_token_accuracy": 0.21146688610315323,
      "num_tokens": 49977506.0,
      "step": 22380
    },
    {
      "entropy": 5.563894844055175,
      "epoch": 2.5140675015443366,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004375995107432077,
      "loss": 5.0301,
      "mean_token_accuracy": 0.21665119826793672,
      "num_tokens": 49988674.0,
      "step": 22385
    },
    {
      "entropy": 5.542116737365722,
      "epoch": 2.5146290784522938,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043757175740105316,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2118400365114212,
      "num_tokens": 50001618.0,
      "step": 22390
    },
    {
      "entropy": 5.526704025268555,
      "epoch": 2.5151906553602514,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004375439988825508,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21965675801038742,
      "num_tokens": 50012556.0,
      "step": 22395
    },
    {
      "entropy": 5.594196796417236,
      "epoch": 2.515752232268209,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004375162351885845,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22241995632648467,
      "num_tokens": 50023399.0,
      "step": 22400
    },
    {
      "entropy": 5.478972673416138,
      "epoch": 2.5163138091761668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043748846632003823,
      "loss": 4.9742,
      "mean_token_accuracy": 0.22121930718421937,
      "num_tokens": 50035209.0,
      "step": 22405
    },
    {
      "entropy": 5.488674831390381,
      "epoch": 2.5168753860841244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043746069227779615,
      "loss": 5.031,
      "mean_token_accuracy": 0.2149332121014595,
      "num_tokens": 50046383.0,
      "step": 22410
    },
    {
      "entropy": 5.534080648422242,
      "epoch": 2.5174369629920816,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004374329130627427,
      "loss": 5.0419,
      "mean_token_accuracy": 0.22077756226062775,
      "num_tokens": 50058785.0,
      "step": 22415
    },
    {
      "entropy": 5.600060892105103,
      "epoch": 2.5179985399000393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043740512867576224,
      "loss": 5.032,
      "mean_token_accuracy": 0.21792866736650468,
      "num_tokens": 50069629.0,
      "step": 22420
    },
    {
      "entropy": 5.5588134765625,
      "epoch": 2.518560116807997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004373773391177395,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22136150598526,
      "num_tokens": 50080923.0,
      "step": 22425
    },
    {
      "entropy": 5.526035261154175,
      "epoch": 2.519121693715954,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004373495443895593,
      "loss": 4.9863,
      "mean_token_accuracy": 0.21923796087503433,
      "num_tokens": 50092133.0,
      "step": 22430
    },
    {
      "entropy": 5.458914709091187,
      "epoch": 2.519683270623912,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004373217444921066,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2202799364924431,
      "num_tokens": 50104001.0,
      "step": 22435
    },
    {
      "entropy": 5.5567680358886715,
      "epoch": 2.5202448475318695,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043729393942626665,
      "loss": 5.0393,
      "mean_token_accuracy": 0.22126491367816925,
      "num_tokens": 50115908.0,
      "step": 22440
    },
    {
      "entropy": 5.593733310699463,
      "epoch": 2.520806424439827,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043726612919292476,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21344430297613143,
      "num_tokens": 50127380.0,
      "step": 22445
    },
    {
      "entropy": 5.6346687316894535,
      "epoch": 2.521368001347785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004372383137929663,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22331000566482545,
      "num_tokens": 50138458.0,
      "step": 22450
    },
    {
      "entropy": 5.600961589813233,
      "epoch": 2.521929578255742,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043721049322727705,
      "loss": 5.1089,
      "mean_token_accuracy": 0.2105940490961075,
      "num_tokens": 50148814.0,
      "step": 22455
    },
    {
      "entropy": 5.6140241622924805,
      "epoch": 2.5224911551636997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004371826674967428,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20563278645277022,
      "num_tokens": 50159233.0,
      "step": 22460
    },
    {
      "entropy": 5.533326196670532,
      "epoch": 2.5230527320716574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043715483660224943,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21893193423748017,
      "num_tokens": 50170752.0,
      "step": 22465
    },
    {
      "entropy": 5.502572441101075,
      "epoch": 2.5236143089796146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004371270005446833,
      "loss": 4.998,
      "mean_token_accuracy": 0.21747928559780122,
      "num_tokens": 50183759.0,
      "step": 22470
    },
    {
      "entropy": 5.55999984741211,
      "epoch": 2.5241758858875722,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004370991593249305,
      "loss": 4.954,
      "mean_token_accuracy": 0.22240064144134522,
      "num_tokens": 50194091.0,
      "step": 22475
    },
    {
      "entropy": 5.5838817119598385,
      "epoch": 2.52473746279553,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043707131294387766,
      "loss": 5.0153,
      "mean_token_accuracy": 0.21740836054086685,
      "num_tokens": 50204591.0,
      "step": 22480
    },
    {
      "entropy": 5.485633182525635,
      "epoch": 2.525299039703487,
      "grad_norm": 1.0,
      "learning_rate": 0.0004370434614024113,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2222868412733078,
      "num_tokens": 50215809.0,
      "step": 22485
    },
    {
      "entropy": 5.526858901977539,
      "epoch": 2.525860616611445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043701560470141824,
      "loss": 5.0594,
      "mean_token_accuracy": 0.2137701228260994,
      "num_tokens": 50227038.0,
      "step": 22490
    },
    {
      "entropy": 5.565846300125122,
      "epoch": 2.5264221935194024,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004369877428417856,
      "loss": 4.9923,
      "mean_token_accuracy": 0.22204449772834778,
      "num_tokens": 50237001.0,
      "step": 22495
    },
    {
      "entropy": 5.516031551361084,
      "epoch": 2.52698377042736,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004369598758244003,
      "loss": 4.9507,
      "mean_token_accuracy": 0.22382864505052566,
      "num_tokens": 50248389.0,
      "step": 22500
    },
    {
      "entropy": 5.507412099838257,
      "epoch": 2.5275453473353178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043693200365014985,
      "loss": 4.9818,
      "mean_token_accuracy": 0.22292576879262924,
      "num_tokens": 50259676.0,
      "step": 22505
    },
    {
      "entropy": 5.658680438995361,
      "epoch": 2.528106924243275,
      "grad_norm": 1.125,
      "learning_rate": 0.00043690412631992154,
      "loss": 5.1128,
      "mean_token_accuracy": 0.21736584603786469,
      "num_tokens": 50270707.0,
      "step": 22510
    },
    {
      "entropy": 5.522495555877685,
      "epoch": 2.5286685011512327,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043687624383460306,
      "loss": 4.8702,
      "mean_token_accuracy": 0.2314695045351982,
      "num_tokens": 50282187.0,
      "step": 22515
    },
    {
      "entropy": 5.546749782562256,
      "epoch": 2.5292300780591903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043684835619508215,
      "loss": 5.0383,
      "mean_token_accuracy": 0.22235967069864274,
      "num_tokens": 50293672.0,
      "step": 22520
    },
    {
      "entropy": 5.492457914352417,
      "epoch": 2.5297916549671475,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004368204634022469,
      "loss": 4.9,
      "mean_token_accuracy": 0.22008581906557084,
      "num_tokens": 50305235.0,
      "step": 22525
    },
    {
      "entropy": 5.586936140060425,
      "epoch": 2.530353231875105,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004367925654569852,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21835442930459975,
      "num_tokens": 50316178.0,
      "step": 22530
    },
    {
      "entropy": 5.507815599441528,
      "epoch": 2.530914808783063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004367646623601856,
      "loss": 4.932,
      "mean_token_accuracy": 0.21777687668800355,
      "num_tokens": 50326713.0,
      "step": 22535
    },
    {
      "entropy": 5.570933485031128,
      "epoch": 2.5314763856910205,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004367367541127364,
      "loss": 5.1398,
      "mean_token_accuracy": 0.21396447569131852,
      "num_tokens": 50338648.0,
      "step": 22540
    },
    {
      "entropy": 5.586802625656128,
      "epoch": 2.532037962598978,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004367088407155261,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2221183806657791,
      "num_tokens": 50350134.0,
      "step": 22545
    },
    {
      "entropy": 5.5024772644042965,
      "epoch": 2.5325995395069354,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004366809221694437,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22921405285596846,
      "num_tokens": 50361101.0,
      "step": 22550
    },
    {
      "entropy": 5.429637289047241,
      "epoch": 2.533161116414893,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043665299847537805,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21311675161123275,
      "num_tokens": 50372376.0,
      "step": 22555
    },
    {
      "entropy": 5.547336149215698,
      "epoch": 2.5337226933228507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043662506963421815,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22261634916067125,
      "num_tokens": 50383494.0,
      "step": 22560
    },
    {
      "entropy": 5.600231695175171,
      "epoch": 2.534284270230808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004365971356468534,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21833306252956391,
      "num_tokens": 50394720.0,
      "step": 22565
    },
    {
      "entropy": 5.629021549224854,
      "epoch": 2.5348458471387656,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043656919651417314,
      "loss": 5.0668,
      "mean_token_accuracy": 0.2166610985994339,
      "num_tokens": 50407109.0,
      "step": 22570
    },
    {
      "entropy": 5.598659372329712,
      "epoch": 2.5354074240467233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000436541252237067,
      "loss": 5.1582,
      "mean_token_accuracy": 0.21158129572868348,
      "num_tokens": 50418664.0,
      "step": 22575
    },
    {
      "entropy": 5.517361259460449,
      "epoch": 2.5359690009546805,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004365133028164248,
      "loss": 4.9551,
      "mean_token_accuracy": 0.21850142776966094,
      "num_tokens": 50431069.0,
      "step": 22580
    },
    {
      "entropy": 5.577747917175293,
      "epoch": 2.536530577862638,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004364853482531363,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21996314078569412,
      "num_tokens": 50441484.0,
      "step": 22585
    },
    {
      "entropy": 5.498120975494385,
      "epoch": 2.537092154770596,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043645738854809177,
      "loss": 4.9574,
      "mean_token_accuracy": 0.22266175597906113,
      "num_tokens": 50452594.0,
      "step": 22590
    },
    {
      "entropy": 5.59347128868103,
      "epoch": 2.5376537316785535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004364294237021813,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21073924601078034,
      "num_tokens": 50463342.0,
      "step": 22595
    },
    {
      "entropy": 5.596143054962158,
      "epoch": 2.538215308586511,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043640145371629544,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20909852683544158,
      "num_tokens": 50475164.0,
      "step": 22600
    },
    {
      "entropy": 5.587312936782837,
      "epoch": 2.5387768854944683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004363734785913246,
      "loss": 5.0896,
      "mean_token_accuracy": 0.212952621281147,
      "num_tokens": 50486793.0,
      "step": 22605
    },
    {
      "entropy": 5.489442300796509,
      "epoch": 2.539338462402426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004363454983281597,
      "loss": 5.016,
      "mean_token_accuracy": 0.22251566350460053,
      "num_tokens": 50499202.0,
      "step": 22610
    },
    {
      "entropy": 5.671355724334717,
      "epoch": 2.5399000393103837,
      "grad_norm": 1.0,
      "learning_rate": 0.0004363175129276915,
      "loss": 5.1098,
      "mean_token_accuracy": 0.21088382452726365,
      "num_tokens": 50511801.0,
      "step": 22615
    },
    {
      "entropy": 5.53126859664917,
      "epoch": 2.540461616218341,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043628952239081113,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21160590201616286,
      "num_tokens": 50522918.0,
      "step": 22620
    },
    {
      "entropy": 5.490508556365967,
      "epoch": 2.5410231931262985,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043626152671840984,
      "loss": 4.9343,
      "mean_token_accuracy": 0.2198411449790001,
      "num_tokens": 50534027.0,
      "step": 22625
    },
    {
      "entropy": 5.53324179649353,
      "epoch": 2.541584770034256,
      "grad_norm": 1.03125,
      "learning_rate": 0.000436233525911379,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22342362999916077,
      "num_tokens": 50543814.0,
      "step": 22630
    },
    {
      "entropy": 5.488043403625488,
      "epoch": 2.542146346942214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004362055199706101,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22127336710691453,
      "num_tokens": 50553818.0,
      "step": 22635
    },
    {
      "entropy": 5.579548597335815,
      "epoch": 2.5427079238501715,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043617750889699496,
      "loss": 5.1117,
      "mean_token_accuracy": 0.20512445271015167,
      "num_tokens": 50564526.0,
      "step": 22640
    },
    {
      "entropy": 5.519158124923706,
      "epoch": 2.5432695007581287,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043614949269142546,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22106749415397645,
      "num_tokens": 50576410.0,
      "step": 22645
    },
    {
      "entropy": 5.550156021118164,
      "epoch": 2.5438310776660864,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043612147135479356,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22267754673957824,
      "num_tokens": 50586378.0,
      "step": 22650
    },
    {
      "entropy": 5.559008502960205,
      "epoch": 2.544392654574044,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004360934448879916,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21966009289026261,
      "num_tokens": 50597157.0,
      "step": 22655
    },
    {
      "entropy": 5.503187847137451,
      "epoch": 2.5449542314820013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004360654132919118,
      "loss": 4.9735,
      "mean_token_accuracy": 0.22161898910999298,
      "num_tokens": 50607390.0,
      "step": 22660
    },
    {
      "entropy": 5.499709224700927,
      "epoch": 2.545515808389959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004360373765674467,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22336798161268234,
      "num_tokens": 50618397.0,
      "step": 22665
    },
    {
      "entropy": 5.621217155456543,
      "epoch": 2.5460773852979166,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043600933471548927,
      "loss": 5.1439,
      "mean_token_accuracy": 0.21095731258392333,
      "num_tokens": 50631186.0,
      "step": 22670
    },
    {
      "entropy": 5.564108848571777,
      "epoch": 2.546638962205874,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004359812877369321,
      "loss": 5.0475,
      "mean_token_accuracy": 0.217476786673069,
      "num_tokens": 50642761.0,
      "step": 22675
    },
    {
      "entropy": 5.4776740074157715,
      "epoch": 2.5472005391138315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043595323563266815,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2289082258939743,
      "num_tokens": 50653374.0,
      "step": 22680
    },
    {
      "entropy": 5.482734346389771,
      "epoch": 2.547762116021789,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043592517840359093,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21800094693899155,
      "num_tokens": 50664331.0,
      "step": 22685
    },
    {
      "entropy": 5.580927515029908,
      "epoch": 2.548323692929747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004358971160505936,
      "loss": 5.0133,
      "mean_token_accuracy": 0.21730758249759674,
      "num_tokens": 50675011.0,
      "step": 22690
    },
    {
      "entropy": 5.494330406188965,
      "epoch": 2.5488852698377045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043586904857456963,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2272218406200409,
      "num_tokens": 50685139.0,
      "step": 22695
    },
    {
      "entropy": 5.553448915481567,
      "epoch": 2.5494468467456617,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043584097597641275,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21763667166233064,
      "num_tokens": 50696610.0,
      "step": 22700
    },
    {
      "entropy": 5.533038902282715,
      "epoch": 2.5500084236536193,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043581289825701683,
      "loss": 4.9846,
      "mean_token_accuracy": 0.22107479870319366,
      "num_tokens": 50708662.0,
      "step": 22705
    },
    {
      "entropy": 5.45192060470581,
      "epoch": 2.550570000561577,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004357848154172759,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22204663902521132,
      "num_tokens": 50719457.0,
      "step": 22710
    },
    {
      "entropy": 5.506832075119019,
      "epoch": 2.5511315774695342,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043575672745808404,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2207427591085434,
      "num_tokens": 50730038.0,
      "step": 22715
    },
    {
      "entropy": 5.562636661529541,
      "epoch": 2.551693154377492,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043572863438033556,
      "loss": 5.0052,
      "mean_token_accuracy": 0.2236795887351036,
      "num_tokens": 50742327.0,
      "step": 22720
    },
    {
      "entropy": 5.593198776245117,
      "epoch": 2.5522547312854496,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004357005361849251,
      "loss": 5.0575,
      "mean_token_accuracy": 0.2203540399670601,
      "num_tokens": 50753056.0,
      "step": 22725
    },
    {
      "entropy": 5.533463954925537,
      "epoch": 2.552816308193407,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043567243287274726,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21763788312673568,
      "num_tokens": 50763961.0,
      "step": 22730
    },
    {
      "entropy": 5.541670227050782,
      "epoch": 2.553377885101365,
      "grad_norm": 1.125,
      "learning_rate": 0.00043564432444469686,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23005380481481552,
      "num_tokens": 50774978.0,
      "step": 22735
    },
    {
      "entropy": 5.580276775360107,
      "epoch": 2.553939462009322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043561621090166885,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21515414416790007,
      "num_tokens": 50786198.0,
      "step": 22740
    },
    {
      "entropy": 5.496707916259766,
      "epoch": 2.5545010389172798,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004355880922445584,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21367162466049194,
      "num_tokens": 50797685.0,
      "step": 22745
    },
    {
      "entropy": 5.396542119979858,
      "epoch": 2.5550626158252374,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043555996847426077,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22792076021432878,
      "num_tokens": 50808518.0,
      "step": 22750
    },
    {
      "entropy": 5.526910400390625,
      "epoch": 2.5556241927331946,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043553183959167154,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21201574206352233,
      "num_tokens": 50819241.0,
      "step": 22755
    },
    {
      "entropy": 5.655047225952148,
      "epoch": 2.5561857696411523,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043550370559768614,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20905033648014068,
      "num_tokens": 50830334.0,
      "step": 22760
    },
    {
      "entropy": 5.673072099685669,
      "epoch": 2.55674734654911,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004354755664932006,
      "loss": 5.049,
      "mean_token_accuracy": 0.21922166794538497,
      "num_tokens": 50840738.0,
      "step": 22765
    },
    {
      "entropy": 5.606631565093994,
      "epoch": 2.557308923457067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043544742227911076,
      "loss": 5.1264,
      "mean_token_accuracy": 0.212349334359169,
      "num_tokens": 50853397.0,
      "step": 22770
    },
    {
      "entropy": 5.6041004180908205,
      "epoch": 2.557870500365025,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043541927295631276,
      "loss": 5.0199,
      "mean_token_accuracy": 0.2153725266456604,
      "num_tokens": 50865786.0,
      "step": 22775
    },
    {
      "entropy": 5.570644235610962,
      "epoch": 2.5584320772729825,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043539111852570296,
      "loss": 5.0346,
      "mean_token_accuracy": 0.222890605032444,
      "num_tokens": 50876029.0,
      "step": 22780
    },
    {
      "entropy": 5.5092438697814945,
      "epoch": 2.55899365418094,
      "grad_norm": 1.0,
      "learning_rate": 0.00043536295898817767,
      "loss": 4.957,
      "mean_token_accuracy": 0.22368922233581542,
      "num_tokens": 50886941.0,
      "step": 22785
    },
    {
      "entropy": 5.564006042480469,
      "epoch": 2.559555231088898,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004353347943446336,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21969368010759355,
      "num_tokens": 50898231.0,
      "step": 22790
    },
    {
      "entropy": 5.539107084274292,
      "epoch": 2.560116807996855,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043530662459596744,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21278483122587205,
      "num_tokens": 50908282.0,
      "step": 22795
    },
    {
      "entropy": 5.505239009857178,
      "epoch": 2.5606783849048127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004352784497430762,
      "loss": 4.9174,
      "mean_token_accuracy": 0.2328803926706314,
      "num_tokens": 50918462.0,
      "step": 22800
    },
    {
      "entropy": 5.519135618209839,
      "epoch": 2.5612399618127704,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043525026978685706,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21930775046348572,
      "num_tokens": 50929160.0,
      "step": 22805
    },
    {
      "entropy": 5.627083730697632,
      "epoch": 2.5618015387207276,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043522208472820706,
      "loss": 5.1802,
      "mean_token_accuracy": 0.2094150811433792,
      "num_tokens": 50943361.0,
      "step": 22810
    },
    {
      "entropy": 5.6335960865020756,
      "epoch": 2.5623631156286852,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004351938945680238,
      "loss": 5.0945,
      "mean_token_accuracy": 0.21598078310489655,
      "num_tokens": 50954515.0,
      "step": 22815
    },
    {
      "entropy": 5.651684999465942,
      "epoch": 2.562924692536643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004351656993072048,
      "loss": 5.0892,
      "mean_token_accuracy": 0.21218037903308867,
      "num_tokens": 50965237.0,
      "step": 22820
    },
    {
      "entropy": 5.513134717941284,
      "epoch": 2.5634862694446006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351374989466477,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2234373614192009,
      "num_tokens": 50976406.0,
      "step": 22825
    },
    {
      "entropy": 5.414370059967041,
      "epoch": 2.5640478463525582,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043510929348725065,
      "loss": 4.915,
      "mean_token_accuracy": 0.21979269981384278,
      "num_tokens": 50987214.0,
      "step": 22830
    },
    {
      "entropy": 5.53629150390625,
      "epoch": 2.5646094232605154,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043508108292991156,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21884544640779496,
      "num_tokens": 50997668.0,
      "step": 22835
    },
    {
      "entropy": 5.593557691574096,
      "epoch": 2.565171000168473,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004350528672755287,
      "loss": 4.9961,
      "mean_token_accuracy": 0.2256360575556755,
      "num_tokens": 51008358.0,
      "step": 22840
    },
    {
      "entropy": 5.544342279434204,
      "epoch": 2.5657325770764308,
      "grad_norm": 1.0,
      "learning_rate": 0.0004350246465250005,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2225613623857498,
      "num_tokens": 51019410.0,
      "step": 22845
    },
    {
      "entropy": 5.641515254974365,
      "epoch": 2.566294153984388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043499642067922544,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20992651283740998,
      "num_tokens": 51032307.0,
      "step": 22850
    },
    {
      "entropy": 5.698702621459961,
      "epoch": 2.5668557308923456,
      "grad_norm": 1.125,
      "learning_rate": 0.00043496818973910235,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21226277351379394,
      "num_tokens": 51043176.0,
      "step": 22855
    },
    {
      "entropy": 5.640099668502808,
      "epoch": 2.5674173078003033,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043493995370553,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21476795077323912,
      "num_tokens": 51053977.0,
      "step": 22860
    },
    {
      "entropy": 5.528575992584228,
      "epoch": 2.5679788847082605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043491171257940755,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21743019968271254,
      "num_tokens": 51064801.0,
      "step": 22865
    },
    {
      "entropy": 5.560040378570557,
      "epoch": 2.568540461616218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043488346636163405,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2315046340227127,
      "num_tokens": 51075400.0,
      "step": 22870
    },
    {
      "entropy": 5.6119688034057615,
      "epoch": 2.569102038524176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004348552150531091,
      "loss": 5.1163,
      "mean_token_accuracy": 0.21162362396717072,
      "num_tokens": 51087077.0,
      "step": 22875
    },
    {
      "entropy": 5.508121490478516,
      "epoch": 2.5696636154321335,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043482695865473207,
      "loss": 5.0153,
      "mean_token_accuracy": 0.21414380222558976,
      "num_tokens": 51099798.0,
      "step": 22880
    },
    {
      "entropy": 5.640077781677246,
      "epoch": 2.570225192340091,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004347986971674026,
      "loss": 5.1282,
      "mean_token_accuracy": 0.2086811199784279,
      "num_tokens": 51111880.0,
      "step": 22885
    },
    {
      "entropy": 5.573119974136352,
      "epoch": 2.5707867692480484,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043477043059202077,
      "loss": 5.0209,
      "mean_token_accuracy": 0.2125042423605919,
      "num_tokens": 51122358.0,
      "step": 22890
    },
    {
      "entropy": 5.564205932617187,
      "epoch": 2.571348346156006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004347421589294864,
      "loss": 5.0523,
      "mean_token_accuracy": 0.2183126300573349,
      "num_tokens": 51132775.0,
      "step": 22895
    },
    {
      "entropy": 5.560975551605225,
      "epoch": 2.5719099230639637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004347138821806997,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2126820519566536,
      "num_tokens": 51144974.0,
      "step": 22900
    },
    {
      "entropy": 5.579027986526489,
      "epoch": 2.572471499971921,
      "grad_norm": 1.0,
      "learning_rate": 0.000434685600346561,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2286905452609062,
      "num_tokens": 51154700.0,
      "step": 22905
    },
    {
      "entropy": 5.615908670425415,
      "epoch": 2.5730330768798786,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000434657313427971,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21188727170228958,
      "num_tokens": 51166910.0,
      "step": 22910
    },
    {
      "entropy": 5.5086753368377686,
      "epoch": 2.5735946537878363,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004346290214258301,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21917138993740082,
      "num_tokens": 51177620.0,
      "step": 22915
    },
    {
      "entropy": 5.605679607391357,
      "epoch": 2.574156230695794,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043460072434103934,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22266880422830582,
      "num_tokens": 51187894.0,
      "step": 22920
    },
    {
      "entropy": 5.546603393554688,
      "epoch": 2.5747178076037516,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004345724221744996,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21220346838235854,
      "num_tokens": 51199625.0,
      "step": 22925
    },
    {
      "entropy": 5.499326753616333,
      "epoch": 2.575279384511709,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043454411492711197,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21729636788368226,
      "num_tokens": 51211932.0,
      "step": 22930
    },
    {
      "entropy": 5.561268424987793,
      "epoch": 2.5758409614196665,
      "grad_norm": 1.0,
      "learning_rate": 0.00043451580259977793,
      "loss": 5.0771,
      "mean_token_accuracy": 0.2177476853132248,
      "num_tokens": 51223196.0,
      "step": 22935
    },
    {
      "entropy": 5.599836254119873,
      "epoch": 2.576402538327624,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043448748519339875,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21368786990642546,
      "num_tokens": 51234246.0,
      "step": 22940
    },
    {
      "entropy": 5.574683761596679,
      "epoch": 2.5769641152355813,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043445916270887626,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22076597213745117,
      "num_tokens": 51245056.0,
      "step": 22945
    },
    {
      "entropy": 5.590896320343018,
      "epoch": 2.577525692143539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004344308351471123,
      "loss": 5.0143,
      "mean_token_accuracy": 0.21549907922744752,
      "num_tokens": 51256513.0,
      "step": 22950
    },
    {
      "entropy": 5.4729650020599365,
      "epoch": 2.5780872690514967,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043440250250900854,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21958441585302352,
      "num_tokens": 51267609.0,
      "step": 22955
    },
    {
      "entropy": 5.602334213256836,
      "epoch": 2.578648845959454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004343741647954673,
      "loss": 5.1232,
      "mean_token_accuracy": 0.2179613769054413,
      "num_tokens": 51277940.0,
      "step": 22960
    },
    {
      "entropy": 5.660164308547974,
      "epoch": 2.5792104228674115,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004343458220073909,
      "loss": 5.102,
      "mean_token_accuracy": 0.2229011595249176,
      "num_tokens": 51288853.0,
      "step": 22965
    },
    {
      "entropy": 5.5715474605560305,
      "epoch": 2.579771999775369,
      "grad_norm": 1.125,
      "learning_rate": 0.0004343174741456817,
      "loss": 5.1209,
      "mean_token_accuracy": 0.2097961962223053,
      "num_tokens": 51299197.0,
      "step": 22970
    },
    {
      "entropy": 5.586898326873779,
      "epoch": 2.580333576683327,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004342891212112423,
      "loss": 5.1191,
      "mean_token_accuracy": 0.21090497970581054,
      "num_tokens": 51310821.0,
      "step": 22975
    },
    {
      "entropy": 5.676619052886963,
      "epoch": 2.5808951535912845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043426076320497557,
      "loss": 5.0478,
      "mean_token_accuracy": 0.20739950090646744,
      "num_tokens": 51321805.0,
      "step": 22980
    },
    {
      "entropy": 5.472146654129029,
      "epoch": 2.5814567304992417,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043423240012778437,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21939619928598403,
      "num_tokens": 51332647.0,
      "step": 22985
    },
    {
      "entropy": 5.458919954299927,
      "epoch": 2.5820183074071994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004342040319805717,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22436649054288865,
      "num_tokens": 51344715.0,
      "step": 22990
    },
    {
      "entropy": 5.614592266082764,
      "epoch": 2.582579884315157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000434175658764241,
      "loss": 5.1137,
      "mean_token_accuracy": 0.21550821214914323,
      "num_tokens": 51355481.0,
      "step": 22995
    },
    {
      "entropy": 5.563913583755493,
      "epoch": 2.5831414612231143,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043414728047969553,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22652142643928527,
      "num_tokens": 51367098.0,
      "step": 23000
    },
    {
      "entropy": 5.577368831634521,
      "epoch": 2.583703038131072,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043411889712783897,
      "loss": 5.13,
      "mean_token_accuracy": 0.21056273430585862,
      "num_tokens": 51379948.0,
      "step": 23005
    },
    {
      "entropy": 5.538464975357056,
      "epoch": 2.5842646150390296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043409050870957505,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2212616980075836,
      "num_tokens": 51391675.0,
      "step": 23010
    },
    {
      "entropy": 5.477495098114014,
      "epoch": 2.5848261919469873,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340621152258075,
      "loss": 4.912,
      "mean_token_accuracy": 0.22346910536289216,
      "num_tokens": 51401809.0,
      "step": 23015
    },
    {
      "entropy": 5.538278102874756,
      "epoch": 2.585387768854945,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043403371667744055,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21504226326942444,
      "num_tokens": 51413184.0,
      "step": 23020
    },
    {
      "entropy": 5.568682336807251,
      "epoch": 2.585949345762902,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004340053130653785,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21159550100564956,
      "num_tokens": 51424751.0,
      "step": 23025
    },
    {
      "entropy": 5.5206437587738035,
      "epoch": 2.58651092267086,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004339769043905255,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2205190032720566,
      "num_tokens": 51436550.0,
      "step": 23030
    },
    {
      "entropy": 5.554955911636353,
      "epoch": 2.5870724995788175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043394849065378624,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21681970357894897,
      "num_tokens": 51447691.0,
      "step": 23035
    },
    {
      "entropy": 5.540941762924194,
      "epoch": 2.5876340764867747,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00043392007185606533,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22031151950359346,
      "num_tokens": 51459715.0,
      "step": 23040
    },
    {
      "entropy": 5.537150478363037,
      "epoch": 2.5881956533947323,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004338916479982678,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22832642495632172,
      "num_tokens": 51470541.0,
      "step": 23045
    },
    {
      "entropy": 5.5268226146698,
      "epoch": 2.58875723030269,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043386321908129856,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22597559094429015,
      "num_tokens": 51480724.0,
      "step": 23050
    },
    {
      "entropy": 5.663091135025025,
      "epoch": 2.589318807210647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043383478510606276,
      "loss": 5.19,
      "mean_token_accuracy": 0.2076091080904007,
      "num_tokens": 51492204.0,
      "step": 23055
    },
    {
      "entropy": 5.572744560241699,
      "epoch": 2.589880384118605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004338063460734658,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21085387021303176,
      "num_tokens": 51503354.0,
      "step": 23060
    },
    {
      "entropy": 5.602397441864014,
      "epoch": 2.5904419610265625,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004337779019844132,
      "loss": 5.0515,
      "mean_token_accuracy": 0.2229481279850006,
      "num_tokens": 51513563.0,
      "step": 23065
    },
    {
      "entropy": 5.5762067317962645,
      "epoch": 2.59100353793452,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043374945283981066,
      "loss": 5.0095,
      "mean_token_accuracy": 0.21784208714962006,
      "num_tokens": 51524996.0,
      "step": 23070
    },
    {
      "entropy": 5.450859117507934,
      "epoch": 2.591565114842478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004337209986405639,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22941445112228392,
      "num_tokens": 51536822.0,
      "step": 23075
    },
    {
      "entropy": 5.494492244720459,
      "epoch": 2.592126691750435,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000433692539387579,
      "loss": 4.9918,
      "mean_token_accuracy": 0.22142852246761321,
      "num_tokens": 51547545.0,
      "step": 23080
    },
    {
      "entropy": 5.570556497573852,
      "epoch": 2.5926882686583927,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004336640750817621,
      "loss": 5.053,
      "mean_token_accuracy": 0.21796246021986007,
      "num_tokens": 51560721.0,
      "step": 23085
    },
    {
      "entropy": 5.507182598114014,
      "epoch": 2.5932498455663504,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043363560572401953,
      "loss": 4.8985,
      "mean_token_accuracy": 0.21892405450344085,
      "num_tokens": 51573226.0,
      "step": 23090
    },
    {
      "entropy": 5.553229236602784,
      "epoch": 2.5938114224743076,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043360713131525774,
      "loss": 5.035,
      "mean_token_accuracy": 0.21942913234233857,
      "num_tokens": 51584338.0,
      "step": 23095
    },
    {
      "entropy": 5.561807918548584,
      "epoch": 2.5943729993822653,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004335786518563834,
      "loss": 5.0017,
      "mean_token_accuracy": 0.2177712619304657,
      "num_tokens": 51593979.0,
      "step": 23100
    },
    {
      "entropy": 5.512984848022461,
      "epoch": 2.594934576290223,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004335501673483033,
      "loss": 4.975,
      "mean_token_accuracy": 0.22419024109840394,
      "num_tokens": 51604713.0,
      "step": 23105
    },
    {
      "entropy": 5.568858575820923,
      "epoch": 2.5954961531981806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043352167779192423,
      "loss": 5.0608,
      "mean_token_accuracy": 0.2224091187119484,
      "num_tokens": 51616034.0,
      "step": 23110
    },
    {
      "entropy": 5.454849004745483,
      "epoch": 2.5960577301061383,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043349318318815363,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21959041506052018,
      "num_tokens": 51627551.0,
      "step": 23115
    },
    {
      "entropy": 5.564099264144898,
      "epoch": 2.5966193070140955,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043346468353789853,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21768305599689483,
      "num_tokens": 51639625.0,
      "step": 23120
    },
    {
      "entropy": 5.535285615921021,
      "epoch": 2.597180883922053,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004334361788420664,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22503838688135147,
      "num_tokens": 51649918.0,
      "step": 23125
    },
    {
      "entropy": 5.498850345611572,
      "epoch": 2.597742460830011,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043340766910156497,
      "loss": 4.9762,
      "mean_token_accuracy": 0.23020596355199813,
      "num_tokens": 51659527.0,
      "step": 23130
    },
    {
      "entropy": 5.55068187713623,
      "epoch": 2.598304037737968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043337915431730195,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21441613733768464,
      "num_tokens": 51670704.0,
      "step": 23135
    },
    {
      "entropy": 5.5246885299682615,
      "epoch": 2.5988656146459257,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043335063449018515,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22032018303871154,
      "num_tokens": 51681606.0,
      "step": 23140
    },
    {
      "entropy": 5.593302297592163,
      "epoch": 2.5994271915538834,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004333221096211227,
      "loss": 5.0788,
      "mean_token_accuracy": 0.2186855912208557,
      "num_tokens": 51692592.0,
      "step": 23145
    },
    {
      "entropy": 5.552587795257568,
      "epoch": 2.599988768461841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043329357971102296,
      "loss": 5.0163,
      "mean_token_accuracy": 0.21461748480796813,
      "num_tokens": 51704990.0,
      "step": 23150
    },
    {
      "entropy": 5.587925958633423,
      "epoch": 2.6005503453697982,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043326504476079425,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2184203714132309,
      "num_tokens": 51715934.0,
      "step": 23155
    },
    {
      "entropy": 5.578255271911621,
      "epoch": 2.601111922277756,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004332365047713451,
      "loss": 5.1276,
      "mean_token_accuracy": 0.2120984047651291,
      "num_tokens": 51727314.0,
      "step": 23160
    },
    {
      "entropy": 5.580967760086059,
      "epoch": 2.6016734991857136,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004332079597435842,
      "loss": 4.923,
      "mean_token_accuracy": 0.22741990983486177,
      "num_tokens": 51738322.0,
      "step": 23165
    },
    {
      "entropy": 5.481619358062744,
      "epoch": 2.602235076093671,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004331794096784206,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22291965931653976,
      "num_tokens": 51749221.0,
      "step": 23170
    },
    {
      "entropy": 5.455865097045899,
      "epoch": 2.6027966530016284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004331508545767633,
      "loss": 4.988,
      "mean_token_accuracy": 0.21668024957180024,
      "num_tokens": 51760568.0,
      "step": 23175
    },
    {
      "entropy": 5.562003517150879,
      "epoch": 2.603358229909586,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004331222944395213,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21956198662519455,
      "num_tokens": 51772652.0,
      "step": 23180
    },
    {
      "entropy": 5.5797374725341795,
      "epoch": 2.6039198068175438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004330937292676042,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2236000657081604,
      "num_tokens": 51783739.0,
      "step": 23185
    },
    {
      "entropy": 5.521290874481201,
      "epoch": 2.604481383725501,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004330651590619214,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21787545830011368,
      "num_tokens": 51795521.0,
      "step": 23190
    },
    {
      "entropy": 5.457558441162109,
      "epoch": 2.6050429606334586,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004330365838233827,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2265913963317871,
      "num_tokens": 51805041.0,
      "step": 23195
    },
    {
      "entropy": 5.557255363464355,
      "epoch": 2.6056045375414163,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043300800355289786,
      "loss": 5.085,
      "mean_token_accuracy": 0.21942793279886247,
      "num_tokens": 51815157.0,
      "step": 23200
    },
    {
      "entropy": 5.638420104980469,
      "epoch": 2.606166114449374,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043297941825137684,
      "loss": 5.0934,
      "mean_token_accuracy": 0.21645734459161758,
      "num_tokens": 51826760.0,
      "step": 23205
    },
    {
      "entropy": 5.712314939498901,
      "epoch": 2.6067276913573316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043295082791972993,
      "loss": 5.1803,
      "mean_token_accuracy": 0.21076213717460632,
      "num_tokens": 51838018.0,
      "step": 23210
    },
    {
      "entropy": 5.645218801498413,
      "epoch": 2.607289268265289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043292223255886737,
      "loss": 5.0824,
      "mean_token_accuracy": 0.2121381714940071,
      "num_tokens": 51848217.0,
      "step": 23215
    },
    {
      "entropy": 5.543339824676513,
      "epoch": 2.6078508451732465,
      "grad_norm": 1.125,
      "learning_rate": 0.00043289363216969973,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22493074536323548,
      "num_tokens": 51857537.0,
      "step": 23220
    },
    {
      "entropy": 5.520798444747925,
      "epoch": 2.608412422081204,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043286502675313757,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2230652704834938,
      "num_tokens": 51867961.0,
      "step": 23225
    },
    {
      "entropy": 5.557815408706665,
      "epoch": 2.6089739989891614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004328364163100917,
      "loss": 5.1061,
      "mean_token_accuracy": 0.20990960896015168,
      "num_tokens": 51879868.0,
      "step": 23230
    },
    {
      "entropy": 5.5892143726348875,
      "epoch": 2.609535575897119,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004328078008414731,
      "loss": 5.0499,
      "mean_token_accuracy": 0.21070678234100343,
      "num_tokens": 51890849.0,
      "step": 23235
    },
    {
      "entropy": 5.544381761550904,
      "epoch": 2.6100971528050767,
      "grad_norm": 1.140625,
      "learning_rate": 0.000432779180348193,
      "loss": 4.9979,
      "mean_token_accuracy": 0.22184484899044038,
      "num_tokens": 51901193.0,
      "step": 23240
    },
    {
      "entropy": 5.5727824687957765,
      "epoch": 2.6106587297130344,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043275055483116254,
      "loss": 5.0016,
      "mean_token_accuracy": 0.2187783032655716,
      "num_tokens": 51912551.0,
      "step": 23245
    },
    {
      "entropy": 5.664476585388184,
      "epoch": 2.6112203066209916,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004327219242912933,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1992790326476097,
      "num_tokens": 51923970.0,
      "step": 23250
    },
    {
      "entropy": 5.46552152633667,
      "epoch": 2.6117818835289492,
      "grad_norm": 1.0,
      "learning_rate": 0.0004326932887294967,
      "loss": 4.8935,
      "mean_token_accuracy": 0.22211546897888185,
      "num_tokens": 51936667.0,
      "step": 23255
    },
    {
      "entropy": 5.5369771957397464,
      "epoch": 2.612343460436907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004326646481466848,
      "loss": 4.9219,
      "mean_token_accuracy": 0.2201874375343323,
      "num_tokens": 51947479.0,
      "step": 23260
    },
    {
      "entropy": 5.510453748703003,
      "epoch": 2.6129050373448646,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004326360025437692,
      "loss": 5.0725,
      "mean_token_accuracy": 0.22357355952262878,
      "num_tokens": 51958669.0,
      "step": 23265
    },
    {
      "entropy": 5.488300514221192,
      "epoch": 2.613466614252822,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043260735192166217,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23000915050506593,
      "num_tokens": 51969273.0,
      "step": 23270
    },
    {
      "entropy": 5.5921101570129395,
      "epoch": 2.6140281911607794,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000432578696281276,
      "loss": 5.07,
      "mean_token_accuracy": 0.21986337900161743,
      "num_tokens": 51981220.0,
      "step": 23275
    },
    {
      "entropy": 5.6041947364807125,
      "epoch": 2.614589768068737,
      "grad_norm": 1.0625,
      "learning_rate": 0.000432550035623523,
      "loss": 5.0725,
      "mean_token_accuracy": 0.21732703745365142,
      "num_tokens": 51992768.0,
      "step": 23280
    },
    {
      "entropy": 5.551333665847778,
      "epoch": 2.6151513449766943,
      "grad_norm": 1.0,
      "learning_rate": 0.0004325213699493157,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21327520161867142,
      "num_tokens": 52002754.0,
      "step": 23285
    },
    {
      "entropy": 5.589683103561401,
      "epoch": 2.615712921884652,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004324926992595669,
      "loss": 4.952,
      "mean_token_accuracy": 0.22488492131233215,
      "num_tokens": 52014503.0,
      "step": 23290
    },
    {
      "entropy": 5.537550735473633,
      "epoch": 2.6162744987926096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004324640235551895,
      "loss": 5.0492,
      "mean_token_accuracy": 0.21802946627140046,
      "num_tokens": 52026151.0,
      "step": 23295
    },
    {
      "entropy": 5.516232967376709,
      "epoch": 2.6168360757005673,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004324353428370965,
      "loss": 5.02,
      "mean_token_accuracy": 0.21305418163537979,
      "num_tokens": 52037704.0,
      "step": 23300
    },
    {
      "entropy": 5.5594230651855465,
      "epoch": 2.617397652608525,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004324066571062012,
      "loss": 5.0102,
      "mean_token_accuracy": 0.22184473574161528,
      "num_tokens": 52049326.0,
      "step": 23305
    },
    {
      "entropy": 5.608141088485718,
      "epoch": 2.617959229516482,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004323779663634168,
      "loss": 5.02,
      "mean_token_accuracy": 0.2220958724617958,
      "num_tokens": 52060030.0,
      "step": 23310
    },
    {
      "entropy": 5.654547023773193,
      "epoch": 2.61852080642444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000432349270609657,
      "loss": 5.1553,
      "mean_token_accuracy": 0.21776966899633407,
      "num_tokens": 52071744.0,
      "step": 23315
    },
    {
      "entropy": 5.610279512405396,
      "epoch": 2.6190823833323975,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043232056984583527,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21775563806295395,
      "num_tokens": 52082443.0,
      "step": 23320
    },
    {
      "entropy": 5.496397066116333,
      "epoch": 2.6196439602403547,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004322918640728656,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21747128367424012,
      "num_tokens": 52094211.0,
      "step": 23325
    },
    {
      "entropy": 5.513851690292358,
      "epoch": 2.6202055371483124,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043226315329166203,
      "loss": 4.9392,
      "mean_token_accuracy": 0.22219696044921874,
      "num_tokens": 52104802.0,
      "step": 23330
    },
    {
      "entropy": 5.611351776123047,
      "epoch": 2.62076711405627,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043223443750313873,
      "loss": 5.0553,
      "mean_token_accuracy": 0.21710754185914993,
      "num_tokens": 52115459.0,
      "step": 23335
    },
    {
      "entropy": 5.544198799133301,
      "epoch": 2.6213286909642277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004322057167082099,
      "loss": 5.08,
      "mean_token_accuracy": 0.2133201152086258,
      "num_tokens": 52125572.0,
      "step": 23340
    },
    {
      "entropy": 5.507091760635376,
      "epoch": 2.621890267872185,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043217699090779006,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21805850118398667,
      "num_tokens": 52137405.0,
      "step": 23345
    },
    {
      "entropy": 5.566605710983277,
      "epoch": 2.6224518447801426,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043214826010279383,
      "loss": 5.0949,
      "mean_token_accuracy": 0.20859762132167817,
      "num_tokens": 52148131.0,
      "step": 23350
    },
    {
      "entropy": 5.56457839012146,
      "epoch": 2.6230134216881003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004321195242941361,
      "loss": 4.9999,
      "mean_token_accuracy": 0.21604792177677154,
      "num_tokens": 52159361.0,
      "step": 23355
    },
    {
      "entropy": 5.549569034576416,
      "epoch": 2.623574998596058,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043209078348273175,
      "loss": 4.9965,
      "mean_token_accuracy": 0.21327283680438996,
      "num_tokens": 52170521.0,
      "step": 23360
    },
    {
      "entropy": 5.630928802490234,
      "epoch": 2.624136575504015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004320620376694959,
      "loss": 5.0993,
      "mean_token_accuracy": 0.21744538396596907,
      "num_tokens": 52182069.0,
      "step": 23365
    },
    {
      "entropy": 5.46106448173523,
      "epoch": 2.624698152411973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043203328685534396,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22708016484975815,
      "num_tokens": 52191867.0,
      "step": 23370
    },
    {
      "entropy": 5.612447071075439,
      "epoch": 2.6252597293199305,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004320045310411912,
      "loss": 5.1735,
      "mean_token_accuracy": 0.20332657396793366,
      "num_tokens": 52204511.0,
      "step": 23375
    },
    {
      "entropy": 5.585911846160888,
      "epoch": 2.6258213062278877,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043197577022795327,
      "loss": 4.9638,
      "mean_token_accuracy": 0.22166026681661605,
      "num_tokens": 52214276.0,
      "step": 23380
    },
    {
      "entropy": 5.567514371871948,
      "epoch": 2.6263828831358453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004319470044165459,
      "loss": 5.0141,
      "mean_token_accuracy": 0.2231723189353943,
      "num_tokens": 52225388.0,
      "step": 23385
    },
    {
      "entropy": 5.524701070785523,
      "epoch": 2.626944460043803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004319182336078852,
      "loss": 4.91,
      "mean_token_accuracy": 0.22930872589349746,
      "num_tokens": 52235300.0,
      "step": 23390
    },
    {
      "entropy": 5.600985383987426,
      "epoch": 2.6275060369517607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000431889457802887,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20827312618494034,
      "num_tokens": 52248392.0,
      "step": 23395
    },
    {
      "entropy": 5.615060091018677,
      "epoch": 2.6280676138597183,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004318606770024675,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21469537913799286,
      "num_tokens": 52259839.0,
      "step": 23400
    },
    {
      "entropy": 5.611004304885864,
      "epoch": 2.6286291907676755,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004318318912075433,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21231815069913865,
      "num_tokens": 52271195.0,
      "step": 23405
    },
    {
      "entropy": 5.489296245574951,
      "epoch": 2.629190767675633,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004318031004190308,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21798120588064193,
      "num_tokens": 52282445.0,
      "step": 23410
    },
    {
      "entropy": 5.591678762435913,
      "epoch": 2.629752344583591,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043177430463784685,
      "loss": 5.015,
      "mean_token_accuracy": 0.21478280127048494,
      "num_tokens": 52294144.0,
      "step": 23415
    },
    {
      "entropy": 5.545351171493531,
      "epoch": 2.630313921491548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043174550386490816,
      "loss": 5.015,
      "mean_token_accuracy": 0.21963077485561372,
      "num_tokens": 52304510.0,
      "step": 23420
    },
    {
      "entropy": 5.518340063095093,
      "epoch": 2.6308754983995057,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004317166981011319,
      "loss": 4.9525,
      "mean_token_accuracy": 0.22122060805559157,
      "num_tokens": 52314507.0,
      "step": 23425
    },
    {
      "entropy": 5.459598255157471,
      "epoch": 2.6314370753074634,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043168788734743516,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22722748816013336,
      "num_tokens": 52325104.0,
      "step": 23430
    },
    {
      "entropy": 5.617460584640503,
      "epoch": 2.631998652215421,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043165907160473525,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21080611199140548,
      "num_tokens": 52336037.0,
      "step": 23435
    },
    {
      "entropy": 5.555622863769531,
      "epoch": 2.6325602291233783,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004316302508739499,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2255492702126503,
      "num_tokens": 52346901.0,
      "step": 23440
    },
    {
      "entropy": 5.540788316726685,
      "epoch": 2.633121806031336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004316014251559965,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2238825261592865,
      "num_tokens": 52359159.0,
      "step": 23445
    },
    {
      "entropy": 5.506995916366577,
      "epoch": 2.6336833829392936,
      "grad_norm": 1.125,
      "learning_rate": 0.00043157259445179304,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2237806349992752,
      "num_tokens": 52369274.0,
      "step": 23450
    },
    {
      "entropy": 5.459559535980224,
      "epoch": 2.6342449598472513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004315437587622574,
      "loss": 4.94,
      "mean_token_accuracy": 0.2214307278394699,
      "num_tokens": 52381360.0,
      "step": 23455
    },
    {
      "entropy": 5.527035188674927,
      "epoch": 2.6348065367552085,
      "grad_norm": 1.125,
      "learning_rate": 0.00043151491808830777,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21505490839481353,
      "num_tokens": 52391550.0,
      "step": 23460
    },
    {
      "entropy": 5.530500030517578,
      "epoch": 2.635368113663166,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043148607243086255,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22150223553180695,
      "num_tokens": 52401339.0,
      "step": 23465
    },
    {
      "entropy": 5.586437177658081,
      "epoch": 2.635929690571124,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004314572217908399,
      "loss": 5.0283,
      "mean_token_accuracy": 0.21825418025255203,
      "num_tokens": 52412062.0,
      "step": 23470
    },
    {
      "entropy": 5.585273265838623,
      "epoch": 2.636491267479081,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004314283661691587,
      "loss": 5.0326,
      "mean_token_accuracy": 0.21393507868051528,
      "num_tokens": 52422524.0,
      "step": 23475
    },
    {
      "entropy": 5.614405488967895,
      "epoch": 2.6370528443870387,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004313995055667377,
      "loss": 5.068,
      "mean_token_accuracy": 0.2178643077611923,
      "num_tokens": 52433422.0,
      "step": 23480
    },
    {
      "entropy": 5.578816556930542,
      "epoch": 2.6376144212949963,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004313706399844958,
      "loss": 5.0789,
      "mean_token_accuracy": 0.21298659294843675,
      "num_tokens": 52443879.0,
      "step": 23485
    },
    {
      "entropy": 5.5811450481414795,
      "epoch": 2.638175998202954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000431341769423352,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21747676134109498,
      "num_tokens": 52455610.0,
      "step": 23490
    },
    {
      "entropy": 5.626401281356811,
      "epoch": 2.6387375751109117,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004313128938842257,
      "loss": 5.0799,
      "mean_token_accuracy": 0.2214831992983818,
      "num_tokens": 52467392.0,
      "step": 23495
    },
    {
      "entropy": 5.5502886295318605,
      "epoch": 2.639299152018869,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043128401336803616,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22276388704776764,
      "num_tokens": 52478812.0,
      "step": 23500
    },
    {
      "entropy": 5.579553079605103,
      "epoch": 2.6398607289268265,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043125512787570305,
      "loss": 5.0509,
      "mean_token_accuracy": 0.2240892991423607,
      "num_tokens": 52488837.0,
      "step": 23505
    },
    {
      "entropy": 5.60763955116272,
      "epoch": 2.640422305834784,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004312262374081461,
      "loss": 5.0987,
      "mean_token_accuracy": 0.21362500339746476,
      "num_tokens": 52501656.0,
      "step": 23510
    },
    {
      "entropy": 5.4945769786834715,
      "epoch": 2.6409838827427414,
      "grad_norm": 1.0,
      "learning_rate": 0.00043119734196628507,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2174015685915947,
      "num_tokens": 52512280.0,
      "step": 23515
    },
    {
      "entropy": 5.529523229598999,
      "epoch": 2.641545459650699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043116844155104025,
      "loss": 5.0101,
      "mean_token_accuracy": 0.2216639831662178,
      "num_tokens": 52523664.0,
      "step": 23520
    },
    {
      "entropy": 5.567110395431518,
      "epoch": 2.6421070365586568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004311395361633316,
      "loss": 5.056,
      "mean_token_accuracy": 0.21190822571516038,
      "num_tokens": 52534913.0,
      "step": 23525
    },
    {
      "entropy": 5.480639886856079,
      "epoch": 2.6426686134666144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004311106258040795,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22213456630706788,
      "num_tokens": 52545576.0,
      "step": 23530
    },
    {
      "entropy": 5.642828416824341,
      "epoch": 2.643230190374572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043108171047420463,
      "loss": 5.0533,
      "mean_token_accuracy": 0.21271274387836456,
      "num_tokens": 52556136.0,
      "step": 23535
    },
    {
      "entropy": 5.610493755340576,
      "epoch": 2.6437917672825293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004310527901746275,
      "loss": 5.0603,
      "mean_token_accuracy": 0.21230973303318024,
      "num_tokens": 52567425.0,
      "step": 23540
    },
    {
      "entropy": 5.511486005783081,
      "epoch": 2.644353344190487,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431023864906269,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21605036407709122,
      "num_tokens": 52578040.0,
      "step": 23545
    },
    {
      "entropy": 5.55917534828186,
      "epoch": 2.6449149210984446,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043099493467005015,
      "loss": 4.9082,
      "mean_token_accuracy": 0.22560404390096664,
      "num_tokens": 52588261.0,
      "step": 23550
    },
    {
      "entropy": 5.54492449760437,
      "epoch": 2.645476498006402,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000430965999466892,
      "loss": 4.966,
      "mean_token_accuracy": 0.2201348900794983,
      "num_tokens": 52599522.0,
      "step": 23555
    },
    {
      "entropy": 5.538984394073486,
      "epoch": 2.6460380749143595,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043093705929771604,
      "loss": 5.1385,
      "mean_token_accuracy": 0.21140238493680955,
      "num_tokens": 52610244.0,
      "step": 23560
    },
    {
      "entropy": 5.627888107299805,
      "epoch": 2.646599651822317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004309081141634436,
      "loss": 5.0895,
      "mean_token_accuracy": 0.21826153248548508,
      "num_tokens": 52621745.0,
      "step": 23565
    },
    {
      "entropy": 5.613685846328735,
      "epoch": 2.6471612287302744,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043087916406499635,
      "loss": 5.0618,
      "mean_token_accuracy": 0.2141012281179428,
      "num_tokens": 52633089.0,
      "step": 23570
    },
    {
      "entropy": 5.596077537536621,
      "epoch": 2.647722805638232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043085020900329604,
      "loss": 5.0565,
      "mean_token_accuracy": 0.21874991208314895,
      "num_tokens": 52644911.0,
      "step": 23575
    },
    {
      "entropy": 5.448034811019897,
      "epoch": 2.6482843825461897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043082124897926463,
      "loss": 4.8642,
      "mean_token_accuracy": 0.2233882576227188,
      "num_tokens": 52656958.0,
      "step": 23580
    },
    {
      "entropy": 5.604281187057495,
      "epoch": 2.6488459594541474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004307922839938242,
      "loss": 5.0932,
      "mean_token_accuracy": 0.21164610683918,
      "num_tokens": 52668988.0,
      "step": 23585
    },
    {
      "entropy": 5.5644083499908445,
      "epoch": 2.649407536362105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043076331404789703,
      "loss": 5.0098,
      "mean_token_accuracy": 0.21681014448404312,
      "num_tokens": 52679948.0,
      "step": 23590
    },
    {
      "entropy": 5.553695869445801,
      "epoch": 2.6499691132700622,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004307343391424055,
      "loss": 4.967,
      "mean_token_accuracy": 0.22045273184776307,
      "num_tokens": 52690249.0,
      "step": 23595
    },
    {
      "entropy": 5.581438493728638,
      "epoch": 2.65053069017802,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004307053592782723,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2141401782631874,
      "num_tokens": 52702766.0,
      "step": 23600
    },
    {
      "entropy": 5.609477424621582,
      "epoch": 2.6510922670859776,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043067637445642004,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21933826357126235,
      "num_tokens": 52714252.0,
      "step": 23605
    },
    {
      "entropy": 5.605251884460449,
      "epoch": 2.651653843993935,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043064738467777166,
      "loss": 5.0967,
      "mean_token_accuracy": 0.21278856843709945,
      "num_tokens": 52724748.0,
      "step": 23610
    },
    {
      "entropy": 5.471751403808594,
      "epoch": 2.6522154209018924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043061838994325007,
      "loss": 5.0627,
      "mean_token_accuracy": 0.21412119567394255,
      "num_tokens": 52735636.0,
      "step": 23615
    },
    {
      "entropy": 5.662180376052857,
      "epoch": 2.65277699780985,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043058939025377867,
      "loss": 5.1787,
      "mean_token_accuracy": 0.2093760699033737,
      "num_tokens": 52747812.0,
      "step": 23620
    },
    {
      "entropy": 5.570288515090942,
      "epoch": 2.6533385747178078,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004305603856102808,
      "loss": 4.9709,
      "mean_token_accuracy": 0.2312754601240158,
      "num_tokens": 52759322.0,
      "step": 23625
    },
    {
      "entropy": 5.555832433700561,
      "epoch": 2.6539001516257654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004305313760136799,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22095671743154527,
      "num_tokens": 52770238.0,
      "step": 23630
    },
    {
      "entropy": 5.53506326675415,
      "epoch": 2.6544617285337226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043050236146489955,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21703654527664185,
      "num_tokens": 52781243.0,
      "step": 23635
    },
    {
      "entropy": 5.538610219955444,
      "epoch": 2.6550233054416803,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004304733419648638,
      "loss": 4.9616,
      "mean_token_accuracy": 0.2216275990009308,
      "num_tokens": 52791341.0,
      "step": 23640
    },
    {
      "entropy": 5.589720153808594,
      "epoch": 2.655584882349638,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043044431751449656,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21517040133476256,
      "num_tokens": 52803322.0,
      "step": 23645
    },
    {
      "entropy": 5.485665225982666,
      "epoch": 2.656146459257595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004304152881147219,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22706061899662017,
      "num_tokens": 52813561.0,
      "step": 23650
    },
    {
      "entropy": 5.565480375289917,
      "epoch": 2.656708036165553,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004303862537664642,
      "loss": 5.0717,
      "mean_token_accuracy": 0.21575930267572402,
      "num_tokens": 52824150.0,
      "step": 23655
    },
    {
      "entropy": 5.45995454788208,
      "epoch": 2.6572696130735105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000430357214470648,
      "loss": 4.875,
      "mean_token_accuracy": 0.22676651030778885,
      "num_tokens": 52835013.0,
      "step": 23660
    },
    {
      "entropy": 5.578522491455078,
      "epoch": 2.6578311899814677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004303281702281977,
      "loss": 5.1009,
      "mean_token_accuracy": 0.2184455156326294,
      "num_tokens": 52846682.0,
      "step": 23665
    },
    {
      "entropy": 5.502072525024414,
      "epoch": 2.6583927668894254,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043029912104003827,
      "loss": 4.9648,
      "mean_token_accuracy": 0.21064316630363464,
      "num_tokens": 52857561.0,
      "step": 23670
    },
    {
      "entropy": 5.594608449935913,
      "epoch": 2.658954343797383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043027006690709457,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22086837738752366,
      "num_tokens": 52868833.0,
      "step": 23675
    },
    {
      "entropy": 5.630242252349854,
      "epoch": 2.6595159207053407,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004302410078302918,
      "loss": 5.0683,
      "mean_token_accuracy": 0.213382788002491,
      "num_tokens": 52879725.0,
      "step": 23680
    },
    {
      "entropy": 5.597429990768433,
      "epoch": 2.6600774976132984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043021194381055505,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20995192974805832,
      "num_tokens": 52890585.0,
      "step": 23685
    },
    {
      "entropy": 5.607091999053955,
      "epoch": 2.6606390745212556,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043018287484880986,
      "loss": 5.1245,
      "mean_token_accuracy": 0.21159467995166778,
      "num_tokens": 52902816.0,
      "step": 23690
    },
    {
      "entropy": 5.650960206985474,
      "epoch": 2.6612006514292132,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004301538009459817,
      "loss": 5.1369,
      "mean_token_accuracy": 0.21640197038650513,
      "num_tokens": 52914523.0,
      "step": 23695
    },
    {
      "entropy": 5.588558197021484,
      "epoch": 2.661762228337171,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004301247221029964,
      "loss": 5.0632,
      "mean_token_accuracy": 0.2190348669886589,
      "num_tokens": 52924953.0,
      "step": 23700
    },
    {
      "entropy": 5.4600865840911865,
      "epoch": 2.662323805245128,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043009563832077977,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22487605214118958,
      "num_tokens": 52935435.0,
      "step": 23705
    },
    {
      "entropy": 5.5230645656585695,
      "epoch": 2.662885382153086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004300665496002578,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2229163885116577,
      "num_tokens": 52946969.0,
      "step": 23710
    },
    {
      "entropy": 5.560060501098633,
      "epoch": 2.6634469590610435,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043003745594235687,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21471706479787828,
      "num_tokens": 52958634.0,
      "step": 23715
    },
    {
      "entropy": 5.524260330200195,
      "epoch": 2.664008535969001,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004300083573480031,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22573625147342682,
      "num_tokens": 52968412.0,
      "step": 23720
    },
    {
      "entropy": 5.487481546401978,
      "epoch": 2.6645701128769588,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042997925381812313,
      "loss": 4.9237,
      "mean_token_accuracy": 0.21485271006822587,
      "num_tokens": 52978653.0,
      "step": 23725
    },
    {
      "entropy": 5.551891231536866,
      "epoch": 2.665131689784916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042995014535364363,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22356362640857697,
      "num_tokens": 52988518.0,
      "step": 23730
    },
    {
      "entropy": 5.5396963596344,
      "epoch": 2.6656932666928737,
      "grad_norm": 1.0,
      "learning_rate": 0.0004299210319554914,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21760943084955214,
      "num_tokens": 52999271.0,
      "step": 23735
    },
    {
      "entropy": 5.569735527038574,
      "epoch": 2.6662548436008313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042989191362459347,
      "loss": 5.0476,
      "mean_token_accuracy": 0.2237466588616371,
      "num_tokens": 53010530.0,
      "step": 23740
    },
    {
      "entropy": 5.546817493438721,
      "epoch": 2.6668164205087885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004298627903618769,
      "loss": 5.0358,
      "mean_token_accuracy": 0.21564223617315292,
      "num_tokens": 53021631.0,
      "step": 23745
    },
    {
      "entropy": 5.520834064483642,
      "epoch": 2.667377997416746,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042983366216826907,
      "loss": 4.9819,
      "mean_token_accuracy": 0.21952608972787857,
      "num_tokens": 53033763.0,
      "step": 23750
    },
    {
      "entropy": 5.5309148788452145,
      "epoch": 2.667939574324704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004298045290446973,
      "loss": 5.048,
      "mean_token_accuracy": 0.21576713025569916,
      "num_tokens": 53045839.0,
      "step": 23755
    },
    {
      "entropy": 5.584755563735962,
      "epoch": 2.668501151232661,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042977539099208937,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21665347814559938,
      "num_tokens": 53057295.0,
      "step": 23760
    },
    {
      "entropy": 5.684488153457641,
      "epoch": 2.6690627281406187,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042974624801137283,
      "loss": 5.2471,
      "mean_token_accuracy": 0.2045261174440384,
      "num_tokens": 53068595.0,
      "step": 23765
    },
    {
      "entropy": 5.552354860305786,
      "epoch": 2.6696243050485764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042971710010347594,
      "loss": 4.975,
      "mean_token_accuracy": 0.22055513709783553,
      "num_tokens": 53080170.0,
      "step": 23770
    },
    {
      "entropy": 5.560001277923584,
      "epoch": 2.670185881956534,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042968794726932634,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22216688841581345,
      "num_tokens": 53090894.0,
      "step": 23775
    },
    {
      "entropy": 5.566949319839478,
      "epoch": 2.6707474588644917,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004296587895098527,
      "loss": 5.1084,
      "mean_token_accuracy": 0.21711268573999404,
      "num_tokens": 53103732.0,
      "step": 23780
    },
    {
      "entropy": 5.582149076461792,
      "epoch": 2.671309035772449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042962962682598316,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21855565756559373,
      "num_tokens": 53115237.0,
      "step": 23785
    },
    {
      "entropy": 5.5595063209533695,
      "epoch": 2.6718706126804066,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042960045921864627,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2214532434940338,
      "num_tokens": 53125638.0,
      "step": 23790
    },
    {
      "entropy": 5.500962400436402,
      "epoch": 2.6724321895883643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004295712866887709,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22397712022066116,
      "num_tokens": 53136153.0,
      "step": 23795
    },
    {
      "entropy": 5.536880254745483,
      "epoch": 2.6729937664963215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042954210923728565,
      "loss": 5.0494,
      "mean_token_accuracy": 0.21690435260534285,
      "num_tokens": 53146562.0,
      "step": 23800
    },
    {
      "entropy": 5.683835887908936,
      "epoch": 2.673555343404279,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004295129268651198,
      "loss": 5.1378,
      "mean_token_accuracy": 0.21323951780796052,
      "num_tokens": 53158460.0,
      "step": 23805
    },
    {
      "entropy": 5.5752012729644775,
      "epoch": 2.674116920312237,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004294837395732025,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22156913578510284,
      "num_tokens": 53170576.0,
      "step": 23810
    },
    {
      "entropy": 5.521656847000122,
      "epoch": 2.6746784972201945,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042945454736246293,
      "loss": 4.9293,
      "mean_token_accuracy": 0.226998433470726,
      "num_tokens": 53181658.0,
      "step": 23815
    },
    {
      "entropy": 5.598465824127198,
      "epoch": 2.675240074128152,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042942535023383075,
      "loss": 5.064,
      "mean_token_accuracy": 0.212347412109375,
      "num_tokens": 53193983.0,
      "step": 23820
    },
    {
      "entropy": 5.5691627025604244,
      "epoch": 2.6758016510361093,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042939614818823547,
      "loss": 5.0014,
      "mean_token_accuracy": 0.2108951836824417,
      "num_tokens": 53205603.0,
      "step": 23825
    },
    {
      "entropy": 5.523241949081421,
      "epoch": 2.676363227944067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004293669412266069,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22274354249238967,
      "num_tokens": 53216295.0,
      "step": 23830
    },
    {
      "entropy": 5.580085039138794,
      "epoch": 2.6769248048520247,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004293377293498751,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22371016591787338,
      "num_tokens": 53227055.0,
      "step": 23835
    },
    {
      "entropy": 5.578140115737915,
      "epoch": 2.677486381759982,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042930851255897014,
      "loss": 5.023,
      "mean_token_accuracy": 0.22000784426927567,
      "num_tokens": 53239050.0,
      "step": 23840
    },
    {
      "entropy": 5.5199408531188965,
      "epoch": 2.6780479586679395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004292792908548223,
      "loss": 4.919,
      "mean_token_accuracy": 0.22172798365354537,
      "num_tokens": 53249551.0,
      "step": 23845
    },
    {
      "entropy": 5.575322532653809,
      "epoch": 2.678609535575897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042925006423836197,
      "loss": 5.094,
      "mean_token_accuracy": 0.21319983303546905,
      "num_tokens": 53261645.0,
      "step": 23850
    },
    {
      "entropy": 5.633069467544556,
      "epoch": 2.6791711124838544,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042922083271051975,
      "loss": 5.0746,
      "mean_token_accuracy": 0.21585700064897537,
      "num_tokens": 53272601.0,
      "step": 23855
    },
    {
      "entropy": 5.601724529266358,
      "epoch": 2.679732689391812,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004291915962722265,
      "loss": 5.0404,
      "mean_token_accuracy": 0.218777571618557,
      "num_tokens": 53283190.0,
      "step": 23860
    },
    {
      "entropy": 5.549609565734864,
      "epoch": 2.6802942662997697,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004291623549244129,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21894527077674866,
      "num_tokens": 53294374.0,
      "step": 23865
    },
    {
      "entropy": 5.601392316818237,
      "epoch": 2.6808558432077274,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042913310866801024,
      "loss": 5.1111,
      "mean_token_accuracy": 0.21685053408145905,
      "num_tokens": 53305609.0,
      "step": 23870
    },
    {
      "entropy": 5.505019998550415,
      "epoch": 2.681417420115685,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042910385750394964,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22423841208219528,
      "num_tokens": 53316251.0,
      "step": 23875
    },
    {
      "entropy": 5.516356706619263,
      "epoch": 2.6819789970236423,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004290746014331623,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22122934460639954,
      "num_tokens": 53326798.0,
      "step": 23880
    },
    {
      "entropy": 5.585136985778808,
      "epoch": 2.6825405739316,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042904534045658,
      "loss": 5.0542,
      "mean_token_accuracy": 0.21137801110744475,
      "num_tokens": 53337586.0,
      "step": 23885
    },
    {
      "entropy": 5.603683185577393,
      "epoch": 2.6831021508395576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004290160745751343,
      "loss": 5.053,
      "mean_token_accuracy": 0.2182625263929367,
      "num_tokens": 53348517.0,
      "step": 23890
    },
    {
      "entropy": 5.5447351932525635,
      "epoch": 2.683663727747515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004289868037897571,
      "loss": 5.0241,
      "mean_token_accuracy": 0.22077293694019318,
      "num_tokens": 53358937.0,
      "step": 23895
    },
    {
      "entropy": 5.537201023101806,
      "epoch": 2.6842253046554725,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004289575281013802,
      "loss": 4.9956,
      "mean_token_accuracy": 0.22470563799142837,
      "num_tokens": 53370007.0,
      "step": 23900
    },
    {
      "entropy": 5.510070848464966,
      "epoch": 2.68478688156343,
      "grad_norm": 1.0625,
      "learning_rate": 0.000428928247510936,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22237492501735687,
      "num_tokens": 53380913.0,
      "step": 23905
    },
    {
      "entropy": 5.519001197814942,
      "epoch": 2.685348458471388,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004288989620193567,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21468696147203445,
      "num_tokens": 53392526.0,
      "step": 23910
    },
    {
      "entropy": 5.575034856796265,
      "epoch": 2.6859100353793455,
      "grad_norm": 1.0,
      "learning_rate": 0.00042886967162757475,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21828781217336654,
      "num_tokens": 53404522.0,
      "step": 23915
    },
    {
      "entropy": 5.566181468963623,
      "epoch": 2.6864716122873027,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042884037633652276,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22057699412107468,
      "num_tokens": 53415156.0,
      "step": 23920
    },
    {
      "entropy": 5.544849634170532,
      "epoch": 2.6870331891952604,
      "grad_norm": 1.0,
      "learning_rate": 0.00042881107614713355,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21106849759817123,
      "num_tokens": 53426297.0,
      "step": 23925
    },
    {
      "entropy": 5.693784236907959,
      "epoch": 2.687594766103218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042878177106034,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20989689230918884,
      "num_tokens": 53438415.0,
      "step": 23930
    },
    {
      "entropy": 5.566653537750244,
      "epoch": 2.6881563430111752,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004287524610770752,
      "loss": 5.0891,
      "mean_token_accuracy": 0.210659396648407,
      "num_tokens": 53450103.0,
      "step": 23935
    },
    {
      "entropy": 5.5405761241912845,
      "epoch": 2.688717919919133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042872314619827234,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2235556349158287,
      "num_tokens": 53461440.0,
      "step": 23940
    },
    {
      "entropy": 5.668830871582031,
      "epoch": 2.6892794968270906,
      "grad_norm": 1.0625,
      "learning_rate": 0.000428693826424865,
      "loss": 5.1323,
      "mean_token_accuracy": 0.20510295480489732,
      "num_tokens": 53472369.0,
      "step": 23945
    },
    {
      "entropy": 5.574690103530884,
      "epoch": 2.6898410737350478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004286645017577866,
      "loss": 5.0513,
      "mean_token_accuracy": 0.22474731504917145,
      "num_tokens": 53483739.0,
      "step": 23950
    },
    {
      "entropy": 5.608912420272827,
      "epoch": 2.6904026506430054,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042863517219797085,
      "loss": 5.049,
      "mean_token_accuracy": 0.2187144637107849,
      "num_tokens": 53494131.0,
      "step": 23955
    },
    {
      "entropy": 5.568646621704102,
      "epoch": 2.690964227550963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042860583774635165,
      "loss": 5.0717,
      "mean_token_accuracy": 0.2091845840215683,
      "num_tokens": 53504923.0,
      "step": 23960
    },
    {
      "entropy": 5.625755405426025,
      "epoch": 2.6915258044589208,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000428576498403863,
      "loss": 5.0657,
      "mean_token_accuracy": 0.21285483241081238,
      "num_tokens": 53516660.0,
      "step": 23965
    },
    {
      "entropy": 5.539487028121949,
      "epoch": 2.6920873813668784,
      "grad_norm": 1.046875,
      "learning_rate": 0.000428547154171439,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22864660918712615,
      "num_tokens": 53527617.0,
      "step": 23970
    },
    {
      "entropy": 5.534249782562256,
      "epoch": 2.6926489582748356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004285178050500142,
      "loss": 5.0242,
      "mean_token_accuracy": 0.21304190158843994,
      "num_tokens": 53537830.0,
      "step": 23975
    },
    {
      "entropy": 5.651747751235962,
      "epoch": 2.6932105351827933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004284884510405228,
      "loss": 5.1357,
      "mean_token_accuracy": 0.21469564586877823,
      "num_tokens": 53549981.0,
      "step": 23980
    },
    {
      "entropy": 5.6523340225219725,
      "epoch": 2.693772112090751,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004284590921438997,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2172248423099518,
      "num_tokens": 53559941.0,
      "step": 23985
    },
    {
      "entropy": 5.501489734649658,
      "epoch": 2.694333688998708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042842972836107954,
      "loss": 4.8958,
      "mean_token_accuracy": 0.219825479388237,
      "num_tokens": 53571611.0,
      "step": 23990
    },
    {
      "entropy": 5.571392297744751,
      "epoch": 2.694895265906666,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004284003596929973,
      "loss": 4.9791,
      "mean_token_accuracy": 0.21499774158000945,
      "num_tokens": 53582425.0,
      "step": 23995
    },
    {
      "entropy": 5.61142897605896,
      "epoch": 2.6954568428146235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042837098614058815,
      "loss": 5.0846,
      "mean_token_accuracy": 0.2119916170835495,
      "num_tokens": 53593576.0,
      "step": 24000
    },
    {
      "epoch": 2.6954568428146235,
      "eval_entropy": 5.333489106246913,
      "eval_loss": 5.106800079345703,
      "eval_mean_token_accuracy": 0.2221290178026217,
      "eval_num_tokens": 53593576.0,
      "eval_runtime": 18.3151,
      "eval_samples_per_second": 1693.299,
      "eval_steps_per_second": 211.683,
      "step": 24000
    },
    {
      "entropy": 5.572452974319458,
      "epoch": 2.696018419722581,
      "grad_norm": 1.0,
      "learning_rate": 0.00042834160770478726,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21769501715898515,
      "num_tokens": 53603768.0,
      "step": 24005
    },
    {
      "entropy": 5.576511430740356,
      "epoch": 2.696579996630539,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042831222438653016,
      "loss": 4.9722,
      "mean_token_accuracy": 0.22322489619255065,
      "num_tokens": 53614464.0,
      "step": 24010
    },
    {
      "entropy": 5.630739545822143,
      "epoch": 2.697141573538496,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004282828361867524,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2122124895453453,
      "num_tokens": 53625641.0,
      "step": 24015
    },
    {
      "entropy": 5.547790431976319,
      "epoch": 2.6977031504464537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042825344310638966,
      "loss": 5.0945,
      "mean_token_accuracy": 0.21605301201343535,
      "num_tokens": 53636741.0,
      "step": 24020
    },
    {
      "entropy": 5.595061302185059,
      "epoch": 2.6982647273544114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042822404514637785,
      "loss": 5.0012,
      "mean_token_accuracy": 0.22136401981115342,
      "num_tokens": 53647295.0,
      "step": 24025
    },
    {
      "entropy": 5.560183668136597,
      "epoch": 2.6988263042623686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000428194642307653,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22381193339824676,
      "num_tokens": 53658371.0,
      "step": 24030
    },
    {
      "entropy": 5.545368957519531,
      "epoch": 2.6993878811703262,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004281652345911513,
      "loss": 5.0407,
      "mean_token_accuracy": 0.21586555540561675,
      "num_tokens": 53669534.0,
      "step": 24035
    },
    {
      "entropy": 5.603540134429932,
      "epoch": 2.699949458078284,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042813582199780913,
      "loss": 5.0577,
      "mean_token_accuracy": 0.213773275911808,
      "num_tokens": 53681247.0,
      "step": 24040
    },
    {
      "entropy": 5.62897219657898,
      "epoch": 2.700511034986241,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004281064045285631,
      "loss": 5.0914,
      "mean_token_accuracy": 0.20999177396297455,
      "num_tokens": 53692197.0,
      "step": 24045
    },
    {
      "entropy": 5.579928874969482,
      "epoch": 2.701072611894199,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042807698218434964,
      "loss": 4.9902,
      "mean_token_accuracy": 0.2182990252971649,
      "num_tokens": 53703734.0,
      "step": 24050
    },
    {
      "entropy": 5.544112491607666,
      "epoch": 2.7016341888021564,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042804755496610576,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2238738939166069,
      "num_tokens": 53714888.0,
      "step": 24055
    },
    {
      "entropy": 5.527381753921508,
      "epoch": 2.702195765710114,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042801812287476836,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21893129050731658,
      "num_tokens": 53726231.0,
      "step": 24060
    },
    {
      "entropy": 5.607086133956909,
      "epoch": 2.7027573426180718,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004279886859112746,
      "loss": 5.0424,
      "mean_token_accuracy": 0.21500605642795562,
      "num_tokens": 53737777.0,
      "step": 24065
    },
    {
      "entropy": 5.590856695175171,
      "epoch": 2.703318919526029,
      "grad_norm": 1.0,
      "learning_rate": 0.00042795924407656175,
      "loss": 5.0276,
      "mean_token_accuracy": 0.21435962170362471,
      "num_tokens": 53748277.0,
      "step": 24070
    },
    {
      "entropy": 5.514821577072143,
      "epoch": 2.7038804964339866,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042792979737156717,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21834062933921813,
      "num_tokens": 53759205.0,
      "step": 24075
    },
    {
      "entropy": 5.560651016235352,
      "epoch": 2.7044420733419443,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004279003457972286,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20839423984289168,
      "num_tokens": 53770984.0,
      "step": 24080
    },
    {
      "entropy": 5.513085508346558,
      "epoch": 2.7050036502499015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004278708893544837,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22643544673919677,
      "num_tokens": 53782092.0,
      "step": 24085
    },
    {
      "entropy": 5.5926683902740475,
      "epoch": 2.705565227157859,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004278414280442704,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21524372845888137,
      "num_tokens": 53794136.0,
      "step": 24090
    },
    {
      "entropy": 5.62717227935791,
      "epoch": 2.706126804065817,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042781196186752676,
      "loss": 5.0605,
      "mean_token_accuracy": 0.21741454899311066,
      "num_tokens": 53804612.0,
      "step": 24095
    },
    {
      "entropy": 5.508116340637207,
      "epoch": 2.7066883809737745,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004277824908251909,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21146359890699387,
      "num_tokens": 53815850.0,
      "step": 24100
    },
    {
      "entropy": 5.531193780899048,
      "epoch": 2.707249957881732,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004277530149182014,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2196519449353218,
      "num_tokens": 53827050.0,
      "step": 24105
    },
    {
      "entropy": 5.60418004989624,
      "epoch": 2.7078115347896894,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042772353414749655,
      "loss": 5.0673,
      "mean_token_accuracy": 0.21215990781784058,
      "num_tokens": 53837697.0,
      "step": 24110
    },
    {
      "entropy": 5.566125345230103,
      "epoch": 2.708373111697647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042769404851401524,
      "loss": 5.003,
      "mean_token_accuracy": 0.22189352810382842,
      "num_tokens": 53849278.0,
      "step": 24115
    },
    {
      "entropy": 5.564800119400024,
      "epoch": 2.7089346886056047,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004276645580186961,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21598363369703294,
      "num_tokens": 53861951.0,
      "step": 24120
    },
    {
      "entropy": 5.620863342285157,
      "epoch": 2.709496265513562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042763506266247834,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2260255604982376,
      "num_tokens": 53873647.0,
      "step": 24125
    },
    {
      "entropy": 5.523846769332886,
      "epoch": 2.7100578424215196,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004276055624463009,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2174445554614067,
      "num_tokens": 53885063.0,
      "step": 24130
    },
    {
      "entropy": 5.52385368347168,
      "epoch": 2.7106194193294773,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042757605737110324,
      "loss": 5.0382,
      "mean_token_accuracy": 0.21579135060310364,
      "num_tokens": 53896955.0,
      "step": 24135
    },
    {
      "entropy": 5.480179738998413,
      "epoch": 2.7111809962374345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004275465474378246,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22779923379421235,
      "num_tokens": 53907862.0,
      "step": 24140
    },
    {
      "entropy": 5.616379165649414,
      "epoch": 2.711742573145392,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004275170326474049,
      "loss": 5.0642,
      "mean_token_accuracy": 0.2180141493678093,
      "num_tokens": 53918918.0,
      "step": 24145
    },
    {
      "entropy": 5.521816110610962,
      "epoch": 2.71230415005335,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042748751300078367,
      "loss": 5.0381,
      "mean_token_accuracy": 0.22044410556554794,
      "num_tokens": 53930488.0,
      "step": 24150
    },
    {
      "entropy": 5.612522745132447,
      "epoch": 2.7128657269613075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004274579884989008,
      "loss": 5.0517,
      "mean_token_accuracy": 0.21975234746932984,
      "num_tokens": 53942314.0,
      "step": 24155
    },
    {
      "entropy": 5.58387017250061,
      "epoch": 2.713427303869265,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042742845914269645,
      "loss": 5.0547,
      "mean_token_accuracy": 0.2284908577799797,
      "num_tokens": 53951915.0,
      "step": 24160
    },
    {
      "entropy": 5.545822525024414,
      "epoch": 2.7139888807772223,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042739892493311095,
      "loss": 5.0606,
      "mean_token_accuracy": 0.22183045893907546,
      "num_tokens": 53962815.0,
      "step": 24165
    },
    {
      "entropy": 5.533163070678711,
      "epoch": 2.71455045768518,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042736938587108447,
      "loss": 4.9864,
      "mean_token_accuracy": 0.21844638735055924,
      "num_tokens": 53972740.0,
      "step": 24170
    },
    {
      "entropy": 5.6149882793426515,
      "epoch": 2.7151120345931377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004273398419575577,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2137635126709938,
      "num_tokens": 53983632.0,
      "step": 24175
    },
    {
      "entropy": 5.655180931091309,
      "epoch": 2.715673611501095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004273102931934713,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2144403412938118,
      "num_tokens": 53994085.0,
      "step": 24180
    },
    {
      "entropy": 5.596403646469116,
      "epoch": 2.7162351884090525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000427280739579766,
      "loss": 5.0096,
      "mean_token_accuracy": 0.22189363539218904,
      "num_tokens": 54004106.0,
      "step": 24185
    },
    {
      "entropy": 5.499273061752319,
      "epoch": 2.71679676531701,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042725118111738305,
      "loss": 4.886,
      "mean_token_accuracy": 0.22676628977060317,
      "num_tokens": 54014582.0,
      "step": 24190
    },
    {
      "entropy": 5.5903040885925295,
      "epoch": 2.717358342224968,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004272216178072633,
      "loss": 5.0131,
      "mean_token_accuracy": 0.217036172747612,
      "num_tokens": 54025626.0,
      "step": 24195
    },
    {
      "entropy": 5.5783709526062015,
      "epoch": 2.7179199191329255,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004271920496503483,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21843255460262298,
      "num_tokens": 54036974.0,
      "step": 24200
    },
    {
      "entropy": 5.614524030685425,
      "epoch": 2.7184814960408827,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042716247664757934,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21634196639060974,
      "num_tokens": 54048404.0,
      "step": 24205
    },
    {
      "entropy": 5.593232011795044,
      "epoch": 2.7190430729488404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042713289879989816,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2170736238360405,
      "num_tokens": 54059210.0,
      "step": 24210
    },
    {
      "entropy": 5.646753549575806,
      "epoch": 2.719604649856798,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004271033161082464,
      "loss": 5.0779,
      "mean_token_accuracy": 0.21819667220115663,
      "num_tokens": 54070406.0,
      "step": 24215
    },
    {
      "entropy": 5.511601495742798,
      "epoch": 2.7201662267647553,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042707372857356615,
      "loss": 4.994,
      "mean_token_accuracy": 0.2175677925348282,
      "num_tokens": 54081497.0,
      "step": 24220
    },
    {
      "entropy": 5.455463123321533,
      "epoch": 2.720727803672713,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004270441361967994,
      "loss": 4.874,
      "mean_token_accuracy": 0.23123635053634645,
      "num_tokens": 54090934.0,
      "step": 24225
    },
    {
      "entropy": 5.461411666870117,
      "epoch": 2.7212893805806706,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004270145389788883,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23075434267520906,
      "num_tokens": 54102011.0,
      "step": 24230
    },
    {
      "entropy": 5.616866874694824,
      "epoch": 2.721850957488628,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042698493692077533,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21805584579706191,
      "num_tokens": 54113066.0,
      "step": 24235
    },
    {
      "entropy": 5.5020520210266115,
      "epoch": 2.7224125343965855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004269553300234031,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2239412322640419,
      "num_tokens": 54124291.0,
      "step": 24240
    },
    {
      "entropy": 5.635113668441773,
      "epoch": 2.722974111304543,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042692571828771423,
      "loss": 5.1423,
      "mean_token_accuracy": 0.21827899366617204,
      "num_tokens": 54136147.0,
      "step": 24245
    },
    {
      "entropy": 5.600636196136475,
      "epoch": 2.723535688212501,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042689610171465147,
      "loss": 5.0315,
      "mean_token_accuracy": 0.2143372967839241,
      "num_tokens": 54147953.0,
      "step": 24250
    },
    {
      "entropy": 5.566958284378051,
      "epoch": 2.7240972651204585,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000426866480305158,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22371114194393157,
      "num_tokens": 54159498.0,
      "step": 24255
    },
    {
      "entropy": 5.613940954208374,
      "epoch": 2.7246588420284157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004268368540601769,
      "loss": 5.003,
      "mean_token_accuracy": 0.21819542348384857,
      "num_tokens": 54170380.0,
      "step": 24260
    },
    {
      "entropy": 5.544346332550049,
      "epoch": 2.7252204189363733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004268072229806514,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21401216685771943,
      "num_tokens": 54181105.0,
      "step": 24265
    },
    {
      "entropy": 5.659995269775391,
      "epoch": 2.725781995844331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004267775870675251,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21240789592266082,
      "num_tokens": 54192142.0,
      "step": 24270
    },
    {
      "entropy": 5.549295711517334,
      "epoch": 2.7263435727522882,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004267479463217415,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21306926012039185,
      "num_tokens": 54202716.0,
      "step": 24275
    },
    {
      "entropy": 5.486480712890625,
      "epoch": 2.726905149660246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042671830074424453,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22646891623735427,
      "num_tokens": 54213035.0,
      "step": 24280
    },
    {
      "entropy": 5.528940105438233,
      "epoch": 2.7274667265682035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000426688650335978,
      "loss": 4.9861,
      "mean_token_accuracy": 0.21878062635660173,
      "num_tokens": 54223999.0,
      "step": 24285
    },
    {
      "entropy": 5.553038740158081,
      "epoch": 2.728028303476161,
      "grad_norm": 1.125,
      "learning_rate": 0.00042665899509788593,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22058823704719543,
      "num_tokens": 54234393.0,
      "step": 24290
    },
    {
      "entropy": 5.5060169219970705,
      "epoch": 2.728589880384119,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004266293350309128,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22042312026023864,
      "num_tokens": 54244787.0,
      "step": 24295
    },
    {
      "entropy": 5.594772386550903,
      "epoch": 2.729151457292076,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004265996701360026,
      "loss": 5.065,
      "mean_token_accuracy": 0.2079576700925827,
      "num_tokens": 54256033.0,
      "step": 24300
    },
    {
      "entropy": 5.607908487319946,
      "epoch": 2.7297130342000337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042657000041410026,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22414944916963578,
      "num_tokens": 54266796.0,
      "step": 24305
    },
    {
      "entropy": 5.640599536895752,
      "epoch": 2.7302746111079914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004265403258661503,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2065444365143776,
      "num_tokens": 54279049.0,
      "step": 24310
    },
    {
      "entropy": 5.576017713546753,
      "epoch": 2.7308361880159486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004265106464930976,
      "loss": 5.037,
      "mean_token_accuracy": 0.22017437666654588,
      "num_tokens": 54290544.0,
      "step": 24315
    },
    {
      "entropy": 5.599188566207886,
      "epoch": 2.7313977649239063,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042648096229588715,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21334056556224823,
      "num_tokens": 54300905.0,
      "step": 24320
    },
    {
      "entropy": 5.5280670642852785,
      "epoch": 2.731959341831864,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042645127327546405,
      "loss": 4.918,
      "mean_token_accuracy": 0.2198034405708313,
      "num_tokens": 54313019.0,
      "step": 24325
    },
    {
      "entropy": 5.637178039550781,
      "epoch": 2.732520918739821,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042642157943277375,
      "loss": 5.0751,
      "mean_token_accuracy": 0.21133809685707092,
      "num_tokens": 54324034.0,
      "step": 24330
    },
    {
      "entropy": 5.531944704055786,
      "epoch": 2.733082495647779,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004263918807687616,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21748074740171433,
      "num_tokens": 54335335.0,
      "step": 24335
    },
    {
      "entropy": 5.617771959304809,
      "epoch": 2.7336440725557365,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004263621772843733,
      "loss": 5.0883,
      "mean_token_accuracy": 0.21539592891931533,
      "num_tokens": 54345659.0,
      "step": 24340
    },
    {
      "entropy": 5.639365196228027,
      "epoch": 2.734205649463694,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004263324689805545,
      "loss": 5.0855,
      "mean_token_accuracy": 0.21422457247972487,
      "num_tokens": 54357959.0,
      "step": 24345
    },
    {
      "entropy": 5.575531768798828,
      "epoch": 2.734767226371652,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004263027558582512,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2188346043229103,
      "num_tokens": 54368779.0,
      "step": 24350
    },
    {
      "entropy": 5.599424409866333,
      "epoch": 2.735328803279609,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004262730379184095,
      "loss": 5.054,
      "mean_token_accuracy": 0.2101218208670616,
      "num_tokens": 54381726.0,
      "step": 24355
    },
    {
      "entropy": 5.622968244552612,
      "epoch": 2.7358903801875667,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042624331516197554,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21746744364500045,
      "num_tokens": 54393201.0,
      "step": 24360
    },
    {
      "entropy": 5.5895304679870605,
      "epoch": 2.7364519570955244,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042621358758989584,
      "loss": 5.0094,
      "mean_token_accuracy": 0.22788841277360916,
      "num_tokens": 54404178.0,
      "step": 24365
    },
    {
      "entropy": 5.489026641845703,
      "epoch": 2.7370135340034816,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004261838552031169,
      "loss": 4.8958,
      "mean_token_accuracy": 0.2265407219529152,
      "num_tokens": 54414963.0,
      "step": 24370
    },
    {
      "entropy": 5.479864406585693,
      "epoch": 2.7375751109114392,
      "grad_norm": 1.0,
      "learning_rate": 0.00042615411800258523,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2286809653043747,
      "num_tokens": 54425333.0,
      "step": 24375
    },
    {
      "entropy": 5.577042293548584,
      "epoch": 2.738136687819397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042612437598924793,
      "loss": 5.1208,
      "mean_token_accuracy": 0.21211368590593338,
      "num_tokens": 54436326.0,
      "step": 24380
    },
    {
      "entropy": 5.560545444488525,
      "epoch": 2.7386982647273546,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004260946291640519,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22010144889354705,
      "num_tokens": 54446664.0,
      "step": 24385
    },
    {
      "entropy": 5.595790672302246,
      "epoch": 2.739259841635312,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042606487752794417,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2246417835354805,
      "num_tokens": 54457135.0,
      "step": 24390
    },
    {
      "entropy": 5.459093523025513,
      "epoch": 2.7398214185432694,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042603512108187224,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2189354494214058,
      "num_tokens": 54468248.0,
      "step": 24395
    },
    {
      "entropy": 5.529761219024659,
      "epoch": 2.740382995451227,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042600535982678347,
      "loss": 4.9236,
      "mean_token_accuracy": 0.23372296690940858,
      "num_tokens": 54478676.0,
      "step": 24400
    },
    {
      "entropy": 5.47899694442749,
      "epoch": 2.7409445723591848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042597559376362544,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21397051960229874,
      "num_tokens": 54490342.0,
      "step": 24405
    },
    {
      "entropy": 5.542941379547119,
      "epoch": 2.741506149267142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004259458228933459,
      "loss": 4.9829,
      "mean_token_accuracy": 0.22426495999097823,
      "num_tokens": 54501543.0,
      "step": 24410
    },
    {
      "entropy": 5.543113803863525,
      "epoch": 2.7420677261750996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042591604721689297,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22226064950227736,
      "num_tokens": 54511777.0,
      "step": 24415
    },
    {
      "entropy": 5.59281587600708,
      "epoch": 2.7426293030830573,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004258862667352145,
      "loss": 5.0065,
      "mean_token_accuracy": 0.22482785433530808,
      "num_tokens": 54523255.0,
      "step": 24420
    },
    {
      "entropy": 5.572739171981811,
      "epoch": 2.7431908799910145,
      "grad_norm": 1.0,
      "learning_rate": 0.0004258564814492587,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21609094589948655,
      "num_tokens": 54534158.0,
      "step": 24425
    },
    {
      "entropy": 5.672663259506225,
      "epoch": 2.743752456898972,
      "grad_norm": 1.078125,
      "learning_rate": 0.000425826691359974,
      "loss": 5.0999,
      "mean_token_accuracy": 0.2128655344247818,
      "num_tokens": 54545735.0,
      "step": 24430
    },
    {
      "entropy": 5.645015811920166,
      "epoch": 2.74431403380693,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004257968964683091,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22110125571489334,
      "num_tokens": 54557112.0,
      "step": 24435
    },
    {
      "entropy": 5.638442420959473,
      "epoch": 2.7448756107148875,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004257670967752125,
      "loss": 5.0855,
      "mean_token_accuracy": 0.21872975528240204,
      "num_tokens": 54568600.0,
      "step": 24440
    },
    {
      "entropy": 5.578036594390869,
      "epoch": 2.745437187622845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000425737292281633,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22351842671632766,
      "num_tokens": 54579008.0,
      "step": 24445
    },
    {
      "entropy": 5.503715085983276,
      "epoch": 2.7459987645308024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004257074829885196,
      "loss": 5.0587,
      "mean_token_accuracy": 0.2173401415348053,
      "num_tokens": 54590306.0,
      "step": 24450
    },
    {
      "entropy": 5.58804874420166,
      "epoch": 2.74656034143876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004256776688968216,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21506866961717605,
      "num_tokens": 54600999.0,
      "step": 24455
    },
    {
      "entropy": 5.622042846679688,
      "epoch": 2.7471219183467177,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042564785000748814,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2219624012708664,
      "num_tokens": 54612977.0,
      "step": 24460
    },
    {
      "entropy": 5.522709846496582,
      "epoch": 2.747683495254675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004256180263214687,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22723279148340225,
      "num_tokens": 54623597.0,
      "step": 24465
    },
    {
      "entropy": 5.477122831344604,
      "epoch": 2.7482450721626326,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042558819783971295,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22777336686849595,
      "num_tokens": 54634618.0,
      "step": 24470
    },
    {
      "entropy": 5.576960802078247,
      "epoch": 2.7488066490705902,
      "grad_norm": 1.125,
      "learning_rate": 0.0004255583645631705,
      "loss": 5.0273,
      "mean_token_accuracy": 0.22223686426877975,
      "num_tokens": 54645291.0,
      "step": 24475
    },
    {
      "entropy": 5.5857186794281,
      "epoch": 2.749368225978548,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004255285264927913,
      "loss": 5.0675,
      "mean_token_accuracy": 0.2140581265091896,
      "num_tokens": 54657294.0,
      "step": 24480
    },
    {
      "entropy": 5.6128990650177,
      "epoch": 2.7499298028865056,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042549868362952556,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21457721889019013,
      "num_tokens": 54668627.0,
      "step": 24485
    },
    {
      "entropy": 5.571594190597534,
      "epoch": 2.750491379794463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042546883597432327,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21791369169950486,
      "num_tokens": 54679763.0,
      "step": 24490
    },
    {
      "entropy": 5.579472541809082,
      "epoch": 2.7510529567024204,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042543898352813485,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21972222924232482,
      "num_tokens": 54690476.0,
      "step": 24495
    },
    {
      "entropy": 5.612585210800171,
      "epoch": 2.751614533610378,
      "grad_norm": 1.03125,
      "learning_rate": 0.000425409126291911,
      "loss": 5.047,
      "mean_token_accuracy": 0.22428634464740754,
      "num_tokens": 54701383.0,
      "step": 24500
    },
    {
      "entropy": 5.581447601318359,
      "epoch": 2.7521761105183353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004253792642666021,
      "loss": 4.9404,
      "mean_token_accuracy": 0.22103769332170486,
      "num_tokens": 54712268.0,
      "step": 24505
    },
    {
      "entropy": 5.564781904220581,
      "epoch": 2.752737687426293,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042534939745315914,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22367111146450042,
      "num_tokens": 54722834.0,
      "step": 24510
    },
    {
      "entropy": 5.576197290420533,
      "epoch": 2.7532992643342507,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000425319525852533,
      "loss": 4.9799,
      "mean_token_accuracy": 0.22555761635303498,
      "num_tokens": 54734628.0,
      "step": 24515
    },
    {
      "entropy": 5.4414551734924315,
      "epoch": 2.753860841242208,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004252896494656749,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2275245189666748,
      "num_tokens": 54745927.0,
      "step": 24520
    },
    {
      "entropy": 5.4690920352935795,
      "epoch": 2.7544224181501655,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004252597682935362,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22259990572929383,
      "num_tokens": 54757822.0,
      "step": 24525
    },
    {
      "entropy": 5.496549081802368,
      "epoch": 2.754983995058123,
      "grad_norm": 1.046875,
      "learning_rate": 0.000425229882337068,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22822325378656388,
      "num_tokens": 54768964.0,
      "step": 24530
    },
    {
      "entropy": 5.555112600326538,
      "epoch": 2.755545571966081,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042519999159722216,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22087000608444213,
      "num_tokens": 54779471.0,
      "step": 24535
    },
    {
      "entropy": 5.5419838428497314,
      "epoch": 2.7561071488740385,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042517009607495035,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22470511347055436,
      "num_tokens": 54790426.0,
      "step": 24540
    },
    {
      "entropy": 5.49076657295227,
      "epoch": 2.7566687257819957,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042514019577120444,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2189055234193802,
      "num_tokens": 54801765.0,
      "step": 24545
    },
    {
      "entropy": 5.563539171218872,
      "epoch": 2.7572303026899534,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004251102906869364,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21498632580041885,
      "num_tokens": 54814272.0,
      "step": 24550
    },
    {
      "entropy": 5.590202760696411,
      "epoch": 2.757791879597911,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004250803808230986,
      "loss": 5.0444,
      "mean_token_accuracy": 0.21512378454208375,
      "num_tokens": 54825610.0,
      "step": 24555
    },
    {
      "entropy": 5.666310882568359,
      "epoch": 2.7583534565058683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042505046618064314,
      "loss": 5.041,
      "mean_token_accuracy": 0.21953587830066681,
      "num_tokens": 54835546.0,
      "step": 24560
    },
    {
      "entropy": 5.596863842010498,
      "epoch": 2.758915033413826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042502054676052273,
      "loss": 5.0506,
      "mean_token_accuracy": 0.2181988000869751,
      "num_tokens": 54845653.0,
      "step": 24565
    },
    {
      "entropy": 5.569756174087525,
      "epoch": 2.7594766103217836,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042499062256368987,
      "loss": 4.9562,
      "mean_token_accuracy": 0.22705154418945311,
      "num_tokens": 54855514.0,
      "step": 24570
    },
    {
      "entropy": 5.510051202774048,
      "epoch": 2.7600381872297413,
      "grad_norm": 1.0,
      "learning_rate": 0.0004249606935910975,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2162928268313408,
      "num_tokens": 54866131.0,
      "step": 24575
    },
    {
      "entropy": 5.492132234573364,
      "epoch": 2.760599764137699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004249307598436984,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2321774333715439,
      "num_tokens": 54876842.0,
      "step": 24580
    },
    {
      "entropy": 5.588446235656738,
      "epoch": 2.761161341045656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004249008213224458,
      "loss": 5.1134,
      "mean_token_accuracy": 0.2104821115732193,
      "num_tokens": 54888036.0,
      "step": 24585
    },
    {
      "entropy": 5.584886646270752,
      "epoch": 2.761722917953614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042487087802829275,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22368431687355042,
      "num_tokens": 54898501.0,
      "step": 24590
    },
    {
      "entropy": 5.589184856414795,
      "epoch": 2.7622844948615715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004248409299621931,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21720079481601715,
      "num_tokens": 54909054.0,
      "step": 24595
    },
    {
      "entropy": 5.573216390609741,
      "epoch": 2.7628460717695287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004248109771251,
      "loss": 5.044,
      "mean_token_accuracy": 0.21608321666717528,
      "num_tokens": 54919572.0,
      "step": 24600
    },
    {
      "entropy": 5.640910911560058,
      "epoch": 2.7634076486774863,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042478101951796716,
      "loss": 5.0316,
      "mean_token_accuracy": 0.2208591103553772,
      "num_tokens": 54930929.0,
      "step": 24605
    },
    {
      "entropy": 5.528279972076416,
      "epoch": 2.763969225585444,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004247510571417487,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22273746579885484,
      "num_tokens": 54941317.0,
      "step": 24610
    },
    {
      "entropy": 5.55590295791626,
      "epoch": 2.764530802493401,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004247210899973985,
      "loss": 4.9574,
      "mean_token_accuracy": 0.22695571333169937,
      "num_tokens": 54951203.0,
      "step": 24615
    },
    {
      "entropy": 5.563991355895996,
      "epoch": 2.765092379401359,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004246911180858707,
      "loss": 5.0146,
      "mean_token_accuracy": 0.2199622169137001,
      "num_tokens": 54961485.0,
      "step": 24620
    },
    {
      "entropy": 5.549723529815674,
      "epoch": 2.7656539563093165,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004246611414081197,
      "loss": 5.007,
      "mean_token_accuracy": 0.21933968961238862,
      "num_tokens": 54973152.0,
      "step": 24625
    },
    {
      "entropy": 5.554009294509887,
      "epoch": 2.766215533217274,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042463115996509984,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2205771952867508,
      "num_tokens": 54984863.0,
      "step": 24630
    },
    {
      "entropy": 5.595094966888428,
      "epoch": 2.766777110125232,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004246011737577658,
      "loss": 5.0079,
      "mean_token_accuracy": 0.2123142346739769,
      "num_tokens": 54997382.0,
      "step": 24635
    },
    {
      "entropy": 5.628778791427612,
      "epoch": 2.767338687033189,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004245711827870724,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21435472071170808,
      "num_tokens": 55007686.0,
      "step": 24640
    },
    {
      "entropy": 5.583675479888916,
      "epoch": 2.7679002639411467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042454118705397454,
      "loss": 5.0942,
      "mean_token_accuracy": 0.21190997064113617,
      "num_tokens": 55018776.0,
      "step": 24645
    },
    {
      "entropy": 5.56077561378479,
      "epoch": 2.7684618408491044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004245111865594273,
      "loss": 5.0635,
      "mean_token_accuracy": 0.21280754655599593,
      "num_tokens": 55031566.0,
      "step": 24650
    },
    {
      "entropy": 5.536598014831543,
      "epoch": 2.7690234177570616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042448118130438583,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23422301113605498,
      "num_tokens": 55043321.0,
      "step": 24655
    },
    {
      "entropy": 5.539708614349365,
      "epoch": 2.7695849946650193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042445117128980553,
      "loss": 4.912,
      "mean_token_accuracy": 0.22180511057376862,
      "num_tokens": 55052655.0,
      "step": 24660
    },
    {
      "entropy": 5.532984781265259,
      "epoch": 2.770146571572977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042442115651664217,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21467611640691758,
      "num_tokens": 55063317.0,
      "step": 24665
    },
    {
      "entropy": 5.512065696716308,
      "epoch": 2.7707081484809346,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042439113698585103,
      "loss": 4.9474,
      "mean_token_accuracy": 0.22369785159826278,
      "num_tokens": 55074916.0,
      "step": 24670
    },
    {
      "entropy": 5.542785739898681,
      "epoch": 2.7712697253888923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004243611126983882,
      "loss": 4.9667,
      "mean_token_accuracy": 0.2206793487071991,
      "num_tokens": 55086456.0,
      "step": 24675
    },
    {
      "entropy": 5.664352893829346,
      "epoch": 2.7718313022968495,
      "grad_norm": 1.0,
      "learning_rate": 0.00042433108365520967,
      "loss": 5.1865,
      "mean_token_accuracy": 0.20690877437591554,
      "num_tokens": 55097268.0,
      "step": 24680
    },
    {
      "entropy": 5.582738780975342,
      "epoch": 2.772392879204807,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042430104985727155,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21955749839544297,
      "num_tokens": 55107451.0,
      "step": 24685
    },
    {
      "entropy": 5.533573770523072,
      "epoch": 2.772954456112765,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042427101130553,
      "loss": 4.9745,
      "mean_token_accuracy": 0.2262762725353241,
      "num_tokens": 55117781.0,
      "step": 24690
    },
    {
      "entropy": 5.607911777496338,
      "epoch": 2.773516033020722,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004242409680009416,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21499606668949128,
      "num_tokens": 55129519.0,
      "step": 24695
    },
    {
      "entropy": 5.4917186260223385,
      "epoch": 2.7740776099286797,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004242109199444629,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2185482621192932,
      "num_tokens": 55138964.0,
      "step": 24700
    },
    {
      "entropy": 5.592500638961792,
      "epoch": 2.7746391868366374,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004241808671370506,
      "loss": 5.01,
      "mean_token_accuracy": 0.21382322907447815,
      "num_tokens": 55149953.0,
      "step": 24705
    },
    {
      "entropy": 5.594241571426392,
      "epoch": 2.7752007637445946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042415080957966176,
      "loss": 5.0675,
      "mean_token_accuracy": 0.21805212795734405,
      "num_tokens": 55162208.0,
      "step": 24710
    },
    {
      "entropy": 5.563787269592285,
      "epoch": 2.7757623406525522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004241207472732531,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2164516717195511,
      "num_tokens": 55174014.0,
      "step": 24715
    },
    {
      "entropy": 5.612440490722657,
      "epoch": 2.77632391756051,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004240906802187821,
      "loss": 5.0908,
      "mean_token_accuracy": 0.20543405264616013,
      "num_tokens": 55185173.0,
      "step": 24720
    },
    {
      "entropy": 5.658352184295654,
      "epoch": 2.7768854944684676,
      "grad_norm": 1.0,
      "learning_rate": 0.0004240606084172061,
      "loss": 5.0708,
      "mean_token_accuracy": 0.21969349235296248,
      "num_tokens": 55198404.0,
      "step": 24725
    },
    {
      "entropy": 5.531762552261353,
      "epoch": 2.777447071376425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042403053186948237,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21393747627735138,
      "num_tokens": 55209126.0,
      "step": 24730
    },
    {
      "entropy": 5.568054294586181,
      "epoch": 2.7780086482843824,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004240004505765688,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21908337026834487,
      "num_tokens": 55220311.0,
      "step": 24735
    },
    {
      "entropy": 5.5500181198120115,
      "epoch": 2.77857022519234,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004239703645394231,
      "loss": 4.9717,
      "mean_token_accuracy": 0.222836272418499,
      "num_tokens": 55230525.0,
      "step": 24740
    },
    {
      "entropy": 5.627074670791626,
      "epoch": 2.7791318021002978,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004239402737590032,
      "loss": 5.1219,
      "mean_token_accuracy": 0.21430562138557435,
      "num_tokens": 55242084.0,
      "step": 24745
    },
    {
      "entropy": 5.557111310958862,
      "epoch": 2.779693379008255,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042391017823626716,
      "loss": 4.9602,
      "mean_token_accuracy": 0.2204621911048889,
      "num_tokens": 55253677.0,
      "step": 24750
    },
    {
      "entropy": 5.520739459991455,
      "epoch": 2.7802549559162126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004238800779721734,
      "loss": 4.9722,
      "mean_token_accuracy": 0.22790087908506393,
      "num_tokens": 55265118.0,
      "step": 24755
    },
    {
      "entropy": 5.5158665657043455,
      "epoch": 2.7808165328241703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042384997296768007,
      "loss": 4.971,
      "mean_token_accuracy": 0.225246798992157,
      "num_tokens": 55277119.0,
      "step": 24760
    },
    {
      "entropy": 5.5899146556854244,
      "epoch": 2.781378109732128,
      "grad_norm": 1.078125,
      "learning_rate": 0.000423819863223746,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22817892879247664,
      "num_tokens": 55287828.0,
      "step": 24765
    },
    {
      "entropy": 5.662408018112183,
      "epoch": 2.7819396866400856,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042378974874132967,
      "loss": 5.091,
      "mean_token_accuracy": 0.21898283660411835,
      "num_tokens": 55299115.0,
      "step": 24770
    },
    {
      "entropy": 5.523025608062744,
      "epoch": 2.782501263548043,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004237596295213901,
      "loss": 4.9942,
      "mean_token_accuracy": 0.21359625160694123,
      "num_tokens": 55310532.0,
      "step": 24775
    },
    {
      "entropy": 5.601618242263794,
      "epoch": 2.7830628404560005,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004237295055648862,
      "loss": 5.0712,
      "mean_token_accuracy": 0.21410947293043137,
      "num_tokens": 55321127.0,
      "step": 24780
    },
    {
      "entropy": 5.626084423065185,
      "epoch": 2.783624417363958,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042369937687277716,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22987637668848038,
      "num_tokens": 55331791.0,
      "step": 24785
    },
    {
      "entropy": 5.604279279708862,
      "epoch": 2.7841859942719154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004236692434460223,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21404936611652375,
      "num_tokens": 55342724.0,
      "step": 24790
    },
    {
      "entropy": 5.52472243309021,
      "epoch": 2.784747571179873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004236391052855811,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21993514746427537,
      "num_tokens": 55354471.0,
      "step": 24795
    },
    {
      "entropy": 5.598960924148559,
      "epoch": 2.7853091480878307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042360896239241306,
      "loss": 4.9748,
      "mean_token_accuracy": 0.22160314470529557,
      "num_tokens": 55366273.0,
      "step": 24800
    },
    {
      "entropy": 5.545458269119263,
      "epoch": 2.785870724995788,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042357881476747807,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2219598487019539,
      "num_tokens": 55377384.0,
      "step": 24805
    },
    {
      "entropy": 5.545984983444214,
      "epoch": 2.7864323019037456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000423548662411736,
      "loss": 5.0053,
      "mean_token_accuracy": 0.22132488042116166,
      "num_tokens": 55387348.0,
      "step": 24810
    },
    {
      "entropy": 5.549749374389648,
      "epoch": 2.7869938788117032,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004235185053261469,
      "loss": 4.9265,
      "mean_token_accuracy": 0.23062274605035782,
      "num_tokens": 55398484.0,
      "step": 24815
    },
    {
      "entropy": 5.539871311187744,
      "epoch": 2.787555455719661,
      "grad_norm": 1.078125,
      "learning_rate": 0.000423488343511671,
      "loss": 4.894,
      "mean_token_accuracy": 0.22464969903230667,
      "num_tokens": 55409347.0,
      "step": 24820
    },
    {
      "entropy": 5.510378646850586,
      "epoch": 2.7881170326276186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004234581769692686,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21947313249111175,
      "num_tokens": 55421048.0,
      "step": 24825
    },
    {
      "entropy": 5.5360705852508545,
      "epoch": 2.788678609535576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004234280056999003,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22744965553283691,
      "num_tokens": 55431189.0,
      "step": 24830
    },
    {
      "entropy": 5.613056468963623,
      "epoch": 2.7892401864435334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004233978297045267,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21342477202415466,
      "num_tokens": 55442695.0,
      "step": 24835
    },
    {
      "entropy": 5.647614336013794,
      "epoch": 2.789801763351491,
      "grad_norm": 1.0,
      "learning_rate": 0.0004233676489841087,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21864757537841797,
      "num_tokens": 55454954.0,
      "step": 24840
    },
    {
      "entropy": 5.5819563388824465,
      "epoch": 2.7903633402594483,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042333746353960723,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2232886254787445,
      "num_tokens": 55466270.0,
      "step": 24845
    },
    {
      "entropy": 5.649620008468628,
      "epoch": 2.790924917167406,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004233072733719833,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21861427277326584,
      "num_tokens": 55476925.0,
      "step": 24850
    },
    {
      "entropy": 5.518258810043335,
      "epoch": 2.7914864940753636,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004232770784821984,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2171281799674034,
      "num_tokens": 55489705.0,
      "step": 24855
    },
    {
      "entropy": 5.627799797058105,
      "epoch": 2.7920480709833213,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042324687887121366,
      "loss": 5.0993,
      "mean_token_accuracy": 0.20990027040243148,
      "num_tokens": 55501362.0,
      "step": 24860
    },
    {
      "entropy": 5.603211450576782,
      "epoch": 2.792609647891279,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004232166745399909,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21785929650068284,
      "num_tokens": 55512754.0,
      "step": 24865
    },
    {
      "entropy": 5.585302448272705,
      "epoch": 2.793171224799236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042318646548949177,
      "loss": 4.981,
      "mean_token_accuracy": 0.21528434306383132,
      "num_tokens": 55524057.0,
      "step": 24870
    },
    {
      "entropy": 5.59127779006958,
      "epoch": 2.793732801707194,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042315625172067794,
      "loss": 5.0025,
      "mean_token_accuracy": 0.2221962869167328,
      "num_tokens": 55534768.0,
      "step": 24875
    },
    {
      "entropy": 5.659724569320678,
      "epoch": 2.7942943786151515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042312603323451173,
      "loss": 5.0721,
      "mean_token_accuracy": 0.21115676909685135,
      "num_tokens": 55547135.0,
      "step": 24880
    },
    {
      "entropy": 5.602214288711548,
      "epoch": 2.7948559555231087,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004230958100319551,
      "loss": 5.0945,
      "mean_token_accuracy": 0.21574446111917495,
      "num_tokens": 55559638.0,
      "step": 24885
    },
    {
      "entropy": 5.573015356063843,
      "epoch": 2.7954175324310664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004230655821139705,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2225364327430725,
      "num_tokens": 55570985.0,
      "step": 24890
    },
    {
      "entropy": 5.544129800796509,
      "epoch": 2.795979109339024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004230353494815203,
      "loss": 4.978,
      "mean_token_accuracy": 0.22083609402179719,
      "num_tokens": 55581701.0,
      "step": 24895
    },
    {
      "entropy": 5.586675214767456,
      "epoch": 2.7965406862469813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004230051121355672,
      "loss": 4.97,
      "mean_token_accuracy": 0.22467518448829651,
      "num_tokens": 55592891.0,
      "step": 24900
    },
    {
      "entropy": 5.479354429244995,
      "epoch": 2.797102263154939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042297487007707387,
      "loss": 5.0,
      "mean_token_accuracy": 0.2241824060678482,
      "num_tokens": 55603762.0,
      "step": 24905
    },
    {
      "entropy": 5.601739692687988,
      "epoch": 2.7976638400628966,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042294462330700336,
      "loss": 5.0287,
      "mean_token_accuracy": 0.21973103731870652,
      "num_tokens": 55614450.0,
      "step": 24910
    },
    {
      "entropy": 5.565270614624024,
      "epoch": 2.7982254169708543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004229143718263186,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2203793540596962,
      "num_tokens": 55625240.0,
      "step": 24915
    },
    {
      "entropy": 5.507693719863892,
      "epoch": 2.798786993878812,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004228841156359829,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2272515594959259,
      "num_tokens": 55635646.0,
      "step": 24920
    },
    {
      "entropy": 5.500457859039306,
      "epoch": 2.799348570786769,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004228538547369596,
      "loss": 4.9162,
      "mean_token_accuracy": 0.2258385419845581,
      "num_tokens": 55646669.0,
      "step": 24925
    },
    {
      "entropy": 5.648530626296997,
      "epoch": 2.799910147694727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042282358913021224,
      "loss": 5.2243,
      "mean_token_accuracy": 0.2106531485915184,
      "num_tokens": 55658792.0,
      "step": 24930
    },
    {
      "entropy": 5.589347457885742,
      "epoch": 2.8004717246026845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042279331881670446,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21786058992147445,
      "num_tokens": 55670037.0,
      "step": 24935
    },
    {
      "entropy": 5.5835710048675535,
      "epoch": 2.8010333015106417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004227630437974,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21184702664613725,
      "num_tokens": 55680405.0,
      "step": 24940
    },
    {
      "entropy": 5.5722685813903805,
      "epoch": 2.8015948784185993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042273276407326304,
      "loss": 5.0045,
      "mean_token_accuracy": 0.22235798686742783,
      "num_tokens": 55692199.0,
      "step": 24945
    },
    {
      "entropy": 5.509901094436645,
      "epoch": 2.802156455326557,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042270247964525755,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22419487088918685,
      "num_tokens": 55702654.0,
      "step": 24950
    },
    {
      "entropy": 5.558305883407593,
      "epoch": 2.8027180322345147,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004226721905143477,
      "loss": 5.005,
      "mean_token_accuracy": 0.22162965983152388,
      "num_tokens": 55712710.0,
      "step": 24955
    },
    {
      "entropy": 5.617483854293823,
      "epoch": 2.8032796091424723,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042264189668149817,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21935369521379472,
      "num_tokens": 55723429.0,
      "step": 24960
    },
    {
      "entropy": 5.596831035614014,
      "epoch": 2.8038411860504295,
      "grad_norm": 1.0,
      "learning_rate": 0.0004226115981476733,
      "loss": 5.0408,
      "mean_token_accuracy": 0.2220585897564888,
      "num_tokens": 55735113.0,
      "step": 24965
    },
    {
      "entropy": 5.418937301635742,
      "epoch": 2.804402762958387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000422581294913838,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22500041872262955,
      "num_tokens": 55745857.0,
      "step": 24970
    },
    {
      "entropy": 5.539986705780029,
      "epoch": 2.804964339866345,
      "grad_norm": 1.0,
      "learning_rate": 0.000422550986980957,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22178531140089036,
      "num_tokens": 55756856.0,
      "step": 24975
    },
    {
      "entropy": 5.628098487854004,
      "epoch": 2.805525916774302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042252067434999524,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21201851665973664,
      "num_tokens": 55768827.0,
      "step": 24980
    },
    {
      "entropy": 5.588510894775391,
      "epoch": 2.8060874936822597,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004224903570219181,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22157487273216248,
      "num_tokens": 55780360.0,
      "step": 24985
    },
    {
      "entropy": 5.594559049606323,
      "epoch": 2.8066490705902174,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042246003499769077,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21708789616823196,
      "num_tokens": 55791874.0,
      "step": 24990
    },
    {
      "entropy": 5.525024366378784,
      "epoch": 2.8072106474981746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042242970827827874,
      "loss": 4.9965,
      "mean_token_accuracy": 0.2198196217417717,
      "num_tokens": 55802689.0,
      "step": 24995
    },
    {
      "entropy": 5.557658338546753,
      "epoch": 2.8077722244061323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004223993768646476,
      "loss": 5.0145,
      "mean_token_accuracy": 0.22290566712617874,
      "num_tokens": 55813945.0,
      "step": 25000
    },
    {
      "entropy": 5.532596778869629,
      "epoch": 2.80833380131409,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004223690407577631,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22224276810884475,
      "num_tokens": 55826403.0,
      "step": 25005
    },
    {
      "entropy": 5.526150894165039,
      "epoch": 2.8088953782220476,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042233869995859127,
      "loss": 4.9788,
      "mean_token_accuracy": 0.22634025067090988,
      "num_tokens": 55836858.0,
      "step": 25010
    },
    {
      "entropy": 5.607988119125366,
      "epoch": 2.8094569551300053,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042230835446809803,
      "loss": 4.9782,
      "mean_token_accuracy": 0.21999674439430236,
      "num_tokens": 55847664.0,
      "step": 25015
    },
    {
      "entropy": 5.533624267578125,
      "epoch": 2.8100185320379625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222780042872496,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22374739795923232,
      "num_tokens": 55858333.0,
      "step": 25020
    },
    {
      "entropy": 5.539474773406982,
      "epoch": 2.81058010894592,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004222476494170125,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2275203749537468,
      "num_tokens": 55868819.0,
      "step": 25025
    },
    {
      "entropy": 5.560573101043701,
      "epoch": 2.811141685853878,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222172898583531,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22400167286396028,
      "num_tokens": 55879133.0,
      "step": 25030
    },
    {
      "entropy": 5.5002296447753904,
      "epoch": 2.811703262761835,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004221869256122381,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22462236881256104,
      "num_tokens": 55890362.0,
      "step": 25035
    },
    {
      "entropy": 5.555015182495117,
      "epoch": 2.8122648396697927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004221565566796343,
      "loss": 4.9735,
      "mean_token_accuracy": 0.2218172311782837,
      "num_tokens": 55899920.0,
      "step": 25040
    },
    {
      "entropy": 5.577810430526734,
      "epoch": 2.8128264165777503,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004221261830615086,
      "loss": 5.02,
      "mean_token_accuracy": 0.2216184064745903,
      "num_tokens": 55911598.0,
      "step": 25045
    },
    {
      "entropy": 5.449719762802124,
      "epoch": 2.813387993485708,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004220958047588282,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22744885832071304,
      "num_tokens": 55922514.0,
      "step": 25050
    },
    {
      "entropy": 5.541629409790039,
      "epoch": 2.8139495703936657,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004220654217725603,
      "loss": 4.9377,
      "mean_token_accuracy": 0.23091571480035783,
      "num_tokens": 55932796.0,
      "step": 25055
    },
    {
      "entropy": 5.533813428878784,
      "epoch": 2.814511147301623,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042203503410367244,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2233802393078804,
      "num_tokens": 55944337.0,
      "step": 25060
    },
    {
      "entropy": 5.637730073928833,
      "epoch": 2.8150727242095805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000422004641753132,
      "loss": 5.1339,
      "mean_token_accuracy": 0.2134704440832138,
      "num_tokens": 55955871.0,
      "step": 25065
    },
    {
      "entropy": 5.629822492599487,
      "epoch": 2.815634301117538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004219742447219068,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2106257662177086,
      "num_tokens": 55966091.0,
      "step": 25070
    },
    {
      "entropy": 5.604153442382812,
      "epoch": 2.8161958780254954,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042194384301096457,
      "loss": 4.985,
      "mean_token_accuracy": 0.21948958337306976,
      "num_tokens": 55976755.0,
      "step": 25075
    },
    {
      "entropy": 5.580384111404419,
      "epoch": 2.816757454933453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042191343662127337,
      "loss": 5.0809,
      "mean_token_accuracy": 0.21938099414110185,
      "num_tokens": 55987579.0,
      "step": 25080
    },
    {
      "entropy": 5.5893807888031,
      "epoch": 2.8173190318414107,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042188302555380144,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21676073223352432,
      "num_tokens": 55997027.0,
      "step": 25085
    },
    {
      "entropy": 5.576378059387207,
      "epoch": 2.817880608749368,
      "grad_norm": 1.015625,
      "learning_rate": 0.000421852609809517,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22023673057556153,
      "num_tokens": 56008928.0,
      "step": 25090
    },
    {
      "entropy": 5.542690277099609,
      "epoch": 2.8184421856573256,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004218221893893884,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22691793739795685,
      "num_tokens": 56021313.0,
      "step": 25095
    },
    {
      "entropy": 5.587626504898071,
      "epoch": 2.8190037625652833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042179176429438454,
      "loss": 4.9786,
      "mean_token_accuracy": 0.224328151345253,
      "num_tokens": 56032002.0,
      "step": 25100
    },
    {
      "entropy": 5.579666185379028,
      "epoch": 2.819565339473241,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004217613345254739,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2341904819011688,
      "num_tokens": 56042905.0,
      "step": 25105
    },
    {
      "entropy": 5.540641403198242,
      "epoch": 2.8201269163811986,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042173090008362535,
      "loss": 4.9668,
      "mean_token_accuracy": 0.21931998580694198,
      "num_tokens": 56052507.0,
      "step": 25110
    },
    {
      "entropy": 5.513655805587769,
      "epoch": 2.820688493289156,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042170046096980807,
      "loss": 5.0502,
      "mean_token_accuracy": 0.21516244411468505,
      "num_tokens": 56065250.0,
      "step": 25115
    },
    {
      "entropy": 5.5741840362548825,
      "epoch": 2.8212500701971135,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042167001718499124,
      "loss": 4.9202,
      "mean_token_accuracy": 0.22367847114801406,
      "num_tokens": 56075858.0,
      "step": 25120
    },
    {
      "entropy": 5.559920787811279,
      "epoch": 2.821811647105071,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004216395687301442,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21813442260026933,
      "num_tokens": 56086647.0,
      "step": 25125
    },
    {
      "entropy": 5.636842060089111,
      "epoch": 2.8223732240130284,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004216091156062364,
      "loss": 5.1969,
      "mean_token_accuracy": 0.20811597108840943,
      "num_tokens": 56098124.0,
      "step": 25130
    },
    {
      "entropy": 5.571738195419312,
      "epoch": 2.822934800920986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004215786578142375,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22047085016965867,
      "num_tokens": 56108505.0,
      "step": 25135
    },
    {
      "entropy": 5.575164842605591,
      "epoch": 2.8234963778289437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042154819535511733,
      "loss": 4.9672,
      "mean_token_accuracy": 0.2236941635608673,
      "num_tokens": 56118423.0,
      "step": 25140
    },
    {
      "entropy": 5.612655305862427,
      "epoch": 2.8240579547369014,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042151772822984567,
      "loss": 5.0326,
      "mean_token_accuracy": 0.22491374015808105,
      "num_tokens": 56130346.0,
      "step": 25145
    },
    {
      "entropy": 5.533376979827881,
      "epoch": 2.824619531644859,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004214872564393928,
      "loss": 5.0437,
      "mean_token_accuracy": 0.22462697625160216,
      "num_tokens": 56141044.0,
      "step": 25150
    },
    {
      "entropy": 5.524428176879883,
      "epoch": 2.8251811085528162,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004214567799847289,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22793195992708207,
      "num_tokens": 56151786.0,
      "step": 25155
    },
    {
      "entropy": 5.572822523117066,
      "epoch": 2.825742685460774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004214262988668242,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21741820126771927,
      "num_tokens": 56164196.0,
      "step": 25160
    },
    {
      "entropy": 5.588466358184815,
      "epoch": 2.8263042623687316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004213958130866494,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21894484460353852,
      "num_tokens": 56176259.0,
      "step": 25165
    },
    {
      "entropy": 5.543361806869507,
      "epoch": 2.8268658392766888,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042136532264517527,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2131354957818985,
      "num_tokens": 56186617.0,
      "step": 25170
    },
    {
      "entropy": 5.444887447357178,
      "epoch": 2.8274274161846464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042133482754337236,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22254674285650253,
      "num_tokens": 56197609.0,
      "step": 25175
    },
    {
      "entropy": 5.494061613082886,
      "epoch": 2.827988993092604,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042130432778221186,
      "loss": 4.8944,
      "mean_token_accuracy": 0.22993071377277374,
      "num_tokens": 56207490.0,
      "step": 25180
    },
    {
      "entropy": 5.582102870941162,
      "epoch": 2.8285505700005613,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042127382336266467,
      "loss": 4.9853,
      "mean_token_accuracy": 0.2176836237311363,
      "num_tokens": 56218167.0,
      "step": 25185
    },
    {
      "entropy": 5.508064413070679,
      "epoch": 2.829112146908519,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004212433142857024,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22484448105096816,
      "num_tokens": 56229199.0,
      "step": 25190
    },
    {
      "entropy": 5.560311508178711,
      "epoch": 2.8296737238164766,
      "grad_norm": 1.0,
      "learning_rate": 0.00042121280055229624,
      "loss": 5.127,
      "mean_token_accuracy": 0.21327512860298156,
      "num_tokens": 56240165.0,
      "step": 25195
    },
    {
      "entropy": 5.635933923721313,
      "epoch": 2.8302353007244343,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004211822821634177,
      "loss": 5.1387,
      "mean_token_accuracy": 0.21098617911338807,
      "num_tokens": 56252447.0,
      "step": 25200
    },
    {
      "entropy": 5.638635635375977,
      "epoch": 2.830796877632392,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004211517591200387,
      "loss": 5.0067,
      "mean_token_accuracy": 0.22357707619667053,
      "num_tokens": 56263116.0,
      "step": 25205
    },
    {
      "entropy": 5.5557092189788815,
      "epoch": 2.831358454540349,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004211212314231309,
      "loss": 5.0509,
      "mean_token_accuracy": 0.2123988926410675,
      "num_tokens": 56276257.0,
      "step": 25210
    },
    {
      "entropy": 5.515365934371948,
      "epoch": 2.831920031448307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042109069907366646,
      "loss": 4.9503,
      "mean_token_accuracy": 0.21880051344633103,
      "num_tokens": 56287865.0,
      "step": 25215
    },
    {
      "entropy": 5.608299446105957,
      "epoch": 2.8324816083562645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042106016207261754,
      "loss": 4.9813,
      "mean_token_accuracy": 0.22467770725488662,
      "num_tokens": 56298122.0,
      "step": 25220
    },
    {
      "entropy": 5.613162040710449,
      "epoch": 2.8330431852642217,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042102962042095643,
      "loss": 5.0899,
      "mean_token_accuracy": 0.21715106666088105,
      "num_tokens": 56309455.0,
      "step": 25225
    },
    {
      "entropy": 5.618431472778321,
      "epoch": 2.8336047621721794,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042099907411965553,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21752699762582778,
      "num_tokens": 56319804.0,
      "step": 25230
    },
    {
      "entropy": 5.584640169143677,
      "epoch": 2.834166339080137,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042096852316968756,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22594480365514755,
      "num_tokens": 56330136.0,
      "step": 25235
    },
    {
      "entropy": 5.700299978256226,
      "epoch": 2.8347279159880947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042093796757202513,
      "loss": 5.1608,
      "mean_token_accuracy": 0.21139343678951264,
      "num_tokens": 56341861.0,
      "step": 25240
    },
    {
      "entropy": 5.532606935501098,
      "epoch": 2.8352894928960524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004209074073276412,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21605881303548813,
      "num_tokens": 56352016.0,
      "step": 25245
    },
    {
      "entropy": 5.465109252929688,
      "epoch": 2.8358510698040096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004208768424375089,
      "loss": 4.915,
      "mean_token_accuracy": 0.22277831733226777,
      "num_tokens": 56363740.0,
      "step": 25250
    },
    {
      "entropy": 5.567902374267578,
      "epoch": 2.8364126467119672,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042084627290260124,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22777475118637086,
      "num_tokens": 56375080.0,
      "step": 25255
    },
    {
      "entropy": 5.598768043518066,
      "epoch": 2.836974223619925,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004208156987238918,
      "loss": 5.0728,
      "mean_token_accuracy": 0.21267261058092118,
      "num_tokens": 56385858.0,
      "step": 25260
    },
    {
      "entropy": 5.6632835388183596,
      "epoch": 2.837535800527882,
      "grad_norm": 1.125,
      "learning_rate": 0.0004207851199023539,
      "loss": 5.0309,
      "mean_token_accuracy": 0.22054607123136521,
      "num_tokens": 56394423.0,
      "step": 25265
    },
    {
      "entropy": 5.554196071624756,
      "epoch": 2.83809737743584,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042075453643896125,
      "loss": 5.0101,
      "mean_token_accuracy": 0.21667972654104234,
      "num_tokens": 56405843.0,
      "step": 25270
    },
    {
      "entropy": 5.533684539794922,
      "epoch": 2.8386589543437974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042072394833468765,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2196830689907074,
      "num_tokens": 56417757.0,
      "step": 25275
    },
    {
      "entropy": 5.5816045761108395,
      "epoch": 2.8392205312517547,
      "grad_norm": 1.046875,
      "learning_rate": 0.000420693355590507,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21871693730354308,
      "num_tokens": 56429937.0,
      "step": 25280
    },
    {
      "entropy": 5.54780216217041,
      "epoch": 2.8397821081597123,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042066275820739333,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22378353476524354,
      "num_tokens": 56441114.0,
      "step": 25285
    },
    {
      "entropy": 5.5242691993713375,
      "epoch": 2.84034368506767,
      "grad_norm": 1.109375,
      "learning_rate": 0.000420632156186321,
      "loss": 4.9952,
      "mean_token_accuracy": 0.21547324061393738,
      "num_tokens": 56451839.0,
      "step": 25290
    },
    {
      "entropy": 5.6241014957427975,
      "epoch": 2.8409052619756276,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004206015495282644,
      "loss": 5.0197,
      "mean_token_accuracy": 0.2218052938580513,
      "num_tokens": 56463384.0,
      "step": 25295
    },
    {
      "entropy": 5.592394685745239,
      "epoch": 2.8414668388835853,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042057093823419795,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21749938130378724,
      "num_tokens": 56474905.0,
      "step": 25300
    },
    {
      "entropy": 5.5512384414672855,
      "epoch": 2.8420284157915425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042054032230509636,
      "loss": 4.9705,
      "mean_token_accuracy": 0.21785319745540618,
      "num_tokens": 56485319.0,
      "step": 25305
    },
    {
      "entropy": 5.551844835281372,
      "epoch": 2.8425899926995,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042050970174193437,
      "loss": 5.0728,
      "mean_token_accuracy": 0.2172453597187996,
      "num_tokens": 56496064.0,
      "step": 25310
    },
    {
      "entropy": 5.548852443695068,
      "epoch": 2.843151569607458,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042047907654568724,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2169172540307045,
      "num_tokens": 56507167.0,
      "step": 25315
    },
    {
      "entropy": 5.6705821514129635,
      "epoch": 2.843713146515415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204484467173297,
      "loss": 5.0834,
      "mean_token_accuracy": 0.21934491097927095,
      "num_tokens": 56518400.0,
      "step": 25320
    },
    {
      "entropy": 5.679170370101929,
      "epoch": 2.8442747234233727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004204178122578373,
      "loss": 5.1195,
      "mean_token_accuracy": 0.21913650929927825,
      "num_tokens": 56529630.0,
      "step": 25325
    },
    {
      "entropy": 5.642560243606567,
      "epoch": 2.8448363003313304,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042038717316818534,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21560747027397156,
      "num_tokens": 56539520.0,
      "step": 25330
    },
    {
      "entropy": 5.57099986076355,
      "epoch": 2.845397877239288,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004203565294493494,
      "loss": 4.983,
      "mean_token_accuracy": 0.22102566063404083,
      "num_tokens": 56550677.0,
      "step": 25335
    },
    {
      "entropy": 5.521417427062988,
      "epoch": 2.8459594541472457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042032588110230525,
      "loss": 5.0631,
      "mean_token_accuracy": 0.21586233675479888,
      "num_tokens": 56561694.0,
      "step": 25340
    },
    {
      "entropy": 5.582666635513306,
      "epoch": 2.846521031055203,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004202952281280286,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2121424436569214,
      "num_tokens": 56572322.0,
      "step": 25345
    },
    {
      "entropy": 5.617684316635132,
      "epoch": 2.8470826079631606,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004202645705274956,
      "loss": 5.021,
      "mean_token_accuracy": 0.21364111453294754,
      "num_tokens": 56583198.0,
      "step": 25350
    },
    {
      "entropy": 5.604371690750122,
      "epoch": 2.8476441848711183,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042023390830168224,
      "loss": 5.0932,
      "mean_token_accuracy": 0.21500879973173143,
      "num_tokens": 56594732.0,
      "step": 25355
    },
    {
      "entropy": 5.59397325515747,
      "epoch": 2.8482057617790755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000420203241451565,
      "loss": 5.0319,
      "mean_token_accuracy": 0.21760542392730714,
      "num_tokens": 56605907.0,
      "step": 25360
    },
    {
      "entropy": 5.652441835403442,
      "epoch": 2.848767338687033,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042017256997812016,
      "loss": 5.0923,
      "mean_token_accuracy": 0.21754715591669083,
      "num_tokens": 56616841.0,
      "step": 25365
    },
    {
      "entropy": 5.617482233047485,
      "epoch": 2.849328915594991,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042014189388232446,
      "loss": 5.032,
      "mean_token_accuracy": 0.21566204279661177,
      "num_tokens": 56628044.0,
      "step": 25370
    },
    {
      "entropy": 5.607907295227051,
      "epoch": 2.849890492502948,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004201112131651545,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21533730626106262,
      "num_tokens": 56638350.0,
      "step": 25375
    },
    {
      "entropy": 5.589438247680664,
      "epoch": 2.8504520694109057,
      "grad_norm": 1.0,
      "learning_rate": 0.0004200805278275873,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21237405240535737,
      "num_tokens": 56649923.0,
      "step": 25380
    },
    {
      "entropy": 5.484613561630249,
      "epoch": 2.8510136463188633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004200498378705997,
      "loss": 4.933,
      "mean_token_accuracy": 0.22003619968891144,
      "num_tokens": 56661210.0,
      "step": 25385
    },
    {
      "entropy": 5.62181921005249,
      "epoch": 2.851575223226821,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004200191432951692,
      "loss": 5.1149,
      "mean_token_accuracy": 0.2090904101729393,
      "num_tokens": 56673106.0,
      "step": 25390
    },
    {
      "entropy": 5.55488543510437,
      "epoch": 2.8521368001347787,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004199884441022728,
      "loss": 5.0195,
      "mean_token_accuracy": 0.22096592485904692,
      "num_tokens": 56684703.0,
      "step": 25395
    },
    {
      "entropy": 5.620799779891968,
      "epoch": 2.852698377042736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041995774029288807,
      "loss": 5.0009,
      "mean_token_accuracy": 0.22020033448934556,
      "num_tokens": 56696881.0,
      "step": 25400
    },
    {
      "entropy": 5.540570402145386,
      "epoch": 2.8532599539506935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041992703186799273,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22649887055158616,
      "num_tokens": 56707372.0,
      "step": 25405
    },
    {
      "entropy": 5.518230724334717,
      "epoch": 2.853821530858651,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041989631882856447,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2277996212244034,
      "num_tokens": 56717954.0,
      "step": 25410
    },
    {
      "entropy": 5.564283752441407,
      "epoch": 2.8543831077666084,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004198656011755812,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2193464994430542,
      "num_tokens": 56728798.0,
      "step": 25415
    },
    {
      "entropy": 5.542463111877441,
      "epoch": 2.854944684674566,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004198348789100211,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21627815514802934,
      "num_tokens": 56740095.0,
      "step": 25420
    },
    {
      "entropy": 5.548013734817505,
      "epoch": 2.8555062615825237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041980415203286223,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2146288648247719,
      "num_tokens": 56751002.0,
      "step": 25425
    },
    {
      "entropy": 5.562070417404175,
      "epoch": 2.8560678384904814,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000419773420545083,
      "loss": 4.9464,
      "mean_token_accuracy": 0.2285858318209648,
      "num_tokens": 56762244.0,
      "step": 25430
    },
    {
      "entropy": 5.6029163837432865,
      "epoch": 2.856629415398439,
      "grad_norm": 1.0,
      "learning_rate": 0.00041974268444766187,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2256344124674797,
      "num_tokens": 56772918.0,
      "step": 25435
    },
    {
      "entropy": 5.518823862075806,
      "epoch": 2.8571909923063963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004197119437415776,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22888641953468322,
      "num_tokens": 56784069.0,
      "step": 25440
    },
    {
      "entropy": 5.5233043193817135,
      "epoch": 2.857752569214354,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004196811984278089,
      "loss": 4.9024,
      "mean_token_accuracy": 0.2309696227312088,
      "num_tokens": 56795187.0,
      "step": 25445
    },
    {
      "entropy": 5.592056608200073,
      "epoch": 2.8583141461223116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004196504485073347,
      "loss": 4.994,
      "mean_token_accuracy": 0.21278913617134093,
      "num_tokens": 56806766.0,
      "step": 25450
    },
    {
      "entropy": 5.620181369781494,
      "epoch": 2.858875723030269,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041961969398113404,
      "loss": 5.0503,
      "mean_token_accuracy": 0.21433290988206863,
      "num_tokens": 56817872.0,
      "step": 25455
    },
    {
      "entropy": 5.621099138259888,
      "epoch": 2.8594372999382265,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041958893485018634,
      "loss": 5.1046,
      "mean_token_accuracy": 0.21248819977045058,
      "num_tokens": 56828291.0,
      "step": 25460
    },
    {
      "entropy": 5.52877287864685,
      "epoch": 2.859998876846184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004195581711154709,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22522586733102798,
      "num_tokens": 56839165.0,
      "step": 25465
    },
    {
      "entropy": 5.530682468414307,
      "epoch": 2.8605604537541414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041952740277796717,
      "loss": 4.929,
      "mean_token_accuracy": 0.2255514532327652,
      "num_tokens": 56850754.0,
      "step": 25470
    },
    {
      "entropy": 5.521741962432861,
      "epoch": 2.861122030662099,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041949662983865486,
      "loss": 5.003,
      "mean_token_accuracy": 0.22144338339567185,
      "num_tokens": 56862798.0,
      "step": 25475
    },
    {
      "entropy": 5.504673337936401,
      "epoch": 2.8616836075700567,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004194658522985139,
      "loss": 4.9284,
      "mean_token_accuracy": 0.23320828825235368,
      "num_tokens": 56873479.0,
      "step": 25480
    },
    {
      "entropy": 5.613129329681397,
      "epoch": 2.8622451844780143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004194350701585241,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22489114701747895,
      "num_tokens": 56883924.0,
      "step": 25485
    },
    {
      "entropy": 5.608363103866577,
      "epoch": 2.862806761385972,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004194042834196656,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2237734392285347,
      "num_tokens": 56895622.0,
      "step": 25490
    },
    {
      "entropy": 5.508962106704712,
      "epoch": 2.8633683382939292,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004193734920829188,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2211256206035614,
      "num_tokens": 56906504.0,
      "step": 25495
    },
    {
      "entropy": 5.488022565841675,
      "epoch": 2.863929915201887,
      "grad_norm": 1.015625,
      "learning_rate": 0.000419342696149264,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21847329139709473,
      "num_tokens": 56917228.0,
      "step": 25500
    },
    {
      "entropy": 5.593483352661133,
      "epoch": 2.8644914921098446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004193118956196818,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22189444899559022,
      "num_tokens": 56927981.0,
      "step": 25505
    },
    {
      "entropy": 5.588128232955933,
      "epoch": 2.8650530690178018,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004192810904951528,
      "loss": 4.9898,
      "mean_token_accuracy": 0.2167352557182312,
      "num_tokens": 56938552.0,
      "step": 25510
    },
    {
      "entropy": 5.565778589248657,
      "epoch": 2.8656146459257594,
      "grad_norm": 1.03125,
      "learning_rate": 0.000419250280776658,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22923055291175842,
      "num_tokens": 56949902.0,
      "step": 25515
    },
    {
      "entropy": 5.616636848449707,
      "epoch": 2.866176222833717,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004192194664651783,
      "loss": 5.1052,
      "mean_token_accuracy": 0.21941858679056167,
      "num_tokens": 56961335.0,
      "step": 25520
    },
    {
      "entropy": 5.520616054534912,
      "epoch": 2.8667377997416748,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004191886475616947,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22722228169441222,
      "num_tokens": 56972542.0,
      "step": 25525
    },
    {
      "entropy": 5.560759449005127,
      "epoch": 2.8672993766496324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041915782406718875,
      "loss": 5.0261,
      "mean_token_accuracy": 0.22453833669424056,
      "num_tokens": 56983838.0,
      "step": 25530
    },
    {
      "entropy": 5.616809511184693,
      "epoch": 2.8678609535575896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004191269959826418,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22325773537158966,
      "num_tokens": 56996125.0,
      "step": 25535
    },
    {
      "entropy": 5.578580284118653,
      "epoch": 2.8684225304655473,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041909616330903544,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21897684335708617,
      "num_tokens": 57007599.0,
      "step": 25540
    },
    {
      "entropy": 5.549531745910644,
      "epoch": 2.868984107373505,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041906532604735127,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22190712690353392,
      "num_tokens": 57018102.0,
      "step": 25545
    },
    {
      "entropy": 5.585144329071045,
      "epoch": 2.869545684281462,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004190344841985713,
      "loss": 5.0821,
      "mean_token_accuracy": 0.21532605290412904,
      "num_tokens": 57029253.0,
      "step": 25550
    },
    {
      "entropy": 5.670800685882568,
      "epoch": 2.87010726118942,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041900363776367747,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21350719183683395,
      "num_tokens": 57040485.0,
      "step": 25555
    },
    {
      "entropy": 5.506251049041748,
      "epoch": 2.8706688380973775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004189727867436519,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22828125357627868,
      "num_tokens": 57051340.0,
      "step": 25560
    },
    {
      "entropy": 5.454259300231934,
      "epoch": 2.8712304150053347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041894193113947703,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21774910390377045,
      "num_tokens": 57061726.0,
      "step": 25565
    },
    {
      "entropy": 5.554666137695312,
      "epoch": 2.8717919919132924,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041891107095213523,
      "loss": 5.0099,
      "mean_token_accuracy": 0.21978580802679062,
      "num_tokens": 57073367.0,
      "step": 25570
    },
    {
      "entropy": 5.615457582473755,
      "epoch": 2.87235356882125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041888020618260913,
      "loss": 5.0926,
      "mean_token_accuracy": 0.22384274005889893,
      "num_tokens": 57085379.0,
      "step": 25575
    },
    {
      "entropy": 5.648680973052978,
      "epoch": 2.8729151457292077,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004188493368318814,
      "loss": 5.0333,
      "mean_token_accuracy": 0.21971053183078765,
      "num_tokens": 57096151.0,
      "step": 25580
    },
    {
      "entropy": 5.583203744888306,
      "epoch": 2.8734767226371654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000418818462900935,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21472398340702056,
      "num_tokens": 57107378.0,
      "step": 25585
    },
    {
      "entropy": 5.554041719436645,
      "epoch": 2.8740382995451226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004187875843907531,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21720563173294066,
      "num_tokens": 57119814.0,
      "step": 25590
    },
    {
      "entropy": 5.764580726623535,
      "epoch": 2.8745998764530802,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041875670130231857,
      "loss": 5.0957,
      "mean_token_accuracy": 0.21878817826509475,
      "num_tokens": 57132070.0,
      "step": 25595
    },
    {
      "entropy": 5.699076461791992,
      "epoch": 2.875161453361038,
      "grad_norm": 1.15625,
      "learning_rate": 0.000418725813636615,
      "loss": 5.0723,
      "mean_token_accuracy": 0.21511399000883102,
      "num_tokens": 57143216.0,
      "step": 25600
    },
    {
      "entropy": 5.4996541976928714,
      "epoch": 2.875723030268995,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041869492139462585,
      "loss": 5.0028,
      "mean_token_accuracy": 0.22251757234334946,
      "num_tokens": 57154485.0,
      "step": 25605
    },
    {
      "entropy": 5.593036079406739,
      "epoch": 2.876284607176953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041866402457733454,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21575814336538315,
      "num_tokens": 57164738.0,
      "step": 25610
    },
    {
      "entropy": 5.626074695587159,
      "epoch": 2.8768461840849104,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041863312318572504,
      "loss": 5.0576,
      "mean_token_accuracy": 0.21503164917230605,
      "num_tokens": 57175790.0,
      "step": 25615
    },
    {
      "entropy": 5.5553820610046385,
      "epoch": 2.877407760992868,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004186022172207812,
      "loss": 5.0941,
      "mean_token_accuracy": 0.21437525749206543,
      "num_tokens": 57187201.0,
      "step": 25620
    },
    {
      "entropy": 5.556068181991577,
      "epoch": 2.8779693379008258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041857130668348705,
      "loss": 4.9455,
      "mean_token_accuracy": 0.2216220200061798,
      "num_tokens": 57197591.0,
      "step": 25625
    },
    {
      "entropy": 5.661864233016968,
      "epoch": 2.878530914808783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004185403915748268,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2155297055840492,
      "num_tokens": 57210194.0,
      "step": 25630
    },
    {
      "entropy": 5.634843587875366,
      "epoch": 2.8790924917167406,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004185094718957848,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2173018604516983,
      "num_tokens": 57220752.0,
      "step": 25635
    },
    {
      "entropy": 5.538264417648316,
      "epoch": 2.8796540686246983,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041847854764734556,
      "loss": 4.922,
      "mean_token_accuracy": 0.2284750372171402,
      "num_tokens": 57233109.0,
      "step": 25640
    },
    {
      "entropy": 5.58117094039917,
      "epoch": 2.8802156455326555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004184476188304937,
      "loss": 5.1038,
      "mean_token_accuracy": 0.21447619646787644,
      "num_tokens": 57244548.0,
      "step": 25645
    },
    {
      "entropy": 5.594193410873413,
      "epoch": 2.880777222440613,
      "grad_norm": 1.125,
      "learning_rate": 0.00041841668544621396,
      "loss": 4.977,
      "mean_token_accuracy": 0.22558897733688354,
      "num_tokens": 57255328.0,
      "step": 25650
    },
    {
      "entropy": 5.60559663772583,
      "epoch": 2.881338799348571,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004183857474954915,
      "loss": 5.0274,
      "mean_token_accuracy": 0.2153394639492035,
      "num_tokens": 57268890.0,
      "step": 25655
    },
    {
      "entropy": 5.652784395217895,
      "epoch": 2.881900376256528,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041835480497931114,
      "loss": 5.0666,
      "mean_token_accuracy": 0.2175716757774353,
      "num_tokens": 57278505.0,
      "step": 25660
    },
    {
      "entropy": 5.681154394149781,
      "epoch": 2.8824619531644857,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004183238578986582,
      "loss": 5.059,
      "mean_token_accuracy": 0.2152912825345993,
      "num_tokens": 57289499.0,
      "step": 25665
    },
    {
      "entropy": 5.705482053756714,
      "epoch": 2.8830235300724434,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041829290625451806,
      "loss": 5.184,
      "mean_token_accuracy": 0.21483875513076783,
      "num_tokens": 57300970.0,
      "step": 25670
    },
    {
      "entropy": 5.576712799072266,
      "epoch": 2.883585106980401,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041826195004787616,
      "loss": 4.9339,
      "mean_token_accuracy": 0.22936901599168777,
      "num_tokens": 57311997.0,
      "step": 25675
    },
    {
      "entropy": 5.5998588562011715,
      "epoch": 2.8841466838883587,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041823098927971827,
      "loss": 5.0467,
      "mean_token_accuracy": 0.22226238250732422,
      "num_tokens": 57323385.0,
      "step": 25680
    },
    {
      "entropy": 5.5819634914398195,
      "epoch": 2.884708260796316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004182000239510301,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21354903131723404,
      "num_tokens": 57335389.0,
      "step": 25685
    },
    {
      "entropy": 5.627308511734009,
      "epoch": 2.8852698377042736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041816905406279756,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2173183187842369,
      "num_tokens": 57346843.0,
      "step": 25690
    },
    {
      "entropy": 5.6221472263336185,
      "epoch": 2.8858314146122312,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041813807961600685,
      "loss": 5.0818,
      "mean_token_accuracy": 0.21935378462076188,
      "num_tokens": 57358253.0,
      "step": 25695
    },
    {
      "entropy": 5.613869571685791,
      "epoch": 2.8863929915201885,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004181071006116442,
      "loss": 5.0974,
      "mean_token_accuracy": 0.21291582584381102,
      "num_tokens": 57369204.0,
      "step": 25700
    },
    {
      "entropy": 5.589415884017944,
      "epoch": 2.886954568428146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041807611705069593,
      "loss": 4.9701,
      "mean_token_accuracy": 0.22392016649246216,
      "num_tokens": 57380214.0,
      "step": 25705
    },
    {
      "entropy": 5.613571453094482,
      "epoch": 2.887516145336104,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041804512893414865,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21565171629190444,
      "num_tokens": 57391848.0,
      "step": 25710
    },
    {
      "entropy": 5.667595481872558,
      "epoch": 2.8880777222440615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041801413626298893,
      "loss": 5.107,
      "mean_token_accuracy": 0.21288004368543625,
      "num_tokens": 57403026.0,
      "step": 25715
    },
    {
      "entropy": 5.565627956390381,
      "epoch": 2.888639299152019,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004179831390382037,
      "loss": 5.0259,
      "mean_token_accuracy": 0.21569849103689193,
      "num_tokens": 57414223.0,
      "step": 25720
    },
    {
      "entropy": 5.5719810962677006,
      "epoch": 2.8892008760599763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004179521372607798,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22517713755369187,
      "num_tokens": 57424661.0,
      "step": 25725
    },
    {
      "entropy": 5.567163467407227,
      "epoch": 2.889762452967934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041792113093170447,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22311867624521256,
      "num_tokens": 57435859.0,
      "step": 25730
    },
    {
      "entropy": 5.555958938598633,
      "epoch": 2.8903240298758917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004178901200519649,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21287058144807816,
      "num_tokens": 57447784.0,
      "step": 25735
    },
    {
      "entropy": 5.562472534179688,
      "epoch": 2.890885606783849,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004178591046225485,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21277427673339844,
      "num_tokens": 57458881.0,
      "step": 25740
    },
    {
      "entropy": 5.558726024627686,
      "epoch": 2.8914471836918065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041782808464444274,
      "loss": 5.0325,
      "mean_token_accuracy": 0.21660642623901366,
      "num_tokens": 57470287.0,
      "step": 25745
    },
    {
      "entropy": 5.636974430084228,
      "epoch": 2.892008760599764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041779706011863546,
      "loss": 5.0872,
      "mean_token_accuracy": 0.21207299083471298,
      "num_tokens": 57481723.0,
      "step": 25750
    },
    {
      "entropy": 5.643393182754517,
      "epoch": 2.8925703375077214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041776603104611427,
      "loss": 5.0709,
      "mean_token_accuracy": 0.21883759200572966,
      "num_tokens": 57492804.0,
      "step": 25755
    },
    {
      "entropy": 5.531958293914795,
      "epoch": 2.893131914415679,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041773499742786736,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22381579726934434,
      "num_tokens": 57504148.0,
      "step": 25760
    },
    {
      "entropy": 5.541896772384644,
      "epoch": 2.8936934913236367,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041770395926488276,
      "loss": 5.0384,
      "mean_token_accuracy": 0.22132462561130523,
      "num_tokens": 57516026.0,
      "step": 25765
    },
    {
      "entropy": 5.586121225357056,
      "epoch": 2.8942550682315944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004176729165581488,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21188684105873107,
      "num_tokens": 57526882.0,
      "step": 25770
    },
    {
      "entropy": 5.537028217315674,
      "epoch": 2.894816645139552,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041764186930865377,
      "loss": 5.0316,
      "mean_token_accuracy": 0.22203753143548965,
      "num_tokens": 57537875.0,
      "step": 25775
    },
    {
      "entropy": 5.617013645172119,
      "epoch": 2.8953782220475093,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004176108175173863,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2171571061015129,
      "num_tokens": 57548143.0,
      "step": 25780
    },
    {
      "entropy": 5.5392743110656735,
      "epoch": 2.895939798955467,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004175797611853351,
      "loss": 4.9831,
      "mean_token_accuracy": 0.22201718837022782,
      "num_tokens": 57558127.0,
      "step": 25785
    },
    {
      "entropy": 5.59144606590271,
      "epoch": 2.8965013758634246,
      "grad_norm": 1.03125,
      "learning_rate": 0.000417548700313489,
      "loss": 5.0523,
      "mean_token_accuracy": 0.22744081616401673,
      "num_tokens": 57570028.0,
      "step": 25790
    },
    {
      "entropy": 5.585885238647461,
      "epoch": 2.897062952771382,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417517634902837,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22208693325519563,
      "num_tokens": 57581089.0,
      "step": 25795
    },
    {
      "entropy": 5.655886936187744,
      "epoch": 2.8976245296793395,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041748656495436834,
      "loss": 5.126,
      "mean_token_accuracy": 0.21554871052503585,
      "num_tokens": 57592555.0,
      "step": 25800
    },
    {
      "entropy": 5.569664669036865,
      "epoch": 2.898186106587297,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041745549046907205,
      "loss": 4.979,
      "mean_token_accuracy": 0.21984843164682388,
      "num_tokens": 57603492.0,
      "step": 25805
    },
    {
      "entropy": 5.518515634536743,
      "epoch": 2.898747683495255,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041742441144793774,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2189430072903633,
      "num_tokens": 57613477.0,
      "step": 25810
    },
    {
      "entropy": 5.579858922958374,
      "epoch": 2.8993092604032125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004173933278919549,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21146612763404846,
      "num_tokens": 57623703.0,
      "step": 25815
    },
    {
      "entropy": 5.628709173202514,
      "epoch": 2.8998708373111697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004173622398021133,
      "loss": 4.9909,
      "mean_token_accuracy": 0.22287665605545043,
      "num_tokens": 57637172.0,
      "step": 25820
    },
    {
      "entropy": 5.624779224395752,
      "epoch": 2.9004324142191273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041733114717940273,
      "loss": 5.0436,
      "mean_token_accuracy": 0.21682052612304686,
      "num_tokens": 57648224.0,
      "step": 25825
    },
    {
      "entropy": 5.566975116729736,
      "epoch": 2.900993991127085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004173000500248133,
      "loss": 5.024,
      "mean_token_accuracy": 0.22818444669246674,
      "num_tokens": 57659093.0,
      "step": 25830
    },
    {
      "entropy": 5.652878761291504,
      "epoch": 2.901555568035042,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004172689483393351,
      "loss": 5.1058,
      "mean_token_accuracy": 0.21668851673603057,
      "num_tokens": 57671399.0,
      "step": 25835
    },
    {
      "entropy": 5.686131572723388,
      "epoch": 2.902117144943,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041723784212395834,
      "loss": 5.0323,
      "mean_token_accuracy": 0.22203932106494903,
      "num_tokens": 57682150.0,
      "step": 25840
    },
    {
      "entropy": 5.584662437438965,
      "epoch": 2.9026787218509575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041720673137967356,
      "loss": 4.9946,
      "mean_token_accuracy": 0.22005308866500856,
      "num_tokens": 57693232.0,
      "step": 25845
    },
    {
      "entropy": 5.551379156112671,
      "epoch": 2.9032402987589148,
      "grad_norm": 1.125,
      "learning_rate": 0.00041717561610747124,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2204699382185936,
      "num_tokens": 57703422.0,
      "step": 25850
    },
    {
      "entropy": 5.5801605701446535,
      "epoch": 2.9038018756668724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041714449630834223,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21819274872541428,
      "num_tokens": 57714454.0,
      "step": 25855
    },
    {
      "entropy": 5.575002765655517,
      "epoch": 2.90436345257483,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004171133719832774,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22660937756299973,
      "num_tokens": 57726243.0,
      "step": 25860
    },
    {
      "entropy": 5.53998212814331,
      "epoch": 2.9049250294827877,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041708224313326747,
      "loss": 4.9839,
      "mean_token_accuracy": 0.21518145203590394,
      "num_tokens": 57737882.0,
      "step": 25865
    },
    {
      "entropy": 5.647746610641479,
      "epoch": 2.9054866063907454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000417051109759304,
      "loss": 5.036,
      "mean_token_accuracy": 0.22101628929376602,
      "num_tokens": 57750205.0,
      "step": 25870
    },
    {
      "entropy": 5.599517440795898,
      "epoch": 2.9060481832987026,
      "grad_norm": 1.109375,
      "learning_rate": 0.000417019971862378,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22639530897140503,
      "num_tokens": 57761762.0,
      "step": 25875
    },
    {
      "entropy": 5.5916694641113285,
      "epoch": 2.9066097602066603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041698882944348105,
      "loss": 5.0282,
      "mean_token_accuracy": 0.21556390523910524,
      "num_tokens": 57771280.0,
      "step": 25880
    },
    {
      "entropy": 5.602675151824951,
      "epoch": 2.907171337114618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041695768250360466,
      "loss": 4.9763,
      "mean_token_accuracy": 0.22077510952949525,
      "num_tokens": 57784099.0,
      "step": 25885
    },
    {
      "entropy": 5.604307413101196,
      "epoch": 2.907732914022575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004169265310437407,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21510854810476304,
      "num_tokens": 57795544.0,
      "step": 25890
    },
    {
      "entropy": 5.543724060058594,
      "epoch": 2.908294490930533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004168953750648809,
      "loss": 4.9277,
      "mean_token_accuracy": 0.22994485050439833,
      "num_tokens": 57805963.0,
      "step": 25895
    },
    {
      "entropy": 5.6285515308380125,
      "epoch": 2.9088560678384905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041686421456801724,
      "loss": 5.0675,
      "mean_token_accuracy": 0.21311390548944473,
      "num_tokens": 57815281.0,
      "step": 25900
    },
    {
      "entropy": 5.632473564147949,
      "epoch": 2.909417644746448,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004168330495541421,
      "loss": 5.0751,
      "mean_token_accuracy": 0.21101573258638381,
      "num_tokens": 57826172.0,
      "step": 25905
    },
    {
      "entropy": 5.650136518478393,
      "epoch": 2.909979221654406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041680188002424755,
      "loss": 4.9487,
      "mean_token_accuracy": 0.21966490447521209,
      "num_tokens": 57836617.0,
      "step": 25910
    },
    {
      "entropy": 5.587276315689087,
      "epoch": 2.910540798562363,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004167707059793262,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2253006100654602,
      "num_tokens": 57847636.0,
      "step": 25915
    },
    {
      "entropy": 5.649514770507812,
      "epoch": 2.9111023754703207,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041673952742037057,
      "loss": 5.1017,
      "mean_token_accuracy": 0.2112824499607086,
      "num_tokens": 57858816.0,
      "step": 25920
    },
    {
      "entropy": 5.595126628875732,
      "epoch": 2.9116639523782784,
      "grad_norm": 1.125,
      "learning_rate": 0.0004167083443483734,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2180323675274849,
      "num_tokens": 57870749.0,
      "step": 25925
    },
    {
      "entropy": 5.571021175384521,
      "epoch": 2.9122255292862356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004166771567643276,
      "loss": 5.039,
      "mean_token_accuracy": 0.22007303982973098,
      "num_tokens": 57883454.0,
      "step": 25930
    },
    {
      "entropy": 5.582950496673584,
      "epoch": 2.9127871061941932,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004166459646692262,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21960506290197374,
      "num_tokens": 57894498.0,
      "step": 25935
    },
    {
      "entropy": 5.487398099899292,
      "epoch": 2.913348683102151,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004166147680640623,
      "loss": 4.8351,
      "mean_token_accuracy": 0.22900986075401306,
      "num_tokens": 57905332.0,
      "step": 25940
    },
    {
      "entropy": 5.5136803150177,
      "epoch": 2.913910260010108,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041658356694982936,
      "loss": 4.9465,
      "mean_token_accuracy": 0.2299311026930809,
      "num_tokens": 57916456.0,
      "step": 25945
    },
    {
      "entropy": 5.552207803726196,
      "epoch": 2.9144718369180658,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004165523613275207,
      "loss": 5.0194,
      "mean_token_accuracy": 0.22633545398712157,
      "num_tokens": 57927615.0,
      "step": 25950
    },
    {
      "entropy": 5.495301532745361,
      "epoch": 2.9150334138260234,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004165211511981299,
      "loss": 4.8846,
      "mean_token_accuracy": 0.2282563790678978,
      "num_tokens": 57937771.0,
      "step": 25955
    },
    {
      "entropy": 5.526004028320313,
      "epoch": 2.915594990733981,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004164899365626507,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22007538378238678,
      "num_tokens": 57948564.0,
      "step": 25960
    },
    {
      "entropy": 5.522476577758789,
      "epoch": 2.9161565676419388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004164587174220771,
      "loss": 4.9636,
      "mean_token_accuracy": 0.2199821352958679,
      "num_tokens": 57959758.0,
      "step": 25965
    },
    {
      "entropy": 5.618832349777222,
      "epoch": 2.916718144549896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041642749377740316,
      "loss": 5.0434,
      "mean_token_accuracy": 0.22226566076278687,
      "num_tokens": 57970988.0,
      "step": 25970
    },
    {
      "entropy": 5.551792812347412,
      "epoch": 2.9172797214578536,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041639626562962293,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22866615056991577,
      "num_tokens": 57981056.0,
      "step": 25975
    },
    {
      "entropy": 5.5215682029724125,
      "epoch": 2.9178412983658113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041636503297973074,
      "loss": 5.0116,
      "mean_token_accuracy": 0.22373811155557632,
      "num_tokens": 57992092.0,
      "step": 25980
    },
    {
      "entropy": 5.5986888885498045,
      "epoch": 2.9184028752737685,
      "grad_norm": 1.171875,
      "learning_rate": 0.000416333795828721,
      "loss": 4.993,
      "mean_token_accuracy": 0.22930201143026352,
      "num_tokens": 58003000.0,
      "step": 25985
    },
    {
      "entropy": 5.536039733886719,
      "epoch": 2.918964452181726,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004163025541775885,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21866042017936707,
      "num_tokens": 58013932.0,
      "step": 25990
    },
    {
      "entropy": 5.6430394649505615,
      "epoch": 2.919526029089684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004162713080273278,
      "loss": 5.0572,
      "mean_token_accuracy": 0.22134116441011428,
      "num_tokens": 58024481.0,
      "step": 25995
    },
    {
      "entropy": 5.619962978363037,
      "epoch": 2.9200876059976415,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041624005737893387,
      "loss": 5.0185,
      "mean_token_accuracy": 0.22093937844038009,
      "num_tokens": 58035753.0,
      "step": 26000
    },
    {
      "entropy": 5.704223108291626,
      "epoch": 2.920649182905599,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004162088022334018,
      "loss": 5.1426,
      "mean_token_accuracy": 0.21296065002679826,
      "num_tokens": 58046599.0,
      "step": 26005
    },
    {
      "entropy": 5.619533061981201,
      "epoch": 2.9212107598135564,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041617754259172665,
      "loss": 5.0776,
      "mean_token_accuracy": 0.21119716316461562,
      "num_tokens": 58058151.0,
      "step": 26010
    },
    {
      "entropy": 5.612974834442139,
      "epoch": 2.921772336721514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041614627845490377,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2157071962952614,
      "num_tokens": 58070116.0,
      "step": 26015
    },
    {
      "entropy": 5.609308052062988,
      "epoch": 2.9223339136294717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041611500982392867,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21962206214666366,
      "num_tokens": 58081089.0,
      "step": 26020
    },
    {
      "entropy": 5.4825715065002445,
      "epoch": 2.922895490537429,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041608373669979696,
      "loss": 4.9203,
      "mean_token_accuracy": 0.2235110208392143,
      "num_tokens": 58092033.0,
      "step": 26025
    },
    {
      "entropy": 5.544528532028198,
      "epoch": 2.9234570674453866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004160524590835042,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22315115332603455,
      "num_tokens": 58102864.0,
      "step": 26030
    },
    {
      "entropy": 5.534972238540649,
      "epoch": 2.9240186443533442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041602117697604664,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2109203666448593,
      "num_tokens": 58114607.0,
      "step": 26035
    },
    {
      "entropy": 5.5864527225494385,
      "epoch": 2.9245802212613015,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041598989037842,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22107511311769484,
      "num_tokens": 58125935.0,
      "step": 26040
    },
    {
      "entropy": 5.566807794570923,
      "epoch": 2.925141798169259,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004159585992916206,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22541995644569396,
      "num_tokens": 58136439.0,
      "step": 26045
    },
    {
      "entropy": 5.599353790283203,
      "epoch": 2.925703375077217,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004159273037166447,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21346690356731415,
      "num_tokens": 58147778.0,
      "step": 26050
    },
    {
      "entropy": 5.6144946098327635,
      "epoch": 2.9262649519851744,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041589600365448884,
      "loss": 5.0179,
      "mean_token_accuracy": 0.2200607642531395,
      "num_tokens": 58159290.0,
      "step": 26055
    },
    {
      "entropy": 5.6046782493591305,
      "epoch": 2.926826528893132,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004158646991061495,
      "loss": 5.0468,
      "mean_token_accuracy": 0.21236024796962738,
      "num_tokens": 58169917.0,
      "step": 26060
    },
    {
      "entropy": 5.539593935012817,
      "epoch": 2.9273881058010893,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004158333900726235,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22637903392314912,
      "num_tokens": 58180905.0,
      "step": 26065
    },
    {
      "entropy": 5.576563930511474,
      "epoch": 2.927949682709047,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041580207655490776,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21662573963403703,
      "num_tokens": 58192312.0,
      "step": 26070
    },
    {
      "entropy": 5.529155015945435,
      "epoch": 2.9285112596170046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004157707585539994,
      "loss": 4.9593,
      "mean_token_accuracy": 0.2309773325920105,
      "num_tokens": 58204739.0,
      "step": 26075
    },
    {
      "entropy": 5.579205894470215,
      "epoch": 2.929072836524962,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041573943607089536,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22167314141988753,
      "num_tokens": 58215663.0,
      "step": 26080
    },
    {
      "entropy": 5.56591420173645,
      "epoch": 2.9296344134329195,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004157081091065932,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21874925047159194,
      "num_tokens": 58226480.0,
      "step": 26085
    },
    {
      "entropy": 5.767205810546875,
      "epoch": 2.930195990340877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004156767776620902,
      "loss": 5.2402,
      "mean_token_accuracy": 0.20726231932640077,
      "num_tokens": 58238087.0,
      "step": 26090
    },
    {
      "entropy": 5.575633955001831,
      "epoch": 2.930757567248835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000415645441738384,
      "loss": 5.0309,
      "mean_token_accuracy": 0.21267254799604415,
      "num_tokens": 58248738.0,
      "step": 26095
    },
    {
      "entropy": 5.519163322448731,
      "epoch": 2.9313191441567925,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041561410133647244,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21795392483472825,
      "num_tokens": 58260440.0,
      "step": 26100
    },
    {
      "entropy": 5.62781834602356,
      "epoch": 2.9318807210647497,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041558275645735336,
      "loss": 5.052,
      "mean_token_accuracy": 0.21572353541851044,
      "num_tokens": 58272704.0,
      "step": 26105
    },
    {
      "entropy": 5.59791932106018,
      "epoch": 2.9324422979727074,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041555140710202484,
      "loss": 5.0543,
      "mean_token_accuracy": 0.21554007828235627,
      "num_tokens": 58284254.0,
      "step": 26110
    },
    {
      "entropy": 5.621101713180542,
      "epoch": 2.933003874880665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041552005327148494,
      "loss": 5.0964,
      "mean_token_accuracy": 0.22161561995744705,
      "num_tokens": 58296472.0,
      "step": 26115
    },
    {
      "entropy": 5.552947902679444,
      "epoch": 2.9335654517886223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004154886949667321,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22825936675071717,
      "num_tokens": 58307734.0,
      "step": 26120
    },
    {
      "entropy": 5.58795166015625,
      "epoch": 2.93412702869658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041545733218876466,
      "loss": 4.9452,
      "mean_token_accuracy": 0.22123263776302338,
      "num_tokens": 58318329.0,
      "step": 26125
    },
    {
      "entropy": 5.506069564819336,
      "epoch": 2.9346886056045376,
      "grad_norm": 1.0,
      "learning_rate": 0.0004154259649385813,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2213407501578331,
      "num_tokens": 58330036.0,
      "step": 26130
    },
    {
      "entropy": 5.568278694152832,
      "epoch": 2.935250182512495,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041539459321718076,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22399204671382905,
      "num_tokens": 58340303.0,
      "step": 26135
    },
    {
      "entropy": 5.561682033538818,
      "epoch": 2.9358117594204525,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004153632170255619,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23447517305612564,
      "num_tokens": 58352394.0,
      "step": 26140
    },
    {
      "entropy": 5.60401291847229,
      "epoch": 2.93637333632841,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041533183636472385,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21153907626867294,
      "num_tokens": 58364605.0,
      "step": 26145
    },
    {
      "entropy": 5.58825364112854,
      "epoch": 2.936934913236368,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004153004512356656,
      "loss": 5.0229,
      "mean_token_accuracy": 0.22536056786775588,
      "num_tokens": 58375514.0,
      "step": 26150
    },
    {
      "entropy": 5.618220090866089,
      "epoch": 2.9374964901443255,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004152690616393866,
      "loss": 4.992,
      "mean_token_accuracy": 0.22333476841449737,
      "num_tokens": 58387135.0,
      "step": 26155
    },
    {
      "entropy": 5.550895643234253,
      "epoch": 2.9380580670522827,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041523766757688634,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21838063150644302,
      "num_tokens": 58399077.0,
      "step": 26160
    },
    {
      "entropy": 5.653751611709595,
      "epoch": 2.9386196439602403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004152062690491643,
      "loss": 5.1222,
      "mean_token_accuracy": 0.21968915015459062,
      "num_tokens": 58409897.0,
      "step": 26165
    },
    {
      "entropy": 5.518504858016968,
      "epoch": 2.939181220868198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004151748660572203,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22614940851926804,
      "num_tokens": 58421189.0,
      "step": 26170
    },
    {
      "entropy": 5.517765665054322,
      "epoch": 2.939742797776155,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004151434586020541,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21793224513530732,
      "num_tokens": 58432258.0,
      "step": 26175
    },
    {
      "entropy": 5.642579174041748,
      "epoch": 2.940304374684113,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041511204668466593,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21964603364467622,
      "num_tokens": 58444288.0,
      "step": 26180
    },
    {
      "entropy": 5.6506102085113525,
      "epoch": 2.9408659515920705,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041508063030605585,
      "loss": 5.011,
      "mean_token_accuracy": 0.22824863195419312,
      "num_tokens": 58453367.0,
      "step": 26185
    },
    {
      "entropy": 5.622426080703735,
      "epoch": 2.941427528500028,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041504920946722413,
      "loss": 5.1527,
      "mean_token_accuracy": 0.21039184480905532,
      "num_tokens": 58464862.0,
      "step": 26190
    },
    {
      "entropy": 5.582870149612427,
      "epoch": 2.941989105407986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004150177841691713,
      "loss": 5.0004,
      "mean_token_accuracy": 0.22930199205875396,
      "num_tokens": 58476353.0,
      "step": 26195
    },
    {
      "entropy": 5.630409574508667,
      "epoch": 2.942550682315943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004149863544128979,
      "loss": 5.0608,
      "mean_token_accuracy": 0.2126747876405716,
      "num_tokens": 58488330.0,
      "step": 26200
    },
    {
      "entropy": 5.596170282363891,
      "epoch": 2.9431122592239007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041495492019940467,
      "loss": 4.9989,
      "mean_token_accuracy": 0.22741350531578064,
      "num_tokens": 58500004.0,
      "step": 26205
    },
    {
      "entropy": 5.541852903366089,
      "epoch": 2.9436738361318584,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004149234815296925,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21544096171855925,
      "num_tokens": 58511433.0,
      "step": 26210
    },
    {
      "entropy": 5.495598220825196,
      "epoch": 2.9442354130398156,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041489203840476243,
      "loss": 4.9303,
      "mean_token_accuracy": 0.21606895625591277,
      "num_tokens": 58522234.0,
      "step": 26215
    },
    {
      "entropy": 5.603760290145874,
      "epoch": 2.9447969899477733,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041486059082561556,
      "loss": 5.015,
      "mean_token_accuracy": 0.2208595171570778,
      "num_tokens": 58534370.0,
      "step": 26220
    },
    {
      "entropy": 5.657776498794556,
      "epoch": 2.945358566855731,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004148291387932533,
      "loss": 4.962,
      "mean_token_accuracy": 0.23000783920288087,
      "num_tokens": 58545859.0,
      "step": 26225
    },
    {
      "entropy": 5.5281473159790036,
      "epoch": 2.9459201437636886,
      "grad_norm": 1.0,
      "learning_rate": 0.00041479768230867696,
      "loss": 4.9261,
      "mean_token_accuracy": 0.2331186279654503,
      "num_tokens": 58557159.0,
      "step": 26230
    },
    {
      "entropy": 5.5488300800323485,
      "epoch": 2.946481720671646,
      "grad_norm": 1.0,
      "learning_rate": 0.00041476622137288823,
      "loss": 5.0182,
      "mean_token_accuracy": 0.21529874801635743,
      "num_tokens": 58568985.0,
      "step": 26235
    },
    {
      "entropy": 5.622339153289795,
      "epoch": 2.9470432975796035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041473475598688886,
      "loss": 4.9994,
      "mean_token_accuracy": 0.22320175170898438,
      "num_tokens": 58579897.0,
      "step": 26240
    },
    {
      "entropy": 5.528769779205322,
      "epoch": 2.947604874487561,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004147032861516805,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2284076690673828,
      "num_tokens": 58590042.0,
      "step": 26245
    },
    {
      "entropy": 5.573739337921142,
      "epoch": 2.948166451395519,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041467181186826546,
      "loss": 5.0591,
      "mean_token_accuracy": 0.22101672738790512,
      "num_tokens": 58601645.0,
      "step": 26250
    },
    {
      "entropy": 5.594943618774414,
      "epoch": 2.948728028303476,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041464033313764575,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21516553461551666,
      "num_tokens": 58613881.0,
      "step": 26255
    },
    {
      "entropy": 5.553203582763672,
      "epoch": 2.9492896052114337,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041460884996082363,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22829726487398147,
      "num_tokens": 58623430.0,
      "step": 26260
    },
    {
      "entropy": 5.541867446899414,
      "epoch": 2.9498511821193913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041457736233880166,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21828176081180573,
      "num_tokens": 58635289.0,
      "step": 26265
    },
    {
      "entropy": 5.505901288986206,
      "epoch": 2.9504127590273486,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004145458702725823,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21869869977235795,
      "num_tokens": 58646086.0,
      "step": 26270
    },
    {
      "entropy": 5.556966257095337,
      "epoch": 2.9509743359353062,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041451437376316825,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21314738243818282,
      "num_tokens": 58657662.0,
      "step": 26275
    },
    {
      "entropy": 5.587275123596191,
      "epoch": 2.951535912843264,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004144828728115624,
      "loss": 4.9796,
      "mean_token_accuracy": 0.223736409842968,
      "num_tokens": 58668054.0,
      "step": 26280
    },
    {
      "entropy": 5.6440675258636475,
      "epoch": 2.9520974897512215,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004144513674187679,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2185098111629486,
      "num_tokens": 58679285.0,
      "step": 26285
    },
    {
      "entropy": 5.625115776062012,
      "epoch": 2.952659066659179,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004144198575857878,
      "loss": 5.0901,
      "mean_token_accuracy": 0.22209459841251372,
      "num_tokens": 58691878.0,
      "step": 26290
    },
    {
      "entropy": 5.491191053390503,
      "epoch": 2.9532206435671364,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004143883433136252,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2253616377711296,
      "num_tokens": 58702130.0,
      "step": 26295
    },
    {
      "entropy": 5.575263500213623,
      "epoch": 2.953782220475094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004143568246032838,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22902376651763917,
      "num_tokens": 58712872.0,
      "step": 26300
    },
    {
      "entropy": 5.598026514053345,
      "epoch": 2.9543437973830518,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004143253014557671,
      "loss": 5.0523,
      "mean_token_accuracy": 0.21337251961231232,
      "num_tokens": 58724783.0,
      "step": 26305
    },
    {
      "entropy": 5.650496912002564,
      "epoch": 2.954905374291009,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041429377387207865,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21597865968942642,
      "num_tokens": 58736473.0,
      "step": 26310
    },
    {
      "entropy": 5.610618162155151,
      "epoch": 2.9554669511989666,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041426224185322255,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2174597591161728,
      "num_tokens": 58748223.0,
      "step": 26315
    },
    {
      "entropy": 5.512318849563599,
      "epoch": 2.9560285281069243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004142307054002025,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22877279222011565,
      "num_tokens": 58759876.0,
      "step": 26320
    },
    {
      "entropy": 5.6148382186889645,
      "epoch": 2.956590105014882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004141991645140228,
      "loss": 5.0328,
      "mean_token_accuracy": 0.2205335631966591,
      "num_tokens": 58771288.0,
      "step": 26325
    },
    {
      "entropy": 5.641060304641724,
      "epoch": 2.957151681922839,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041416761919568784,
      "loss": 5.0968,
      "mean_token_accuracy": 0.2179628849029541,
      "num_tokens": 58781664.0,
      "step": 26330
    },
    {
      "entropy": 5.601086616516113,
      "epoch": 2.957713258830797,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041413606944620183,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21538540869951248,
      "num_tokens": 58792808.0,
      "step": 26335
    },
    {
      "entropy": 5.550546646118164,
      "epoch": 2.9582748357387545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004141045152665695,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23040433079004288,
      "num_tokens": 58804455.0,
      "step": 26340
    },
    {
      "entropy": 5.529049634933472,
      "epoch": 2.958836412646712,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041407295665779533,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22595041394233703,
      "num_tokens": 58814718.0,
      "step": 26345
    },
    {
      "entropy": 5.586625099182129,
      "epoch": 2.9593979895546694,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041404139362088434,
      "loss": 5.0682,
      "mean_token_accuracy": 0.21869244426488876,
      "num_tokens": 58825649.0,
      "step": 26350
    },
    {
      "entropy": 5.609233236312866,
      "epoch": 2.959959566462627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004140098261568414,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21534397155046464,
      "num_tokens": 58836529.0,
      "step": 26355
    },
    {
      "entropy": 5.600598764419556,
      "epoch": 2.9605211433705847,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041397825426667175,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22086244374513625,
      "num_tokens": 58847879.0,
      "step": 26360
    },
    {
      "entropy": 5.646154689788818,
      "epoch": 2.961082720278542,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004139466779513805,
      "loss": 5.1575,
      "mean_token_accuracy": 0.21078127920627593,
      "num_tokens": 58859283.0,
      "step": 26365
    },
    {
      "entropy": 5.6125109672546385,
      "epoch": 2.9616442971864996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004139150972119732,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2143400177359581,
      "num_tokens": 58870529.0,
      "step": 26370
    },
    {
      "entropy": 5.596693134307861,
      "epoch": 2.9622058740944572,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041388351204945534,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2229960799217224,
      "num_tokens": 58880828.0,
      "step": 26375
    },
    {
      "entropy": 5.682480525970459,
      "epoch": 2.962767451002415,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041385192246483257,
      "loss": 5.0877,
      "mean_token_accuracy": 0.22313906699419023,
      "num_tokens": 58892121.0,
      "step": 26380
    },
    {
      "entropy": 5.602609252929687,
      "epoch": 2.9633290279103726,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041382032845911073,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21667010337114334,
      "num_tokens": 58902768.0,
      "step": 26385
    },
    {
      "entropy": 5.648215293884277,
      "epoch": 2.9638906048183298,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041378873003329575,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2186923071742058,
      "num_tokens": 58913404.0,
      "step": 26390
    },
    {
      "entropy": 5.687697458267212,
      "epoch": 2.9644521817262874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004137571271883938,
      "loss": 5.1314,
      "mean_token_accuracy": 0.215249964594841,
      "num_tokens": 58925086.0,
      "step": 26395
    },
    {
      "entropy": 5.551814460754395,
      "epoch": 2.965013758634245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004137255199254112,
      "loss": 5.005,
      "mean_token_accuracy": 0.21374452263116836,
      "num_tokens": 58935425.0,
      "step": 26400
    },
    {
      "entropy": 5.605255460739135,
      "epoch": 2.9655753355422023,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041369390824535417,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21672741174697877,
      "num_tokens": 58945528.0,
      "step": 26405
    },
    {
      "entropy": 5.566385507583618,
      "epoch": 2.96613691245016,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041366229214922926,
      "loss": 4.9975,
      "mean_token_accuracy": 0.2219054564833641,
      "num_tokens": 58957008.0,
      "step": 26410
    },
    {
      "entropy": 5.628593873977661,
      "epoch": 2.9666984893581176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004136306716380432,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21100234091281891,
      "num_tokens": 58968648.0,
      "step": 26415
    },
    {
      "entropy": 5.60047025680542,
      "epoch": 2.9672600662660753,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041359904671280285,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2123686581850052,
      "num_tokens": 58979794.0,
      "step": 26420
    },
    {
      "entropy": 5.570560169219971,
      "epoch": 2.9678216431740325,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004135674173745151,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22488165497779847,
      "num_tokens": 58990534.0,
      "step": 26425
    },
    {
      "entropy": 5.577532958984375,
      "epoch": 2.96838322008199,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041353578362418695,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22167059630155564,
      "num_tokens": 59001717.0,
      "step": 26430
    },
    {
      "entropy": 5.509610652923584,
      "epoch": 2.968944796989948,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004135041454628258,
      "loss": 4.9639,
      "mean_token_accuracy": 0.21911582797765733,
      "num_tokens": 59012118.0,
      "step": 26435
    },
    {
      "entropy": 5.635817813873291,
      "epoch": 2.9695063738979055,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004134725028914389,
      "loss": 5.054,
      "mean_token_accuracy": 0.2168222650885582,
      "num_tokens": 59023940.0,
      "step": 26440
    },
    {
      "entropy": 5.637709093093872,
      "epoch": 2.9700679508058627,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041344085591103383,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22222494184970856,
      "num_tokens": 59035147.0,
      "step": 26445
    },
    {
      "entropy": 5.497382497787475,
      "epoch": 2.9706295277138204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004134092045226182,
      "loss": 4.908,
      "mean_token_accuracy": 0.2278777092695236,
      "num_tokens": 59045531.0,
      "step": 26450
    },
    {
      "entropy": 5.621706628799439,
      "epoch": 2.971191104621778,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041337754872719984,
      "loss": 5.108,
      "mean_token_accuracy": 0.2249643012881279,
      "num_tokens": 59056364.0,
      "step": 26455
    },
    {
      "entropy": 5.56344428062439,
      "epoch": 2.9717526815297353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004133458885257866,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22375116795301436,
      "num_tokens": 59066504.0,
      "step": 26460
    },
    {
      "entropy": 5.572540616989135,
      "epoch": 2.972314258437693,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041331422391938663,
      "loss": 4.9891,
      "mean_token_accuracy": 0.22185958176851273,
      "num_tokens": 59077888.0,
      "step": 26465
    },
    {
      "entropy": 5.624399995803833,
      "epoch": 2.9728758353456506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004132825549090082,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21959903389215468,
      "num_tokens": 59088222.0,
      "step": 26470
    },
    {
      "entropy": 5.603703737258911,
      "epoch": 2.9734374122536082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004132508814956595,
      "loss": 5.011,
      "mean_token_accuracy": 0.21700435131788254,
      "num_tokens": 59099302.0,
      "step": 26475
    },
    {
      "entropy": 5.585181856155396,
      "epoch": 2.973998989161566,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004132192036803491,
      "loss": 4.987,
      "mean_token_accuracy": 0.2150997444987297,
      "num_tokens": 59109238.0,
      "step": 26480
    },
    {
      "entropy": 5.485153532028198,
      "epoch": 2.974560566069523,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041318752146408575,
      "loss": 4.9215,
      "mean_token_accuracy": 0.21950652003288268,
      "num_tokens": 59119795.0,
      "step": 26485
    },
    {
      "entropy": 5.520458364486695,
      "epoch": 2.975122142977481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004131558348478781,
      "loss": 5.0035,
      "mean_token_accuracy": 0.2174706995487213,
      "num_tokens": 59129649.0,
      "step": 26490
    },
    {
      "entropy": 5.610235977172851,
      "epoch": 2.9756837198854384,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041312414383273507,
      "loss": 5.0657,
      "mean_token_accuracy": 0.21650943011045456,
      "num_tokens": 59140802.0,
      "step": 26495
    },
    {
      "entropy": 5.6205446243286135,
      "epoch": 2.9762452967933957,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041309244841966565,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22316927164793016,
      "num_tokens": 59151333.0,
      "step": 26500
    },
    {
      "entropy": 5.613866806030273,
      "epoch": 2.9768068737013533,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041306074860967916,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2100142791867256,
      "num_tokens": 59163414.0,
      "step": 26505
    },
    {
      "entropy": 5.615428113937378,
      "epoch": 2.977368450609311,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041302904440378494,
      "loss": 5.0295,
      "mean_token_accuracy": 0.22036806643009185,
      "num_tokens": 59173884.0,
      "step": 26510
    },
    {
      "entropy": 5.637410640716553,
      "epoch": 2.9779300275172687,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041299733580299236,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21469663381576537,
      "num_tokens": 59185263.0,
      "step": 26515
    },
    {
      "entropy": 5.678665447235107,
      "epoch": 2.978491604425226,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041296562280831106,
      "loss": 5.1452,
      "mean_token_accuracy": 0.2054934874176979,
      "num_tokens": 59196279.0,
      "step": 26520
    },
    {
      "entropy": 5.480897092819214,
      "epoch": 2.9790531813331835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041293390542075095,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22765495628118515,
      "num_tokens": 59207537.0,
      "step": 26525
    },
    {
      "entropy": 5.501996231079102,
      "epoch": 2.979614758241141,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004129021836413217,
      "loss": 4.9477,
      "mean_token_accuracy": 0.2224799320101738,
      "num_tokens": 59218583.0,
      "step": 26530
    },
    {
      "entropy": 5.553728866577148,
      "epoch": 2.980176335149099,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004128704574710334,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22505664676427842,
      "num_tokens": 59228638.0,
      "step": 26535
    },
    {
      "entropy": 5.677872037887573,
      "epoch": 2.980737912057056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004128387269108963,
      "loss": 5.1172,
      "mean_token_accuracy": 0.21747157871723174,
      "num_tokens": 59240574.0,
      "step": 26540
    },
    {
      "entropy": 5.567373514175415,
      "epoch": 2.9812994889650137,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041280699196192074,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22396643012762069,
      "num_tokens": 59252986.0,
      "step": 26545
    },
    {
      "entropy": 5.57235164642334,
      "epoch": 2.9818610658729714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041277525262511705,
      "loss": 5.067,
      "mean_token_accuracy": 0.2181283250451088,
      "num_tokens": 59263034.0,
      "step": 26550
    },
    {
      "entropy": 5.598926544189453,
      "epoch": 2.9824226427809286,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041274350890149584,
      "loss": 5.0887,
      "mean_token_accuracy": 0.21375298202037812,
      "num_tokens": 59275170.0,
      "step": 26555
    },
    {
      "entropy": 5.6085728168487545,
      "epoch": 2.9829842196888863,
      "grad_norm": 1.0625,
      "learning_rate": 0.000412711760792068,
      "loss": 5.0499,
      "mean_token_accuracy": 0.2210780054330826,
      "num_tokens": 59287168.0,
      "step": 26560
    },
    {
      "entropy": 5.670382452011109,
      "epoch": 2.983545796596844,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004126800082978442,
      "loss": 5.0614,
      "mean_token_accuracy": 0.22260820120573044,
      "num_tokens": 59298043.0,
      "step": 26565
    },
    {
      "entropy": 5.529120969772339,
      "epoch": 2.9841073735048016,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041264825141983547,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2268499791622162,
      "num_tokens": 59310454.0,
      "step": 26570
    },
    {
      "entropy": 5.635623264312744,
      "epoch": 2.9846689504127593,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041261649015905314,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2167591482400894,
      "num_tokens": 59323275.0,
      "step": 26575
    },
    {
      "entropy": 5.570049858093261,
      "epoch": 2.9852305273207165,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004125847245165084,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21821131259202958,
      "num_tokens": 59334003.0,
      "step": 26580
    },
    {
      "entropy": 5.545033597946167,
      "epoch": 2.985792104228674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004125529544932126,
      "loss": 4.948,
      "mean_token_accuracy": 0.22576240301132203,
      "num_tokens": 59345403.0,
      "step": 26585
    },
    {
      "entropy": 5.570388841629028,
      "epoch": 2.986353681136632,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004125211800901774,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21354705542325975,
      "num_tokens": 59357457.0,
      "step": 26590
    },
    {
      "entropy": 5.579530668258667,
      "epoch": 2.986915258044589,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004124894013084145,
      "loss": 4.9303,
      "mean_token_accuracy": 0.22503935992717744,
      "num_tokens": 59369059.0,
      "step": 26595
    },
    {
      "entropy": 5.616629886627197,
      "epoch": 2.9874768349525467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041245761814893576,
      "loss": 4.9845,
      "mean_token_accuracy": 0.21994331181049348,
      "num_tokens": 59380700.0,
      "step": 26600
    },
    {
      "entropy": 5.561266231536865,
      "epoch": 2.9880384118605043,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004124258306127531,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2159329980611801,
      "num_tokens": 59391504.0,
      "step": 26605
    },
    {
      "entropy": 5.621786642074585,
      "epoch": 2.988599988768462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004123940387008787,
      "loss": 5.0463,
      "mean_token_accuracy": 0.21490940749645232,
      "num_tokens": 59402648.0,
      "step": 26610
    },
    {
      "entropy": 5.598483467102051,
      "epoch": 2.9891615656764197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004123622424143248,
      "loss": 5.0732,
      "mean_token_accuracy": 0.21554801911115645,
      "num_tokens": 59413842.0,
      "step": 26615
    },
    {
      "entropy": 5.624752330780029,
      "epoch": 2.989723142584377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041233044175410387,
      "loss": 5.1079,
      "mean_token_accuracy": 0.21546099036931993,
      "num_tokens": 59425630.0,
      "step": 26620
    },
    {
      "entropy": 5.569988441467285,
      "epoch": 2.9902847194923345,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041229863672122844,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22226320207118988,
      "num_tokens": 59437140.0,
      "step": 26625
    },
    {
      "entropy": 5.570414686203003,
      "epoch": 2.990846296400292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004122668273167111,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2230682909488678,
      "num_tokens": 59447629.0,
      "step": 26630
    },
    {
      "entropy": 5.539887523651123,
      "epoch": 2.9914078733082494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004122350135415648,
      "loss": 5.0435,
      "mean_token_accuracy": 0.20820383429527284,
      "num_tokens": 59458469.0,
      "step": 26635
    },
    {
      "entropy": 5.561685228347779,
      "epoch": 2.991969450216207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004122031953968024,
      "loss": 4.9843,
      "mean_token_accuracy": 0.22044271528720855,
      "num_tokens": 59468864.0,
      "step": 26640
    },
    {
      "entropy": 5.538389492034912,
      "epoch": 2.9925310271241647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041217137288343704,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21897886246442794,
      "num_tokens": 59480689.0,
      "step": 26645
    },
    {
      "entropy": 5.6484619140625,
      "epoch": 2.993092604032122,
      "grad_norm": 1.171875,
      "learning_rate": 0.000412139546002482,
      "loss": 5.0842,
      "mean_token_accuracy": 0.2144427329301834,
      "num_tokens": 59491510.0,
      "step": 26650
    },
    {
      "entropy": 5.616292667388916,
      "epoch": 2.9936541809400796,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041210771475495064,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21332893073558806,
      "num_tokens": 59502838.0,
      "step": 26655
    },
    {
      "entropy": 5.60566520690918,
      "epoch": 2.9942157578480373,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004120758791418564,
      "loss": 4.9949,
      "mean_token_accuracy": 0.22539695352315903,
      "num_tokens": 59514356.0,
      "step": 26660
    },
    {
      "entropy": 5.557268190383911,
      "epoch": 2.994777334755995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004120440391642131,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21648603230714797,
      "num_tokens": 59525933.0,
      "step": 26665
    },
    {
      "entropy": 5.660343790054322,
      "epoch": 2.9953389116639526,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004120121948230343,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2218962073326111,
      "num_tokens": 59536856.0,
      "step": 26670
    },
    {
      "entropy": 5.618624877929688,
      "epoch": 2.99590048857191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004119803461193342,
      "loss": 5.0253,
      "mean_token_accuracy": 0.22435340732336045,
      "num_tokens": 59546023.0,
      "step": 26675
    },
    {
      "entropy": 5.567879486083984,
      "epoch": 2.9964620654798675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004119484930541267,
      "loss": 4.977,
      "mean_token_accuracy": 0.22302140593528746,
      "num_tokens": 59557515.0,
      "step": 26680
    },
    {
      "entropy": 5.567043495178223,
      "epoch": 2.997023642387825,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041191663562842605,
      "loss": 5.0361,
      "mean_token_accuracy": 0.21538899093866348,
      "num_tokens": 59570198.0,
      "step": 26685
    },
    {
      "entropy": 5.635405588150024,
      "epoch": 2.9975852192957824,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004118847738432467,
      "loss": 5.1009,
      "mean_token_accuracy": 0.21203597038984298,
      "num_tokens": 59580365.0,
      "step": 26690
    },
    {
      "entropy": 5.543608236312866,
      "epoch": 2.99814679620374,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004118529076996029,
      "loss": 4.8923,
      "mean_token_accuracy": 0.2234704986214638,
      "num_tokens": 59591304.0,
      "step": 26695
    },
    {
      "entropy": 5.6318567276000975,
      "epoch": 2.9987083731116977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041182103719850967,
      "loss": 5.0827,
      "mean_token_accuracy": 0.21497339904308319,
      "num_tokens": 59603005.0,
      "step": 26700
    },
    {
      "entropy": 5.571457004547119,
      "epoch": 2.9992699500196554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041178916234098134,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2288787245750427,
      "num_tokens": 59614211.0,
      "step": 26705
    },
    {
      "entropy": 5.549999809265136,
      "epoch": 2.999831526927613,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004117572831280332,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2151966705918312,
      "num_tokens": 59624162.0,
      "step": 26710
    },
    {
      "entropy": 5.551123089260525,
      "epoch": 3.0003369461447744,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041172539956068,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21254568464226192,
      "num_tokens": 59633165.0,
      "step": 26715
    },
    {
      "entropy": 5.673091411590576,
      "epoch": 3.000898523052732,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041169351163993705,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21330106109380723,
      "num_tokens": 59645130.0,
      "step": 26720
    },
    {
      "entropy": 5.545118522644043,
      "epoch": 3.0014600999606897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041166161936681973,
      "loss": 4.8888,
      "mean_token_accuracy": 0.23387515395879746,
      "num_tokens": 59656826.0,
      "step": 26725
    },
    {
      "entropy": 5.609489679336548,
      "epoch": 3.002021676868647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004116297227423435,
      "loss": 5.0595,
      "mean_token_accuracy": 0.2225101634860039,
      "num_tokens": 59667743.0,
      "step": 26730
    },
    {
      "entropy": 5.4989766597747805,
      "epoch": 3.0025832537766046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041159782176752375,
      "loss": 4.8301,
      "mean_token_accuracy": 0.2330364391207695,
      "num_tokens": 59679750.0,
      "step": 26735
    },
    {
      "entropy": 5.490346908569336,
      "epoch": 3.0031448306845623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004115659164433765,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22495172321796417,
      "num_tokens": 59689113.0,
      "step": 26740
    },
    {
      "entropy": 5.522772741317749,
      "epoch": 3.00370640759252,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041153400677091745,
      "loss": 4.874,
      "mean_token_accuracy": 0.22304388731718064,
      "num_tokens": 59701282.0,
      "step": 26745
    },
    {
      "entropy": 5.577074289321899,
      "epoch": 3.004267984500477,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041150209275116266,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22000405043363572,
      "num_tokens": 59713536.0,
      "step": 26750
    },
    {
      "entropy": 5.668290901184082,
      "epoch": 3.004829561408435,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041147017438512834,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2148563027381897,
      "num_tokens": 59723818.0,
      "step": 26755
    },
    {
      "entropy": 5.66949462890625,
      "epoch": 3.0053911383163925,
      "grad_norm": 1.125,
      "learning_rate": 0.00041143825167383073,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2245111122727394,
      "num_tokens": 59735486.0,
      "step": 26760
    },
    {
      "entropy": 5.656681251525879,
      "epoch": 3.00595271522435,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004114063246182863,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22068793922662736,
      "num_tokens": 59746006.0,
      "step": 26765
    },
    {
      "entropy": 5.614851951599121,
      "epoch": 3.0065142921323074,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041137439321951147,
      "loss": 4.9858,
      "mean_token_accuracy": 0.22162686735391618,
      "num_tokens": 59756346.0,
      "step": 26770
    },
    {
      "entropy": 5.514579486846924,
      "epoch": 3.007075869040265,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041134245747852314,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22775866091251373,
      "num_tokens": 59768038.0,
      "step": 26775
    },
    {
      "entropy": 5.609335947036743,
      "epoch": 3.0076374459482227,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004113105173963382,
      "loss": 4.992,
      "mean_token_accuracy": 0.22167981564998626,
      "num_tokens": 59779211.0,
      "step": 26780
    },
    {
      "entropy": 5.595012760162353,
      "epoch": 3.0081990228561803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004112785729739733,
      "loss": 4.9259,
      "mean_token_accuracy": 0.2248649999499321,
      "num_tokens": 59790777.0,
      "step": 26785
    },
    {
      "entropy": 5.595809602737427,
      "epoch": 3.0087605997641376,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041124662421244593,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21771931797266006,
      "num_tokens": 59802030.0,
      "step": 26790
    },
    {
      "entropy": 5.584261798858643,
      "epoch": 3.0093221766720952,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004112146711127732,
      "loss": 5.0144,
      "mean_token_accuracy": 0.21843783557415009,
      "num_tokens": 59814616.0,
      "step": 26795
    },
    {
      "entropy": 5.551241445541382,
      "epoch": 3.009883753580053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004111827136759725,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22612346559762955,
      "num_tokens": 59825797.0,
      "step": 26800
    },
    {
      "entropy": 5.52638726234436,
      "epoch": 3.0104453304880106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004111507519030613,
      "loss": 4.9775,
      "mean_token_accuracy": 0.22599918395280838,
      "num_tokens": 59836699.0,
      "step": 26805
    },
    {
      "entropy": 5.630412292480469,
      "epoch": 3.0110069073959678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004111187857950574,
      "loss": 5.006,
      "mean_token_accuracy": 0.22149114459753036,
      "num_tokens": 59848575.0,
      "step": 26810
    },
    {
      "entropy": 5.571379661560059,
      "epoch": 3.0115684843039254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004110868153529786,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23313240259885787,
      "num_tokens": 59860471.0,
      "step": 26815
    },
    {
      "entropy": 5.578571367263794,
      "epoch": 3.012130061211883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004110548405778428,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22667839676141738,
      "num_tokens": 59871759.0,
      "step": 26820
    },
    {
      "entropy": 5.616928911209106,
      "epoch": 3.0126916381198403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004110228614706681,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2214425891637802,
      "num_tokens": 59884375.0,
      "step": 26825
    },
    {
      "entropy": 5.60373706817627,
      "epoch": 3.013253215027798,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041099087803247276,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21849762201309203,
      "num_tokens": 59895606.0,
      "step": 26830
    },
    {
      "entropy": 5.594703626632691,
      "epoch": 3.0138147919357556,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004109588902642751,
      "loss": 5.001,
      "mean_token_accuracy": 0.2242535799741745,
      "num_tokens": 59908090.0,
      "step": 26835
    },
    {
      "entropy": 5.5913042545318605,
      "epoch": 3.0143763688437133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004109268981670936,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21997361183166503,
      "num_tokens": 59919218.0,
      "step": 26840
    },
    {
      "entropy": 5.646936988830566,
      "epoch": 3.0149379457516705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000410894901741947,
      "loss": 5.095,
      "mean_token_accuracy": 0.2203273668885231,
      "num_tokens": 59931195.0,
      "step": 26845
    },
    {
      "entropy": 5.598254585266114,
      "epoch": 3.015499522659628,
      "grad_norm": 1.015625,
      "learning_rate": 0.000410862900989854,
      "loss": 4.9893,
      "mean_token_accuracy": 0.22394278794527053,
      "num_tokens": 59943653.0,
      "step": 26850
    },
    {
      "entropy": 5.581298971176148,
      "epoch": 3.016061099567586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004108308959118336,
      "loss": 4.941,
      "mean_token_accuracy": 0.22091638445854186,
      "num_tokens": 59954677.0,
      "step": 26855
    },
    {
      "entropy": 5.628510999679565,
      "epoch": 3.0166226764755435,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004107988865089047,
      "loss": 5.0223,
      "mean_token_accuracy": 0.2202321320772171,
      "num_tokens": 59966135.0,
      "step": 26860
    },
    {
      "entropy": 5.533756875991822,
      "epoch": 3.0171842533835007,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041076687278208664,
      "loss": 4.8201,
      "mean_token_accuracy": 0.23404287099838256,
      "num_tokens": 59977828.0,
      "step": 26865
    },
    {
      "entropy": 5.5013174533844,
      "epoch": 3.0177458302914584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004107348547323987,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22732071429491044,
      "num_tokens": 59989226.0,
      "step": 26870
    },
    {
      "entropy": 5.598773002624512,
      "epoch": 3.018307407199416,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041070283236086024,
      "loss": 4.9356,
      "mean_token_accuracy": 0.23319011479616164,
      "num_tokens": 60000461.0,
      "step": 26875
    },
    {
      "entropy": 5.628406715393067,
      "epoch": 3.0188689841073737,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004106708056684911,
      "loss": 5.0945,
      "mean_token_accuracy": 0.22042309492826462,
      "num_tokens": 60012364.0,
      "step": 26880
    },
    {
      "entropy": 5.594395685195923,
      "epoch": 3.019430561015331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041063877465631084,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22683245837688445,
      "num_tokens": 60022913.0,
      "step": 26885
    },
    {
      "entropy": 5.588278722763062,
      "epoch": 3.0199921379232886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004106067393253394,
      "loss": 4.9507,
      "mean_token_accuracy": 0.22788898199796676,
      "num_tokens": 60034407.0,
      "step": 26890
    },
    {
      "entropy": 5.562199878692627,
      "epoch": 3.0205537148312462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004105746996765968,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21904523372650148,
      "num_tokens": 60044063.0,
      "step": 26895
    },
    {
      "entropy": 5.581045007705688,
      "epoch": 3.021115291739204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004105426557111031,
      "loss": 4.9077,
      "mean_token_accuracy": 0.2280976340174675,
      "num_tokens": 60054432.0,
      "step": 26900
    },
    {
      "entropy": 5.607355546951294,
      "epoch": 3.021676868647161,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004105106074298787,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22199981063604354,
      "num_tokens": 60065876.0,
      "step": 26905
    },
    {
      "entropy": 5.584501934051514,
      "epoch": 3.022238445555119,
      "grad_norm": 1.15625,
      "learning_rate": 0.000410478554833944,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22155465334653854,
      "num_tokens": 60076108.0,
      "step": 26910
    },
    {
      "entropy": 5.552416658401489,
      "epoch": 3.0228000224630764,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041044649792431963,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2254222959280014,
      "num_tokens": 60087386.0,
      "step": 26915
    },
    {
      "entropy": 5.617202806472778,
      "epoch": 3.0233615993710337,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041041443670202616,
      "loss": 4.9301,
      "mean_token_accuracy": 0.23225386887788774,
      "num_tokens": 60099017.0,
      "step": 26920
    },
    {
      "entropy": 5.573901557922364,
      "epoch": 3.0239231762789913,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004103823711680846,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22734050303697587,
      "num_tokens": 60110676.0,
      "step": 26925
    },
    {
      "entropy": 5.485046529769898,
      "epoch": 3.024484753186949,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041035030132351575,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22829129844903945,
      "num_tokens": 60120807.0,
      "step": 26930
    },
    {
      "entropy": 5.615556097030639,
      "epoch": 3.0250463300949066,
      "grad_norm": 1.125,
      "learning_rate": 0.00041031822716934086,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22648153305053711,
      "num_tokens": 60132174.0,
      "step": 26935
    },
    {
      "entropy": 5.662944507598877,
      "epoch": 3.025607907002864,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041028614870658106,
      "loss": 5.0704,
      "mean_token_accuracy": 0.22019724994897844,
      "num_tokens": 60144311.0,
      "step": 26940
    },
    {
      "entropy": 5.608879804611206,
      "epoch": 3.0261694839108215,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004102540659362579,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22396618872880936,
      "num_tokens": 60155368.0,
      "step": 26945
    },
    {
      "entropy": 5.640507459640503,
      "epoch": 3.026731060818779,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041022197885939285,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2184576079249382,
      "num_tokens": 60167633.0,
      "step": 26950
    },
    {
      "entropy": 5.540636348724365,
      "epoch": 3.027292637726737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041018988747700745,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23219280689954758,
      "num_tokens": 60180596.0,
      "step": 26955
    },
    {
      "entropy": 5.6685950756073,
      "epoch": 3.027854214634694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004101577917901236,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21951965540647506,
      "num_tokens": 60191811.0,
      "step": 26960
    },
    {
      "entropy": 5.683190584182739,
      "epoch": 3.0284157915426517,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041012569179976326,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2252937525510788,
      "num_tokens": 60202575.0,
      "step": 26965
    },
    {
      "entropy": 5.5764542579650875,
      "epoch": 3.0289773684506094,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004100935875069485,
      "loss": 4.9095,
      "mean_token_accuracy": 0.22569277435541152,
      "num_tokens": 60212541.0,
      "step": 26970
    },
    {
      "entropy": 5.587900972366333,
      "epoch": 3.029538945358567,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004100614789127015,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2298298180103302,
      "num_tokens": 60223018.0,
      "step": 26975
    },
    {
      "entropy": 5.6346477508544925,
      "epoch": 3.0301005222665243,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004100293660180447,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21820174306631088,
      "num_tokens": 60234104.0,
      "step": 26980
    },
    {
      "entropy": 5.5943248748779295,
      "epoch": 3.030662099174482,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040999724882400046,
      "loss": 4.9697,
      "mean_token_accuracy": 0.21949550062417983,
      "num_tokens": 60245119.0,
      "step": 26985
    },
    {
      "entropy": 5.592558240890503,
      "epoch": 3.0312236760824396,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004099651273315915,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22859519869089126,
      "num_tokens": 60255984.0,
      "step": 26990
    },
    {
      "entropy": 5.607346677780152,
      "epoch": 3.0317852529903973,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004099330015418405,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2295822784304619,
      "num_tokens": 60266764.0,
      "step": 26995
    },
    {
      "entropy": 5.553027105331421,
      "epoch": 3.0323468298983545,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040990087145577044,
      "loss": 4.929,
      "mean_token_accuracy": 0.2284087657928467,
      "num_tokens": 60278284.0,
      "step": 27000
    },
    {
      "epoch": 3.0323468298983545,
      "eval_entropy": 5.320054965010617,
      "eval_loss": 5.090383529663086,
      "eval_mean_token_accuracy": 0.22469001224176582,
      "eval_num_tokens": 60278284.0,
      "eval_runtime": 18.4951,
      "eval_samples_per_second": 1676.818,
      "eval_steps_per_second": 209.623,
      "step": 27000
    },
    {
      "entropy": 5.536643314361572,
      "epoch": 3.032908406806312,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004098687370744042,
      "loss": 4.8955,
      "mean_token_accuracy": 0.22871344089508056,
      "num_tokens": 60287993.0,
      "step": 27005
    },
    {
      "entropy": 5.635091829299927,
      "epoch": 3.03346998371427,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004098365983987652,
      "loss": 5.0634,
      "mean_token_accuracy": 0.21669960767030716,
      "num_tokens": 60300040.0,
      "step": 27010
    },
    {
      "entropy": 5.630225563049317,
      "epoch": 3.034031560622227,
      "grad_norm": 1.0,
      "learning_rate": 0.00040980445542987653,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22055701315402984,
      "num_tokens": 60311008.0,
      "step": 27015
    },
    {
      "entropy": 5.551601600646973,
      "epoch": 3.0345931375301847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040977230816876176,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22208404839038848,
      "num_tokens": 60321071.0,
      "step": 27020
    },
    {
      "entropy": 5.579111576080322,
      "epoch": 3.0351547144381423,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040974015661644433,
      "loss": 4.9502,
      "mean_token_accuracy": 0.22285210490226745,
      "num_tokens": 60333075.0,
      "step": 27025
    },
    {
      "entropy": 5.6268034934997555,
      "epoch": 3.0357162913461,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004097080007739481,
      "loss": 5.0235,
      "mean_token_accuracy": 0.2158128499984741,
      "num_tokens": 60343393.0,
      "step": 27030
    },
    {
      "entropy": 5.613195896148682,
      "epoch": 3.036277868254057,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040967584064229695,
      "loss": 5.0461,
      "mean_token_accuracy": 0.21175061911344528,
      "num_tokens": 60353981.0,
      "step": 27035
    },
    {
      "entropy": 5.614744663238525,
      "epoch": 3.036839445162015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040964367622251473,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22934891432523727,
      "num_tokens": 60364544.0,
      "step": 27040
    },
    {
      "entropy": 5.65191159248352,
      "epoch": 3.0374010220699725,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004096115075156257,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2254611998796463,
      "num_tokens": 60376031.0,
      "step": 27045
    },
    {
      "entropy": 5.595260953903198,
      "epoch": 3.03796259897793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004095793345226541,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21652601510286332,
      "num_tokens": 60387320.0,
      "step": 27050
    },
    {
      "entropy": 5.632719039916992,
      "epoch": 3.0385241758858874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004095471572446242,
      "loss": 4.9624,
      "mean_token_accuracy": 0.2218483045697212,
      "num_tokens": 60397345.0,
      "step": 27055
    },
    {
      "entropy": 5.597931814193726,
      "epoch": 3.039085752793845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004095149756825606,
      "loss": 5.0023,
      "mean_token_accuracy": 0.2208438351750374,
      "num_tokens": 60408194.0,
      "step": 27060
    },
    {
      "entropy": 5.643911409378052,
      "epoch": 3.0396473297018027,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004094827898374881,
      "loss": 5.0115,
      "mean_token_accuracy": 0.2174660861492157,
      "num_tokens": 60418928.0,
      "step": 27065
    },
    {
      "entropy": 5.544269561767578,
      "epoch": 3.0402089066097604,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004094505997104314,
      "loss": 4.8701,
      "mean_token_accuracy": 0.22558345943689345,
      "num_tokens": 60428819.0,
      "step": 27070
    },
    {
      "entropy": 5.48356990814209,
      "epoch": 3.0407704835177176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040941840530241537,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22819630205631256,
      "num_tokens": 60440575.0,
      "step": 27075
    },
    {
      "entropy": 5.603362655639648,
      "epoch": 3.0413320604256753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040938620661446526,
      "loss": 4.9877,
      "mean_token_accuracy": 0.21838736087083815,
      "num_tokens": 60452977.0,
      "step": 27080
    },
    {
      "entropy": 5.660623598098755,
      "epoch": 3.041893637333633,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004093540036476062,
      "loss": 4.9467,
      "mean_token_accuracy": 0.22112692147493362,
      "num_tokens": 60463552.0,
      "step": 27085
    },
    {
      "entropy": 5.642055988311768,
      "epoch": 3.0424552142415906,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040932179640286356,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21881585866212844,
      "num_tokens": 60474674.0,
      "step": 27090
    },
    {
      "entropy": 5.577290868759155,
      "epoch": 3.043016791149548,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004092895848812628,
      "loss": 4.9233,
      "mean_token_accuracy": 0.2317307099699974,
      "num_tokens": 60486140.0,
      "step": 27095
    },
    {
      "entropy": 5.562660551071167,
      "epoch": 3.0435783680575055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040925736908382953,
      "loss": 4.9295,
      "mean_token_accuracy": 0.23104745894670486,
      "num_tokens": 60497386.0,
      "step": 27100
    },
    {
      "entropy": 5.661973237991333,
      "epoch": 3.044139944965463,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004092251490115896,
      "loss": 5.074,
      "mean_token_accuracy": 0.21354657113552095,
      "num_tokens": 60508535.0,
      "step": 27105
    },
    {
      "entropy": 5.618186664581299,
      "epoch": 3.0447015218734204,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040919292466556895,
      "loss": 4.9942,
      "mean_token_accuracy": 0.21825923025608063,
      "num_tokens": 60520791.0,
      "step": 27110
    },
    {
      "entropy": 5.545917129516601,
      "epoch": 3.045263098781378,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004091606960467934,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22964697629213332,
      "num_tokens": 60533676.0,
      "step": 27115
    },
    {
      "entropy": 5.610679197311401,
      "epoch": 3.0458246756893357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040912846315628934,
      "loss": 4.9913,
      "mean_token_accuracy": 0.220120207965374,
      "num_tokens": 60545216.0,
      "step": 27120
    },
    {
      "entropy": 5.502521181106568,
      "epoch": 3.0463862525972933,
      "grad_norm": 1.046875,
      "learning_rate": 0.000409096225995083,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23004188388586044,
      "num_tokens": 60554730.0,
      "step": 27125
    },
    {
      "entropy": 5.466436719894409,
      "epoch": 3.0469478295052506,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040906398456420077,
      "loss": 4.8343,
      "mean_token_accuracy": 0.2267439365386963,
      "num_tokens": 60565494.0,
      "step": 27130
    },
    {
      "entropy": 5.54768009185791,
      "epoch": 3.047509406413208,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040903173886466934,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22437616884708406,
      "num_tokens": 60576249.0,
      "step": 27135
    },
    {
      "entropy": 5.5391517162323,
      "epoch": 3.048070983321166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040899948889751525,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22381734400987624,
      "num_tokens": 60587253.0,
      "step": 27140
    },
    {
      "entropy": 5.620519351959229,
      "epoch": 3.0486325602291235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040896723466376557,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21473684161901474,
      "num_tokens": 60598158.0,
      "step": 27145
    },
    {
      "entropy": 5.5933239459991455,
      "epoch": 3.0491941371370808,
      "grad_norm": 1.0,
      "learning_rate": 0.00040893497616444714,
      "loss": 4.8822,
      "mean_token_accuracy": 0.23179082274436952,
      "num_tokens": 60608812.0,
      "step": 27150
    },
    {
      "entropy": 5.583954524993897,
      "epoch": 3.0497557140450384,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004089027134005871,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2315080553293228,
      "num_tokens": 60619264.0,
      "step": 27155
    },
    {
      "entropy": 5.484221267700195,
      "epoch": 3.050317290952996,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004088704463732127,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23011974692344667,
      "num_tokens": 60629688.0,
      "step": 27160
    },
    {
      "entropy": 5.5685454368591305,
      "epoch": 3.0508788678609537,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004088381750833514,
      "loss": 4.9257,
      "mean_token_accuracy": 0.22430949807167053,
      "num_tokens": 60640790.0,
      "step": 27165
    },
    {
      "entropy": 5.5860882759094235,
      "epoch": 3.051440444768911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004088058995320307,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22561365365982056,
      "num_tokens": 60651952.0,
      "step": 27170
    },
    {
      "entropy": 5.657443428039551,
      "epoch": 3.0520020216768686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004087736197202784,
      "loss": 5.0372,
      "mean_token_accuracy": 0.22354716509580613,
      "num_tokens": 60665388.0,
      "step": 27175
    },
    {
      "entropy": 5.513414335250855,
      "epoch": 3.0525635985848263,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040874133564912195,
      "loss": 4.8656,
      "mean_token_accuracy": 0.2311892107129097,
      "num_tokens": 60676407.0,
      "step": 27180
    },
    {
      "entropy": 5.552424526214599,
      "epoch": 3.053125175492784,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040870904731958965,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22134810835123062,
      "num_tokens": 60686796.0,
      "step": 27185
    },
    {
      "entropy": 5.600311422348023,
      "epoch": 3.053686752400741,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004086767547327094,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22149686366319657,
      "num_tokens": 60697770.0,
      "step": 27190
    },
    {
      "entropy": 5.572322654724121,
      "epoch": 3.054248329308699,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040864445788950935,
      "loss": 4.9426,
      "mean_token_accuracy": 0.22640556544065477,
      "num_tokens": 60709059.0,
      "step": 27195
    },
    {
      "entropy": 5.536814546585083,
      "epoch": 3.0548099062166565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000408612156791018,
      "loss": 4.933,
      "mean_token_accuracy": 0.22578003108501435,
      "num_tokens": 60720228.0,
      "step": 27200
    },
    {
      "entropy": 5.653487062454223,
      "epoch": 3.0553714831246137,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004085798514382637,
      "loss": 5.0516,
      "mean_token_accuracy": 0.2210460975766182,
      "num_tokens": 60730658.0,
      "step": 27205
    },
    {
      "entropy": 5.617294502258301,
      "epoch": 3.0559330600325714,
      "grad_norm": 1.125,
      "learning_rate": 0.00040854754183227516,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2270521640777588,
      "num_tokens": 60741757.0,
      "step": 27210
    },
    {
      "entropy": 5.58541350364685,
      "epoch": 3.056494636940529,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004085152279740811,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22331530302762986,
      "num_tokens": 60755452.0,
      "step": 27215
    },
    {
      "entropy": 5.624424028396606,
      "epoch": 3.0570562138484867,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040848290986471037,
      "loss": 4.968,
      "mean_token_accuracy": 0.2214927226305008,
      "num_tokens": 60765715.0,
      "step": 27220
    },
    {
      "entropy": 5.622122573852539,
      "epoch": 3.057617790756444,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040845058750519205,
      "loss": 5.0359,
      "mean_token_accuracy": 0.220183564722538,
      "num_tokens": 60777021.0,
      "step": 27225
    },
    {
      "entropy": 5.535579490661621,
      "epoch": 3.0581793676644016,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040841826089655525,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22092769891023636,
      "num_tokens": 60787309.0,
      "step": 27230
    },
    {
      "entropy": 5.59045901298523,
      "epoch": 3.0587409445723592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040838593003982926,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2195867195725441,
      "num_tokens": 60797566.0,
      "step": 27235
    },
    {
      "entropy": 5.630866813659668,
      "epoch": 3.059302521480317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004083535949360436,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2271500676870346,
      "num_tokens": 60808213.0,
      "step": 27240
    },
    {
      "entropy": 5.628676080703736,
      "epoch": 3.059864098388274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004083212555862277,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21762911081314087,
      "num_tokens": 60820656.0,
      "step": 27245
    },
    {
      "entropy": 5.599404239654541,
      "epoch": 3.0604256752962318,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004082889119914113,
      "loss": 5.0232,
      "mean_token_accuracy": 0.22864385098218917,
      "num_tokens": 60831572.0,
      "step": 27250
    },
    {
      "entropy": 5.530362749099732,
      "epoch": 3.0609872522041894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004082565641526242,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23229837864637376,
      "num_tokens": 60842443.0,
      "step": 27255
    },
    {
      "entropy": 5.553641605377197,
      "epoch": 3.061548829112147,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040822421207089654,
      "loss": 4.891,
      "mean_token_accuracy": 0.22962461709976195,
      "num_tokens": 60853427.0,
      "step": 27260
    },
    {
      "entropy": 5.641279077529907,
      "epoch": 3.0621104060201043,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004081918557472582,
      "loss": 5.034,
      "mean_token_accuracy": 0.21973099410533906,
      "num_tokens": 60864517.0,
      "step": 27265
    },
    {
      "entropy": 5.610202169418335,
      "epoch": 3.062671982928062,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004081594951827396,
      "loss": 5.0004,
      "mean_token_accuracy": 0.22317808568477632,
      "num_tokens": 60874280.0,
      "step": 27270
    },
    {
      "entropy": 5.611862850189209,
      "epoch": 3.0632335598360196,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040812713037837096,
      "loss": 4.988,
      "mean_token_accuracy": 0.2223164767026901,
      "num_tokens": 60885656.0,
      "step": 27275
    },
    {
      "entropy": 5.637243318557739,
      "epoch": 3.0637951367439773,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040809476133518284,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2281840205192566,
      "num_tokens": 60896856.0,
      "step": 27280
    },
    {
      "entropy": 5.495713329315185,
      "epoch": 3.0643567136519345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040806238805420595,
      "loss": 4.9148,
      "mean_token_accuracy": 0.2278126984834671,
      "num_tokens": 60907464.0,
      "step": 27285
    },
    {
      "entropy": 5.595905923843384,
      "epoch": 3.064918290559892,
      "grad_norm": 1.125,
      "learning_rate": 0.000408030010536471,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21881770193576813,
      "num_tokens": 60918321.0,
      "step": 27290
    },
    {
      "entropy": 5.657015991210938,
      "epoch": 3.06547986746785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040799762878300895,
      "loss": 5.0184,
      "mean_token_accuracy": 0.21071440130472183,
      "num_tokens": 60928823.0,
      "step": 27295
    },
    {
      "entropy": 5.732596588134766,
      "epoch": 3.066041444375807,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040796524279485084,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2223851874470711,
      "num_tokens": 60940161.0,
      "step": 27300
    },
    {
      "entropy": 5.569212436676025,
      "epoch": 3.0666030212837647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040793285257302776,
      "loss": 5.0295,
      "mean_token_accuracy": 0.21862450540065764,
      "num_tokens": 60951968.0,
      "step": 27305
    },
    {
      "entropy": 5.633648490905761,
      "epoch": 3.0671645981917224,
      "grad_norm": 1.0625,
      "learning_rate": 0.000407900458118571,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21324760615825653,
      "num_tokens": 60963619.0,
      "step": 27310
    },
    {
      "entropy": 5.631049633026123,
      "epoch": 3.06772617509968,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004078680594325123,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2268174931406975,
      "num_tokens": 60974615.0,
      "step": 27315
    },
    {
      "entropy": 5.625506448745727,
      "epoch": 3.0682877520076373,
      "grad_norm": 1.09375,
      "learning_rate": 0.000407835656515883,
      "loss": 5.0485,
      "mean_token_accuracy": 0.21945664286613464,
      "num_tokens": 60986943.0,
      "step": 27320
    },
    {
      "entropy": 5.594842004776001,
      "epoch": 3.068849328915595,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040780324936971486,
      "loss": 5.07,
      "mean_token_accuracy": 0.22292616218328476,
      "num_tokens": 60998306.0,
      "step": 27325
    },
    {
      "entropy": 5.656533193588257,
      "epoch": 3.0694109058235526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004077708379950397,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21661709249019623,
      "num_tokens": 61010102.0,
      "step": 27330
    },
    {
      "entropy": 5.637247371673584,
      "epoch": 3.0699724827315102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004077384223928896,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22366156727075576,
      "num_tokens": 61019474.0,
      "step": 27335
    },
    {
      "entropy": 5.5437836170196535,
      "epoch": 3.0705340596394675,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004077060025642967,
      "loss": 4.9675,
      "mean_token_accuracy": 0.22760785073041917,
      "num_tokens": 61030401.0,
      "step": 27340
    },
    {
      "entropy": 5.677547121047974,
      "epoch": 3.071095636547425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004076735785102931,
      "loss": 5.0412,
      "mean_token_accuracy": 0.22089191079139708,
      "num_tokens": 61042139.0,
      "step": 27345
    },
    {
      "entropy": 5.56537184715271,
      "epoch": 3.071657213455383,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004076411502319114,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22324636876583098,
      "num_tokens": 61052468.0,
      "step": 27350
    },
    {
      "entropy": 5.613861751556397,
      "epoch": 3.0722187903633404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040760871773018405,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22577202767133714,
      "num_tokens": 61063625.0,
      "step": 27355
    },
    {
      "entropy": 5.506431341171265,
      "epoch": 3.0727803672712977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004075762810061437,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22708746790885925,
      "num_tokens": 61074246.0,
      "step": 27360
    },
    {
      "entropy": 5.637124967575073,
      "epoch": 3.0733419441792553,
      "grad_norm": 1.015625,
      "learning_rate": 0.000407543840060823,
      "loss": 5.0469,
      "mean_token_accuracy": 0.210322205722332,
      "num_tokens": 61086096.0,
      "step": 27365
    },
    {
      "entropy": 5.633491516113281,
      "epoch": 3.073903521087213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004075113948952551,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22437421530485152,
      "num_tokens": 61097176.0,
      "step": 27370
    },
    {
      "entropy": 5.703715038299561,
      "epoch": 3.0744650979951706,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004074789455104731,
      "loss": 5.0916,
      "mean_token_accuracy": 0.21944895386695862,
      "num_tokens": 61107865.0,
      "step": 27375
    },
    {
      "entropy": 5.61688265800476,
      "epoch": 3.075026674903128,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040744649190750995,
      "loss": 4.915,
      "mean_token_accuracy": 0.2219466969370842,
      "num_tokens": 61118682.0,
      "step": 27380
    },
    {
      "entropy": 5.602292203903199,
      "epoch": 3.0755882518110855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004074140340873992,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2356874853372574,
      "num_tokens": 61130210.0,
      "step": 27385
    },
    {
      "entropy": 5.542510223388672,
      "epoch": 3.076149828719043,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004073815720511743,
      "loss": 5.0123,
      "mean_token_accuracy": 0.22001791447401048,
      "num_tokens": 61141216.0,
      "step": 27390
    },
    {
      "entropy": 5.57162504196167,
      "epoch": 3.0767114056270004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040734910579986875,
      "loss": 5.0247,
      "mean_token_accuracy": 0.22108874768018721,
      "num_tokens": 61153233.0,
      "step": 27395
    },
    {
      "entropy": 5.609178447723389,
      "epoch": 3.077272982534958,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040731663533451624,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22266195118427276,
      "num_tokens": 61163044.0,
      "step": 27400
    },
    {
      "entropy": 5.560253858566284,
      "epoch": 3.0778345594429157,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040728416065615087,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22336438149213791,
      "num_tokens": 61174595.0,
      "step": 27405
    },
    {
      "entropy": 5.542100667953491,
      "epoch": 3.0783961363508734,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004072516817658065,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22478233575820922,
      "num_tokens": 61185215.0,
      "step": 27410
    },
    {
      "entropy": 5.497819709777832,
      "epoch": 3.0789577132588306,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040721919866451724,
      "loss": 4.8164,
      "mean_token_accuracy": 0.23512892425060272,
      "num_tokens": 61195504.0,
      "step": 27415
    },
    {
      "entropy": 5.473890686035157,
      "epoch": 3.0795192901667883,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004071867113533174,
      "loss": 4.8984,
      "mean_token_accuracy": 0.22756849974393845,
      "num_tokens": 61206040.0,
      "step": 27420
    },
    {
      "entropy": 5.627643775939942,
      "epoch": 3.080080867074746,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040715421983324147,
      "loss": 5.0283,
      "mean_token_accuracy": 0.22107081413269042,
      "num_tokens": 61217797.0,
      "step": 27425
    },
    {
      "entropy": 5.6685467720031735,
      "epoch": 3.0806424439827036,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040712172410532397,
      "loss": 5.0116,
      "mean_token_accuracy": 0.22408568412065505,
      "num_tokens": 61230158.0,
      "step": 27430
    },
    {
      "entropy": 5.661988019943237,
      "epoch": 3.081204020890661,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004070892241705994,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22732712775468827,
      "num_tokens": 61241072.0,
      "step": 27435
    },
    {
      "entropy": 5.509804105758667,
      "epoch": 3.0817655977986185,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040705672003010275,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22991017401218414,
      "num_tokens": 61251057.0,
      "step": 27440
    },
    {
      "entropy": 5.490853691101075,
      "epoch": 3.082327174706576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000407024211684869,
      "loss": 4.9097,
      "mean_token_accuracy": 0.23172446340322495,
      "num_tokens": 61262729.0,
      "step": 27445
    },
    {
      "entropy": 5.54977650642395,
      "epoch": 3.082888751614534,
      "grad_norm": 1.0,
      "learning_rate": 0.00040699169913593306,
      "loss": 4.9435,
      "mean_token_accuracy": 0.23059425950050355,
      "num_tokens": 61274665.0,
      "step": 27450
    },
    {
      "entropy": 5.626300382614136,
      "epoch": 3.083450328522491,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004069591823843302,
      "loss": 4.937,
      "mean_token_accuracy": 0.22472518235445021,
      "num_tokens": 61285244.0,
      "step": 27455
    },
    {
      "entropy": 5.574486970901489,
      "epoch": 3.0840119054304487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004069266614310959,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22071963250637056,
      "num_tokens": 61295739.0,
      "step": 27460
    },
    {
      "entropy": 5.558404874801636,
      "epoch": 3.0845734823384063,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040689413627726544,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22574560046195985,
      "num_tokens": 61305964.0,
      "step": 27465
    },
    {
      "entropy": 5.6564699649810795,
      "epoch": 3.085135059246364,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040686160692387455,
      "loss": 5.0483,
      "mean_token_accuracy": 0.22117441147565842,
      "num_tokens": 61317004.0,
      "step": 27470
    },
    {
      "entropy": 5.667507457733154,
      "epoch": 3.085696636154321,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040682907337195894,
      "loss": 5.0775,
      "mean_token_accuracy": 0.21187164038419723,
      "num_tokens": 61329014.0,
      "step": 27475
    },
    {
      "entropy": 5.558926582336426,
      "epoch": 3.086258213062279,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004067965356225545,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22168709188699723,
      "num_tokens": 61340522.0,
      "step": 27480
    },
    {
      "entropy": 5.638939619064331,
      "epoch": 3.0868197899702365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040676399367669733,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22271748930215834,
      "num_tokens": 61351595.0,
      "step": 27485
    },
    {
      "entropy": 5.647454404830933,
      "epoch": 3.0873813668781938,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004067314475354234,
      "loss": 5.0239,
      "mean_token_accuracy": 0.22501692771911622,
      "num_tokens": 61362432.0,
      "step": 27490
    },
    {
      "entropy": 5.571636629104614,
      "epoch": 3.0879429437861514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040669889719976916,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22430279850959778,
      "num_tokens": 61372717.0,
      "step": 27495
    },
    {
      "entropy": 5.612591791152954,
      "epoch": 3.088504520694109,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004066663426707709,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2222135469317436,
      "num_tokens": 61383150.0,
      "step": 27500
    },
    {
      "entropy": 5.608088493347168,
      "epoch": 3.0890660976020667,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040663378394946534,
      "loss": 5.0803,
      "mean_token_accuracy": 0.21098088026046752,
      "num_tokens": 61395538.0,
      "step": 27505
    },
    {
      "entropy": 5.627812194824219,
      "epoch": 3.089627674510024,
      "grad_norm": 1.125,
      "learning_rate": 0.0004066012210368889,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2224595218896866,
      "num_tokens": 61407523.0,
      "step": 27510
    },
    {
      "entropy": 5.587844896316528,
      "epoch": 3.0901892514179816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004065686539340786,
      "loss": 4.9333,
      "mean_token_accuracy": 0.2199319288134575,
      "num_tokens": 61418680.0,
      "step": 27515
    },
    {
      "entropy": 5.567046308517456,
      "epoch": 3.0907508283259393,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040653608264207134,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22757115513086318,
      "num_tokens": 61430006.0,
      "step": 27520
    },
    {
      "entropy": 5.482967185974121,
      "epoch": 3.091312405233897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040650350716190417,
      "loss": 4.7927,
      "mean_token_accuracy": 0.24096384942531585,
      "num_tokens": 61440564.0,
      "step": 27525
    },
    {
      "entropy": 5.480120754241943,
      "epoch": 3.091873982141854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004064709274946143,
      "loss": 4.7931,
      "mean_token_accuracy": 0.22896968126296996,
      "num_tokens": 61451589.0,
      "step": 27530
    },
    {
      "entropy": 5.521594762802124,
      "epoch": 3.092435559049812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004064383436412392,
      "loss": 4.8329,
      "mean_token_accuracy": 0.2346409261226654,
      "num_tokens": 61462177.0,
      "step": 27535
    },
    {
      "entropy": 5.634721755981445,
      "epoch": 3.0929971359577695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004064057556028162,
      "loss": 4.9628,
      "mean_token_accuracy": 0.22386689633131027,
      "num_tokens": 61472626.0,
      "step": 27540
    },
    {
      "entropy": 5.528603649139404,
      "epoch": 3.093558712865727,
      "grad_norm": 1.046875,
      "learning_rate": 0.000406373163380383,
      "loss": 4.8866,
      "mean_token_accuracy": 0.22752842009067537,
      "num_tokens": 61483839.0,
      "step": 27545
    },
    {
      "entropy": 5.543665885925293,
      "epoch": 3.0941202897736844,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040634056697497726,
      "loss": 4.974,
      "mean_token_accuracy": 0.23412736654281616,
      "num_tokens": 61495099.0,
      "step": 27550
    },
    {
      "entropy": 5.570893573760986,
      "epoch": 3.094681866681642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000406307966387637,
      "loss": 4.9639,
      "mean_token_accuracy": 0.22541098445653915,
      "num_tokens": 61506902.0,
      "step": 27555
    },
    {
      "entropy": 5.612701845169068,
      "epoch": 3.0952434435895997,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040627536161940016,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21987846344709397,
      "num_tokens": 61519357.0,
      "step": 27560
    },
    {
      "entropy": 5.587075805664062,
      "epoch": 3.0958050204975573,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004062427526713049,
      "loss": 4.8481,
      "mean_token_accuracy": 0.237555430829525,
      "num_tokens": 61529510.0,
      "step": 27565
    },
    {
      "entropy": 5.510145044326782,
      "epoch": 3.0963665974055146,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004062101395443894,
      "loss": 4.8202,
      "mean_token_accuracy": 0.2309665083885193,
      "num_tokens": 61539679.0,
      "step": 27570
    },
    {
      "entropy": 5.497973918914795,
      "epoch": 3.0969281743134722,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004061775222396922,
      "loss": 4.8884,
      "mean_token_accuracy": 0.2311049684882164,
      "num_tokens": 61549488.0,
      "step": 27575
    },
    {
      "entropy": 5.516869163513183,
      "epoch": 3.09748975122143,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004061449007582519,
      "loss": 4.955,
      "mean_token_accuracy": 0.22306123226881028,
      "num_tokens": 61561167.0,
      "step": 27580
    },
    {
      "entropy": 5.597123861312866,
      "epoch": 3.098051328129387,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000406112275101107,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2273518681526184,
      "num_tokens": 61572730.0,
      "step": 27585
    },
    {
      "entropy": 5.530239200592041,
      "epoch": 3.0986129050373448,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040607964526929655,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23198584765195845,
      "num_tokens": 61583846.0,
      "step": 27590
    },
    {
      "entropy": 5.569707870483398,
      "epoch": 3.0991744819453024,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040604701126385924,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2223135158419609,
      "num_tokens": 61594360.0,
      "step": 27595
    },
    {
      "entropy": 5.601286792755127,
      "epoch": 3.09973605885326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004060143730858342,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22885841727256775,
      "num_tokens": 61606750.0,
      "step": 27600
    },
    {
      "entropy": 5.568647575378418,
      "epoch": 3.1002976357612173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040598173073626086,
      "loss": 4.9944,
      "mean_token_accuracy": 0.20993604511022568,
      "num_tokens": 61618238.0,
      "step": 27605
    },
    {
      "entropy": 5.60656361579895,
      "epoch": 3.100859212669175,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004059490842161783,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2183505490422249,
      "num_tokens": 61629568.0,
      "step": 27610
    },
    {
      "entropy": 5.64889760017395,
      "epoch": 3.1014207895771326,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040591643352662615,
      "loss": 5.0322,
      "mean_token_accuracy": 0.22278825789690018,
      "num_tokens": 61640103.0,
      "step": 27615
    },
    {
      "entropy": 5.550869369506836,
      "epoch": 3.1019823664850903,
      "grad_norm": 1.171875,
      "learning_rate": 0.000405883778668644,
      "loss": 4.8674,
      "mean_token_accuracy": 0.23478101044893265,
      "num_tokens": 61651392.0,
      "step": 27620
    },
    {
      "entropy": 5.627930450439453,
      "epoch": 3.1025439433930475,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040585111964327157,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2136969417333603,
      "num_tokens": 61664932.0,
      "step": 27625
    },
    {
      "entropy": 5.530371856689453,
      "epoch": 3.103105520301005,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040581845645154877,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22446806877851486,
      "num_tokens": 61675719.0,
      "step": 27630
    },
    {
      "entropy": 5.639584732055664,
      "epoch": 3.103667097208963,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004057857890945155,
      "loss": 5.0562,
      "mean_token_accuracy": 0.22127631604671477,
      "num_tokens": 61687298.0,
      "step": 27635
    },
    {
      "entropy": 5.6594507694244385,
      "epoch": 3.1042286741169205,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040575311757321203,
      "loss": 5.03,
      "mean_token_accuracy": 0.2176799088716507,
      "num_tokens": 61697943.0,
      "step": 27640
    },
    {
      "entropy": 5.621898794174195,
      "epoch": 3.1047902510248777,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040572044188867863,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22157568484544754,
      "num_tokens": 61708897.0,
      "step": 27645
    },
    {
      "entropy": 5.566775465011597,
      "epoch": 3.1053518279328354,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004056877620419556,
      "loss": 5.0745,
      "mean_token_accuracy": 0.22475274950265883,
      "num_tokens": 61720409.0,
      "step": 27650
    },
    {
      "entropy": 5.658451271057129,
      "epoch": 3.105913404840793,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004056550780340835,
      "loss": 4.9818,
      "mean_token_accuracy": 0.22224654853343964,
      "num_tokens": 61731378.0,
      "step": 27655
    },
    {
      "entropy": 5.608495235443115,
      "epoch": 3.1064749817487507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004056223898661031,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22313157171010972,
      "num_tokens": 61742713.0,
      "step": 27660
    },
    {
      "entropy": 5.618476247787475,
      "epoch": 3.107036558656708,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040558969753905514,
      "loss": 5.0075,
      "mean_token_accuracy": 0.22490881085395814,
      "num_tokens": 61753561.0,
      "step": 27665
    },
    {
      "entropy": 5.600550508499145,
      "epoch": 3.1075981355646656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004055570010539806,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2214084655046463,
      "num_tokens": 61764966.0,
      "step": 27670
    },
    {
      "entropy": 5.62105975151062,
      "epoch": 3.1081597124726232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004055243004119204,
      "loss": 5.0092,
      "mean_token_accuracy": 0.21095353960990906,
      "num_tokens": 61776988.0,
      "step": 27675
    },
    {
      "entropy": 5.617335414886474,
      "epoch": 3.108721289380581,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040549159561391595,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21865565478801727,
      "num_tokens": 61788327.0,
      "step": 27680
    },
    {
      "entropy": 5.543080568313599,
      "epoch": 3.109282866288538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004054588866610085,
      "loss": 4.8786,
      "mean_token_accuracy": 0.23291634023189545,
      "num_tokens": 61799872.0,
      "step": 27685
    },
    {
      "entropy": 5.5314737319946286,
      "epoch": 3.1098444431964958,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040542617355423947,
      "loss": 4.942,
      "mean_token_accuracy": 0.2249660983681679,
      "num_tokens": 61810736.0,
      "step": 27690
    },
    {
      "entropy": 5.628012704849243,
      "epoch": 3.1104060201044534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040539345629465043,
      "loss": 5.0514,
      "mean_token_accuracy": 0.22055621594190597,
      "num_tokens": 61821236.0,
      "step": 27695
    },
    {
      "entropy": 5.71139760017395,
      "epoch": 3.1109675970124107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040536073488328325,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21921157985925674,
      "num_tokens": 61832282.0,
      "step": 27700
    },
    {
      "entropy": 5.7147626876831055,
      "epoch": 3.1115291739203683,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004053280093211796,
      "loss": 5.0465,
      "mean_token_accuracy": 0.21290226131677628,
      "num_tokens": 61842376.0,
      "step": 27705
    },
    {
      "entropy": 5.591224813461304,
      "epoch": 3.112090750828326,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004052952796093817,
      "loss": 4.9873,
      "mean_token_accuracy": 0.22302421033382416,
      "num_tokens": 61852227.0,
      "step": 27710
    },
    {
      "entropy": 5.646606922149658,
      "epoch": 3.1126523277362836,
      "grad_norm": 1.125,
      "learning_rate": 0.00040526254574893154,
      "loss": 5.026,
      "mean_token_accuracy": 0.22457720786333085,
      "num_tokens": 61863119.0,
      "step": 27715
    },
    {
      "entropy": 5.569517374038696,
      "epoch": 3.113213904644241,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040522980774087143,
      "loss": 4.916,
      "mean_token_accuracy": 0.22283942699432374,
      "num_tokens": 61874036.0,
      "step": 27720
    },
    {
      "entropy": 5.484189081192016,
      "epoch": 3.1137754815521985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004051970655862437,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23022659718990326,
      "num_tokens": 61884744.0,
      "step": 27725
    },
    {
      "entropy": 5.539526510238647,
      "epoch": 3.114337058460156,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004051643192860908,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21934515982866287,
      "num_tokens": 61896866.0,
      "step": 27730
    },
    {
      "entropy": 5.595732879638672,
      "epoch": 3.114898635368114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040513156884145564,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22646194994449614,
      "num_tokens": 61907384.0,
      "step": 27735
    },
    {
      "entropy": 5.4654113292694095,
      "epoch": 3.115460212276071,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004050988142533808,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23183571994304658,
      "num_tokens": 61917719.0,
      "step": 27740
    },
    {
      "entropy": 5.549846076965332,
      "epoch": 3.1160217891840287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004050660555229093,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22829806804656982,
      "num_tokens": 61928159.0,
      "step": 27745
    },
    {
      "entropy": 5.563377809524536,
      "epoch": 3.1165833660919864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004050332926510841,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2270898014307022,
      "num_tokens": 61938748.0,
      "step": 27750
    },
    {
      "entropy": 5.488481712341309,
      "epoch": 3.117144942999944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004050005256389484,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23569443672895432,
      "num_tokens": 61950429.0,
      "step": 27755
    },
    {
      "entropy": 5.55371732711792,
      "epoch": 3.1177065199079013,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004049677544875456,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22379663288593293,
      "num_tokens": 61961156.0,
      "step": 27760
    },
    {
      "entropy": 5.559010314941406,
      "epoch": 3.118268096815859,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040493497919791906,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22323964834213256,
      "num_tokens": 61972673.0,
      "step": 27765
    },
    {
      "entropy": 5.569641590118408,
      "epoch": 3.1188296737238166,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040490219977111237,
      "loss": 4.898,
      "mean_token_accuracy": 0.22608653455972672,
      "num_tokens": 61984481.0,
      "step": 27770
    },
    {
      "entropy": 5.576093339920044,
      "epoch": 3.1193912506317742,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004048694162081693,
      "loss": 4.9273,
      "mean_token_accuracy": 0.224456325173378,
      "num_tokens": 61994919.0,
      "step": 27775
    },
    {
      "entropy": 5.578846502304077,
      "epoch": 3.1199528275397315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004048366285101336,
      "loss": 4.9678,
      "mean_token_accuracy": 0.2260999247431755,
      "num_tokens": 62005303.0,
      "step": 27780
    },
    {
      "entropy": 5.524591493606567,
      "epoch": 3.120514404447689,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004048038366780494,
      "loss": 4.7988,
      "mean_token_accuracy": 0.23282495886087418,
      "num_tokens": 62015104.0,
      "step": 27785
    },
    {
      "entropy": 5.51768012046814,
      "epoch": 3.121075981355647,
      "grad_norm": 1.125,
      "learning_rate": 0.0004047710407129606,
      "loss": 4.9007,
      "mean_token_accuracy": 0.21965146958827972,
      "num_tokens": 62025577.0,
      "step": 27790
    },
    {
      "entropy": 5.718463277816772,
      "epoch": 3.121637558263604,
      "grad_norm": 1.125,
      "learning_rate": 0.0004047382406159115,
      "loss": 5.1046,
      "mean_token_accuracy": 0.21266368180513381,
      "num_tokens": 62037001.0,
      "step": 27795
    },
    {
      "entropy": 5.660603094100952,
      "epoch": 3.1221991351715617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004047054363879466,
      "loss": 5.0163,
      "mean_token_accuracy": 0.2208824187517166,
      "num_tokens": 62049520.0,
      "step": 27800
    },
    {
      "entropy": 5.584346151351928,
      "epoch": 3.1227607120795193,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040467262803011025,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2254662960767746,
      "num_tokens": 62061650.0,
      "step": 27805
    },
    {
      "entropy": 5.589773225784302,
      "epoch": 3.123322288987477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040463981554344716,
      "loss": 5.0432,
      "mean_token_accuracy": 0.22179077416658402,
      "num_tokens": 62074433.0,
      "step": 27810
    },
    {
      "entropy": 5.619342231750489,
      "epoch": 3.123883865895434,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040460699892900197,
      "loss": 4.9698,
      "mean_token_accuracy": 0.23077136725187303,
      "num_tokens": 62084885.0,
      "step": 27815
    },
    {
      "entropy": 5.615165042877197,
      "epoch": 3.124445442803392,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004045741781878198,
      "loss": 4.991,
      "mean_token_accuracy": 0.22212199866771698,
      "num_tokens": 62096847.0,
      "step": 27820
    },
    {
      "entropy": 5.6312590599060055,
      "epoch": 3.1250070197113495,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040454135332094543,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21971203237771988,
      "num_tokens": 62108131.0,
      "step": 27825
    },
    {
      "entropy": 5.57465705871582,
      "epoch": 3.125568596619307,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004045085243294242,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22497215121984482,
      "num_tokens": 62118450.0,
      "step": 27830
    },
    {
      "entropy": 5.581702089309692,
      "epoch": 3.1261301735272644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004044756912143013,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21654727309942245,
      "num_tokens": 62130702.0,
      "step": 27835
    },
    {
      "entropy": 5.610519599914551,
      "epoch": 3.126691750435222,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040444285397662225,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2184680387377739,
      "num_tokens": 62142863.0,
      "step": 27840
    },
    {
      "entropy": 5.580997514724731,
      "epoch": 3.1272533273431797,
      "grad_norm": 1.0,
      "learning_rate": 0.0004044100126174324,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22229348570108415,
      "num_tokens": 62153787.0,
      "step": 27845
    },
    {
      "entropy": 5.649285888671875,
      "epoch": 3.1278149042511374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040437716713777756,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21836046427488326,
      "num_tokens": 62166862.0,
      "step": 27850
    },
    {
      "entropy": 5.610101509094238,
      "epoch": 3.1283764811590946,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004043443175387036,
      "loss": 4.9677,
      "mean_token_accuracy": 0.22507266998291015,
      "num_tokens": 62177431.0,
      "step": 27855
    },
    {
      "entropy": 5.543297624588012,
      "epoch": 3.1289380580670523,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040431146382125644,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22473741620779036,
      "num_tokens": 62189465.0,
      "step": 27860
    },
    {
      "entropy": 5.5632177829742435,
      "epoch": 3.12949963497501,
      "grad_norm": 1.15625,
      "learning_rate": 0.000404278605986482,
      "loss": 4.8691,
      "mean_token_accuracy": 0.2317176267504692,
      "num_tokens": 62201327.0,
      "step": 27865
    },
    {
      "entropy": 5.565255737304687,
      "epoch": 3.1300612118829676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004042457440354267,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22197538763284683,
      "num_tokens": 62212689.0,
      "step": 27870
    },
    {
      "entropy": 5.542959499359131,
      "epoch": 3.130622788790925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004042128779691368,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22105631679296495,
      "num_tokens": 62223771.0,
      "step": 27875
    },
    {
      "entropy": 5.628036785125732,
      "epoch": 3.1311843656988825,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004041800077886587,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22232648581266404,
      "num_tokens": 62234137.0,
      "step": 27880
    },
    {
      "entropy": 5.615486907958984,
      "epoch": 3.13174594260684,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040414713349503907,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21978787034749986,
      "num_tokens": 62244978.0,
      "step": 27885
    },
    {
      "entropy": 5.643204164505005,
      "epoch": 3.1323075195147974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040411425508932457,
      "loss": 5.02,
      "mean_token_accuracy": 0.22021542638540267,
      "num_tokens": 62257364.0,
      "step": 27890
    },
    {
      "entropy": 5.564228534698486,
      "epoch": 3.132869096422755,
      "grad_norm": 1.125,
      "learning_rate": 0.00040408137257256225,
      "loss": 4.8366,
      "mean_token_accuracy": 0.22904119938611983,
      "num_tokens": 62268135.0,
      "step": 27895
    },
    {
      "entropy": 5.556001281738281,
      "epoch": 3.1334306733307127,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004040484859457988,
      "loss": 5.0106,
      "mean_token_accuracy": 0.22657999396324158,
      "num_tokens": 62279096.0,
      "step": 27900
    },
    {
      "entropy": 5.53884482383728,
      "epoch": 3.1339922502386703,
      "grad_norm": 1.125,
      "learning_rate": 0.00040401559521008155,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21969729214906691,
      "num_tokens": 62289640.0,
      "step": 27905
    },
    {
      "entropy": 5.642285966873169,
      "epoch": 3.1345538271466276,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040398270036645775,
      "loss": 5.0799,
      "mean_token_accuracy": 0.2132007971405983,
      "num_tokens": 62301085.0,
      "step": 27910
    },
    {
      "entropy": 5.505818605422974,
      "epoch": 3.135115404054585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004039498014159748,
      "loss": 4.8327,
      "mean_token_accuracy": 0.2279985159635544,
      "num_tokens": 62312667.0,
      "step": 27915
    },
    {
      "entropy": 5.625954055786133,
      "epoch": 3.135676980962543,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040391689835968004,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21820063292980194,
      "num_tokens": 62324640.0,
      "step": 27920
    },
    {
      "entropy": 5.634156942367554,
      "epoch": 3.1362385578705005,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004038839911986213,
      "loss": 4.9838,
      "mean_token_accuracy": 0.22492291778326035,
      "num_tokens": 62335940.0,
      "step": 27925
    },
    {
      "entropy": 5.673445081710815,
      "epoch": 3.1368001347784578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004038510799338462,
      "loss": 5.0185,
      "mean_token_accuracy": 0.22131550163030625,
      "num_tokens": 62346745.0,
      "step": 27930
    },
    {
      "entropy": 5.648662376403808,
      "epoch": 3.1373617116864154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004038181645664029,
      "loss": 4.9946,
      "mean_token_accuracy": 0.2207603707909584,
      "num_tokens": 62357694.0,
      "step": 27935
    },
    {
      "entropy": 5.664141654968262,
      "epoch": 3.137923288594373,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040378524509733916,
      "loss": 5.101,
      "mean_token_accuracy": 0.2127330034971237,
      "num_tokens": 62370621.0,
      "step": 27940
    },
    {
      "entropy": 5.597820138931274,
      "epoch": 3.1384848655023307,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004037523215277033,
      "loss": 4.9661,
      "mean_token_accuracy": 0.2210215613245964,
      "num_tokens": 62381248.0,
      "step": 27945
    },
    {
      "entropy": 5.676266145706177,
      "epoch": 3.139046442410288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004037193938585436,
      "loss": 4.979,
      "mean_token_accuracy": 0.21832115799188614,
      "num_tokens": 62392102.0,
      "step": 27950
    },
    {
      "entropy": 5.632308721542358,
      "epoch": 3.1396080193182456,
      "grad_norm": 1.0,
      "learning_rate": 0.0004036864620909085,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22537059634923934,
      "num_tokens": 62403254.0,
      "step": 27955
    },
    {
      "entropy": 5.564277076721192,
      "epoch": 3.1401695962262033,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004036535262258465,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23930185586214064,
      "num_tokens": 62413686.0,
      "step": 27960
    },
    {
      "entropy": 5.6123872756958,
      "epoch": 3.140731173134161,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040362058626440637,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22682064026594162,
      "num_tokens": 62424119.0,
      "step": 27965
    },
    {
      "entropy": 5.5764953136444095,
      "epoch": 3.141292750042118,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040358764220763685,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22524441480636598,
      "num_tokens": 62434900.0,
      "step": 27970
    },
    {
      "entropy": 5.580014324188232,
      "epoch": 3.141854326950076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040355469405658687,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2303948387503624,
      "num_tokens": 62446766.0,
      "step": 27975
    },
    {
      "entropy": 5.57128529548645,
      "epoch": 3.1424159038580335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004035217418123056,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22508161664009094,
      "num_tokens": 62457863.0,
      "step": 27980
    },
    {
      "entropy": 5.6430917263031,
      "epoch": 3.1429774807659907,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040348878547584227,
      "loss": 5.0626,
      "mean_token_accuracy": 0.21412102580070497,
      "num_tokens": 62470380.0,
      "step": 27985
    },
    {
      "entropy": 5.55840277671814,
      "epoch": 3.1435390576739484,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040345582504824624,
      "loss": 4.9306,
      "mean_token_accuracy": 0.22271953672170638,
      "num_tokens": 62482135.0,
      "step": 27990
    },
    {
      "entropy": 5.721755504608154,
      "epoch": 3.144100634581906,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004034228605305668,
      "loss": 5.1564,
      "mean_token_accuracy": 0.21153751015663147,
      "num_tokens": 62494409.0,
      "step": 27995
    },
    {
      "entropy": 5.691534805297851,
      "epoch": 3.1446622114898637,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040338989192385377,
      "loss": 5.0555,
      "mean_token_accuracy": 0.2173632875084877,
      "num_tokens": 62506247.0,
      "step": 28000
    },
    {
      "entropy": 5.708321237564087,
      "epoch": 3.145223788397821,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004033569192291567,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2121024951338768,
      "num_tokens": 62517580.0,
      "step": 28005
    },
    {
      "entropy": 5.625176954269409,
      "epoch": 3.1457853653057786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004033239424475256,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22136607468128205,
      "num_tokens": 62529086.0,
      "step": 28010
    },
    {
      "entropy": 5.527010726928711,
      "epoch": 3.1463469422137362,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040329096158001033,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22569512873888015,
      "num_tokens": 62540381.0,
      "step": 28015
    },
    {
      "entropy": 5.604122304916382,
      "epoch": 3.146908519121694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004032579766276612,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21780080050230027,
      "num_tokens": 62551756.0,
      "step": 28020
    },
    {
      "entropy": 5.654383373260498,
      "epoch": 3.147470096029651,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040322498759152824,
      "loss": 5.0238,
      "mean_token_accuracy": 0.21749861389398575,
      "num_tokens": 62563666.0,
      "step": 28025
    },
    {
      "entropy": 5.701459646224976,
      "epoch": 3.1480316729376088,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040319199447266205,
      "loss": 5.1174,
      "mean_token_accuracy": 0.21344360560178757,
      "num_tokens": 62576326.0,
      "step": 28030
    },
    {
      "entropy": 5.64227557182312,
      "epoch": 3.1485932498455664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004031589972721129,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22434263676404953,
      "num_tokens": 62586911.0,
      "step": 28035
    },
    {
      "entropy": 5.532308578491211,
      "epoch": 3.149154826753524,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004031259959909317,
      "loss": 4.9929,
      "mean_token_accuracy": 0.22304200381040573,
      "num_tokens": 62596694.0,
      "step": 28040
    },
    {
      "entropy": 5.658222484588623,
      "epoch": 3.1497164036614813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000403092990630169,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22339562326669693,
      "num_tokens": 62607576.0,
      "step": 28045
    },
    {
      "entropy": 5.5906905174255375,
      "epoch": 3.150277980569439,
      "grad_norm": 1.125,
      "learning_rate": 0.0004030599811908758,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22956023812294007,
      "num_tokens": 62618015.0,
      "step": 28050
    },
    {
      "entropy": 5.535467720031738,
      "epoch": 3.1508395574773966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004030269676741032,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22490967363119124,
      "num_tokens": 62629317.0,
      "step": 28055
    },
    {
      "entropy": 5.573162460327149,
      "epoch": 3.1514011343853543,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004029939500809023,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21625725477933883,
      "num_tokens": 62640077.0,
      "step": 28060
    },
    {
      "entropy": 5.800023937225342,
      "epoch": 3.1519627112933115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004029609284123243,
      "loss": 5.1643,
      "mean_token_accuracy": 0.21657926738262176,
      "num_tokens": 62652399.0,
      "step": 28065
    },
    {
      "entropy": 5.632557487487793,
      "epoch": 3.152524288201269,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040292790266942074,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22758538872003556,
      "num_tokens": 62663489.0,
      "step": 28070
    },
    {
      "entropy": 5.586713552474976,
      "epoch": 3.153085865109227,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040289487285324324,
      "loss": 5.0323,
      "mean_token_accuracy": 0.217813341319561,
      "num_tokens": 62674975.0,
      "step": 28075
    },
    {
      "entropy": 5.721684122085572,
      "epoch": 3.153647442017184,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004028618389648433,
      "loss": 5.1071,
      "mean_token_accuracy": 0.207653146982193,
      "num_tokens": 62686608.0,
      "step": 28080
    },
    {
      "entropy": 5.60950026512146,
      "epoch": 3.1542090189251417,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004028288010052728,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22086980789899827,
      "num_tokens": 62697844.0,
      "step": 28085
    },
    {
      "entropy": 5.517769765853882,
      "epoch": 3.1547705958330994,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004027957589755837,
      "loss": 4.901,
      "mean_token_accuracy": 0.22702545821666717,
      "num_tokens": 62709092.0,
      "step": 28090
    },
    {
      "entropy": 5.514061307907104,
      "epoch": 3.155332172741057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040276271287682813,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22163843214511872,
      "num_tokens": 62720047.0,
      "step": 28095
    },
    {
      "entropy": 5.636418962478638,
      "epoch": 3.1558937496490143,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040272966271005816,
      "loss": 4.9787,
      "mean_token_accuracy": 0.2247130513191223,
      "num_tokens": 62729952.0,
      "step": 28100
    },
    {
      "entropy": 5.663022708892822,
      "epoch": 3.156455326556972,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040269660847632626,
      "loss": 5.008,
      "mean_token_accuracy": 0.2184970647096634,
      "num_tokens": 62741961.0,
      "step": 28105
    },
    {
      "entropy": 5.583188676834107,
      "epoch": 3.1570169034649296,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040266355017668475,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22328823059797287,
      "num_tokens": 62753552.0,
      "step": 28110
    },
    {
      "entropy": 5.521012496948242,
      "epoch": 3.1575784803728872,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004026304878121864,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2327189177274704,
      "num_tokens": 62765239.0,
      "step": 28115
    },
    {
      "entropy": 5.516909170150757,
      "epoch": 3.1581400572808445,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040259742138388366,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22582112550735473,
      "num_tokens": 62776975.0,
      "step": 28120
    },
    {
      "entropy": 5.5180285453796385,
      "epoch": 3.158701634188802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040256435089282966,
      "loss": 4.9082,
      "mean_token_accuracy": 0.21965551227331162,
      "num_tokens": 62787435.0,
      "step": 28125
    },
    {
      "entropy": 5.577269697189331,
      "epoch": 3.15926321109676,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004025312763400772,
      "loss": 4.9586,
      "mean_token_accuracy": 0.2229569062590599,
      "num_tokens": 62799014.0,
      "step": 28130
    },
    {
      "entropy": 5.579418802261353,
      "epoch": 3.1598247880047174,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004024981977266794,
      "loss": 5.0298,
      "mean_token_accuracy": 0.2144472748041153,
      "num_tokens": 62810196.0,
      "step": 28135
    },
    {
      "entropy": 5.510231351852417,
      "epoch": 3.1603863649126747,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004024651150536896,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22330477684736252,
      "num_tokens": 62821416.0,
      "step": 28140
    },
    {
      "entropy": 5.5668562889099125,
      "epoch": 3.1609479418206323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004024320283221611,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22660741657018663,
      "num_tokens": 62831736.0,
      "step": 28145
    },
    {
      "entropy": 5.54722032546997,
      "epoch": 3.16150951872859,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040239893753314735,
      "loss": 4.8775,
      "mean_token_accuracy": 0.23188028186559678,
      "num_tokens": 62842017.0,
      "step": 28150
    },
    {
      "entropy": 5.642906951904297,
      "epoch": 3.1620710956365476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004023658426877021,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2125438556075096,
      "num_tokens": 62853594.0,
      "step": 28155
    },
    {
      "entropy": 5.584921550750733,
      "epoch": 3.162632672544505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040233274378687897,
      "loss": 4.8986,
      "mean_token_accuracy": 0.22650066763162613,
      "num_tokens": 62864831.0,
      "step": 28160
    },
    {
      "entropy": 5.723721551895141,
      "epoch": 3.1631942494524625,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040229964083173187,
      "loss": 5.0634,
      "mean_token_accuracy": 0.22132209986448287,
      "num_tokens": 62877557.0,
      "step": 28165
    },
    {
      "entropy": 5.584205436706543,
      "epoch": 3.16375582636042,
      "grad_norm": 1.125,
      "learning_rate": 0.00040226653382331484,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22227217704057695,
      "num_tokens": 62889162.0,
      "step": 28170
    },
    {
      "entropy": 5.499041938781739,
      "epoch": 3.1643174032683774,
      "grad_norm": 1.171875,
      "learning_rate": 0.000402233422762682,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22643920183181762,
      "num_tokens": 62899376.0,
      "step": 28175
    },
    {
      "entropy": 5.58070387840271,
      "epoch": 3.164878980176335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040220030765088764,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22601508796215058,
      "num_tokens": 62912448.0,
      "step": 28180
    },
    {
      "entropy": 5.565707206726074,
      "epoch": 3.1654405570842927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004021671884889862,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22553615868091584,
      "num_tokens": 62924021.0,
      "step": 28185
    },
    {
      "entropy": 5.6508325099945065,
      "epoch": 3.1660021339922504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004021340652780321,
      "loss": 5.0144,
      "mean_token_accuracy": 0.22226702123880387,
      "num_tokens": 62935330.0,
      "step": 28190
    },
    {
      "entropy": 5.599165534973144,
      "epoch": 3.1665637109002076,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004021009380190801,
      "loss": 4.949,
      "mean_token_accuracy": 0.22381659150123595,
      "num_tokens": 62947211.0,
      "step": 28195
    },
    {
      "entropy": 5.736557912826538,
      "epoch": 3.1671252878081653,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040206780671318495,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21668854653835296,
      "num_tokens": 62959012.0,
      "step": 28200
    },
    {
      "entropy": 5.648465538024903,
      "epoch": 3.167686864716123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004020346713614015,
      "loss": 5.0648,
      "mean_token_accuracy": 0.22376121282577516,
      "num_tokens": 62970608.0,
      "step": 28205
    },
    {
      "entropy": 5.644004392623901,
      "epoch": 3.1682484416240806,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040200153196478486,
      "loss": 5.0461,
      "mean_token_accuracy": 0.22215369045734407,
      "num_tokens": 62982613.0,
      "step": 28210
    },
    {
      "entropy": 5.51151762008667,
      "epoch": 3.168810018532038,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004019683885243902,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22447944134473802,
      "num_tokens": 62994394.0,
      "step": 28215
    },
    {
      "entropy": 5.605480527877807,
      "epoch": 3.1693715954399955,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004019352410412728,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2173373818397522,
      "num_tokens": 63007017.0,
      "step": 28220
    },
    {
      "entropy": 5.647416400909424,
      "epoch": 3.169933172347953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040190208951648807,
      "loss": 4.9632,
      "mean_token_accuracy": 0.22101624310016632,
      "num_tokens": 63018325.0,
      "step": 28225
    },
    {
      "entropy": 5.686427116394043,
      "epoch": 3.170494749255911,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004018689339510916,
      "loss": 5.0074,
      "mean_token_accuracy": 0.22214700877666474,
      "num_tokens": 63030694.0,
      "step": 28230
    },
    {
      "entropy": 5.565698194503784,
      "epoch": 3.171056326163868,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040183577434613904,
      "loss": 4.929,
      "mean_token_accuracy": 0.22448519617319107,
      "num_tokens": 63040889.0,
      "step": 28235
    },
    {
      "entropy": 5.504220819473266,
      "epoch": 3.1716179030718257,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004018026107026863,
      "loss": 4.8252,
      "mean_token_accuracy": 0.23131299614906312,
      "num_tokens": 63051385.0,
      "step": 28240
    },
    {
      "entropy": 5.557776689529419,
      "epoch": 3.1721794799797833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004017694430217892,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22168779373168945,
      "num_tokens": 63061989.0,
      "step": 28245
    },
    {
      "entropy": 5.664997911453247,
      "epoch": 3.172741056887741,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004017362713045039,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21617548614740373,
      "num_tokens": 63073031.0,
      "step": 28250
    },
    {
      "entropy": 5.586092329025268,
      "epoch": 3.173302633795698,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040170309555188655,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22863075882196426,
      "num_tokens": 63083674.0,
      "step": 28255
    },
    {
      "entropy": 5.548306465148926,
      "epoch": 3.173864210703656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004016699157649936,
      "loss": 4.939,
      "mean_token_accuracy": 0.23731489330530167,
      "num_tokens": 63094916.0,
      "step": 28260
    },
    {
      "entropy": 5.57163519859314,
      "epoch": 3.1744257876116135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004016367319448812,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21311400979757308,
      "num_tokens": 63106307.0,
      "step": 28265
    },
    {
      "entropy": 5.570258283615113,
      "epoch": 3.1749873645195708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040160354409260625,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2239832729101181,
      "num_tokens": 63115871.0,
      "step": 28270
    },
    {
      "entropy": 5.497352933883667,
      "epoch": 3.1755489414275284,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004015703522092254,
      "loss": 4.8836,
      "mean_token_accuracy": 0.2285323143005371,
      "num_tokens": 63126282.0,
      "step": 28275
    },
    {
      "entropy": 5.542381715774536,
      "epoch": 3.176110518335486,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004015371562957954,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2247241646051407,
      "num_tokens": 63138061.0,
      "step": 28280
    },
    {
      "entropy": 5.650219964981079,
      "epoch": 3.1766720952434437,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004015039563533732,
      "loss": 4.9862,
      "mean_token_accuracy": 0.2207380548119545,
      "num_tokens": 63148478.0,
      "step": 28285
    },
    {
      "entropy": 5.62716965675354,
      "epoch": 3.177233672151401,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004014707523830161,
      "loss": 5.001,
      "mean_token_accuracy": 0.2288699895143509,
      "num_tokens": 63161205.0,
      "step": 28290
    },
    {
      "entropy": 5.590086555480957,
      "epoch": 3.1777952490593586,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004014375443857811,
      "loss": 4.9426,
      "mean_token_accuracy": 0.22796505987644194,
      "num_tokens": 63173532.0,
      "step": 28295
    },
    {
      "entropy": 5.616058826446533,
      "epoch": 3.1783568259673163,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004014043323627257,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21776774078607558,
      "num_tokens": 63185254.0,
      "step": 28300
    },
    {
      "entropy": 5.674463367462158,
      "epoch": 3.178918402875274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004013711163149072,
      "loss": 5.0416,
      "mean_token_accuracy": 0.21476495414972305,
      "num_tokens": 63196481.0,
      "step": 28305
    },
    {
      "entropy": 5.5812735080719,
      "epoch": 3.179479979783231,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040133789624338353,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22745772451162338,
      "num_tokens": 63207411.0,
      "step": 28310
    },
    {
      "entropy": 5.578214406967163,
      "epoch": 3.180041556691189,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040130467214921216,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22127503007650376,
      "num_tokens": 63219097.0,
      "step": 28315
    },
    {
      "entropy": 5.640768671035767,
      "epoch": 3.1806031335991465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000401271444033451,
      "loss": 5.0169,
      "mean_token_accuracy": 0.2183619037270546,
      "num_tokens": 63230530.0,
      "step": 28320
    },
    {
      "entropy": 5.602543020248413,
      "epoch": 3.181164710507104,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004012382118971582,
      "loss": 4.9503,
      "mean_token_accuracy": 0.2334381103515625,
      "num_tokens": 63241918.0,
      "step": 28325
    },
    {
      "entropy": 5.66152400970459,
      "epoch": 3.1817262874150614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004012049757413917,
      "loss": 5.0568,
      "mean_token_accuracy": 0.21838285475969316,
      "num_tokens": 63253821.0,
      "step": 28330
    },
    {
      "entropy": 5.633574485778809,
      "epoch": 3.182287864323019,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040117173556720976,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22459098547697068,
      "num_tokens": 63265463.0,
      "step": 28335
    },
    {
      "entropy": 5.533801937103272,
      "epoch": 3.1828494412309767,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040113849137567094,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22860385477542877,
      "num_tokens": 63276010.0,
      "step": 28340
    },
    {
      "entropy": 5.539884185791015,
      "epoch": 3.1834110181389343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004011052431678336,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22055017054080964,
      "num_tokens": 63287179.0,
      "step": 28345
    },
    {
      "entropy": 5.592981958389283,
      "epoch": 3.1839725950468916,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004010719909447564,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22512660175561905,
      "num_tokens": 63298453.0,
      "step": 28350
    },
    {
      "entropy": 5.612341594696045,
      "epoch": 3.1845341719548492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004010387347074981,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22070101648569107,
      "num_tokens": 63310130.0,
      "step": 28355
    },
    {
      "entropy": 5.60241847038269,
      "epoch": 3.185095748862807,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040100547445711765,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22827959060668945,
      "num_tokens": 63319809.0,
      "step": 28360
    },
    {
      "entropy": 5.5884932518005375,
      "epoch": 3.185657325770764,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040097221019467393,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22006217241287232,
      "num_tokens": 63331263.0,
      "step": 28365
    },
    {
      "entropy": 5.602005672454834,
      "epoch": 3.1862189026787218,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004009389419212262,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22322552651166916,
      "num_tokens": 63341983.0,
      "step": 28370
    },
    {
      "entropy": 5.693082332611084,
      "epoch": 3.1867804795866794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040090566963783375,
      "loss": 4.982,
      "mean_token_accuracy": 0.22695057094097137,
      "num_tokens": 63352856.0,
      "step": 28375
    },
    {
      "entropy": 5.609148836135864,
      "epoch": 3.187342056494637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040087239334555603,
      "loss": 5.0141,
      "mean_token_accuracy": 0.22454948723316193,
      "num_tokens": 63364893.0,
      "step": 28380
    },
    {
      "entropy": 5.602074813842774,
      "epoch": 3.1879036334025943,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040083911304545234,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2255350261926651,
      "num_tokens": 63375222.0,
      "step": 28385
    },
    {
      "entropy": 5.503733396530151,
      "epoch": 3.188465210310552,
      "grad_norm": 1.0,
      "learning_rate": 0.00040080582873858254,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22422086000442504,
      "num_tokens": 63387056.0,
      "step": 28390
    },
    {
      "entropy": 5.620356273651123,
      "epoch": 3.1890267872185096,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004007725404260064,
      "loss": 5.021,
      "mean_token_accuracy": 0.2214551493525505,
      "num_tokens": 63397475.0,
      "step": 28395
    },
    {
      "entropy": 5.596714305877685,
      "epoch": 3.1895883641264673,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004007392481087837,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22579072117805482,
      "num_tokens": 63407920.0,
      "step": 28400
    },
    {
      "entropy": 5.535333871841431,
      "epoch": 3.1901499410344245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004007059517879747,
      "loss": 4.8786,
      "mean_token_accuracy": 0.23339123129844666,
      "num_tokens": 63418537.0,
      "step": 28405
    },
    {
      "entropy": 5.501137733459473,
      "epoch": 3.190711517942382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004006726514646393,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2241858348250389,
      "num_tokens": 63429844.0,
      "step": 28410
    },
    {
      "entropy": 5.6002137660980225,
      "epoch": 3.19127309485034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040063934713983803,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22269269227981567,
      "num_tokens": 63441417.0,
      "step": 28415
    },
    {
      "entropy": 5.535176849365234,
      "epoch": 3.1918346717582975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040060603881463127,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22341168224811553,
      "num_tokens": 63451772.0,
      "step": 28420
    },
    {
      "entropy": 5.556821966171265,
      "epoch": 3.1923962486662547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040057272649007944,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22586388438940047,
      "num_tokens": 63462683.0,
      "step": 28425
    },
    {
      "entropy": 5.660488605499268,
      "epoch": 3.1929578255742124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040053941016724326,
      "loss": 4.975,
      "mean_token_accuracy": 0.21928298622369766,
      "num_tokens": 63473627.0,
      "step": 28430
    },
    {
      "entropy": 5.618064117431641,
      "epoch": 3.19351940248217,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040050608984718364,
      "loss": 5.0529,
      "mean_token_accuracy": 0.21929014921188356,
      "num_tokens": 63485554.0,
      "step": 28435
    },
    {
      "entropy": 5.660331964492798,
      "epoch": 3.1940809793901277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004004727655309614,
      "loss": 5.0144,
      "mean_token_accuracy": 0.22489432692527772,
      "num_tokens": 63495881.0,
      "step": 28440
    },
    {
      "entropy": 5.653698348999024,
      "epoch": 3.194642556298085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040043943721963766,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21601244956254959,
      "num_tokens": 63506639.0,
      "step": 28445
    },
    {
      "entropy": 5.621528911590576,
      "epoch": 3.1952041332060426,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040040610491427357,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22117209285497666,
      "num_tokens": 63518482.0,
      "step": 28450
    },
    {
      "entropy": 5.5074108123779295,
      "epoch": 3.1957657101140002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004003727686159305,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22609093487262727,
      "num_tokens": 63529019.0,
      "step": 28455
    },
    {
      "entropy": 5.591524505615235,
      "epoch": 3.1963272870219575,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040033942832566984,
      "loss": 4.964,
      "mean_token_accuracy": 0.2272086814045906,
      "num_tokens": 63540330.0,
      "step": 28460
    },
    {
      "entropy": 5.530486440658569,
      "epoch": 3.196888863929915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004003060840445531,
      "loss": 4.962,
      "mean_token_accuracy": 0.21798104047775269,
      "num_tokens": 63552689.0,
      "step": 28465
    },
    {
      "entropy": 5.5713146209716795,
      "epoch": 3.1974504408378728,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004002727357736422,
      "loss": 4.9123,
      "mean_token_accuracy": 0.227738356590271,
      "num_tokens": 63562954.0,
      "step": 28470
    },
    {
      "entropy": 5.642990159988403,
      "epoch": 3.1980120177458304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040023938351399865,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2224881038069725,
      "num_tokens": 63574697.0,
      "step": 28475
    },
    {
      "entropy": 5.593879842758179,
      "epoch": 3.1985735946537877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004002060272666846,
      "loss": 4.9981,
      "mean_token_accuracy": 0.22885361313819885,
      "num_tokens": 63585060.0,
      "step": 28480
    },
    {
      "entropy": 5.592276000976563,
      "epoch": 3.1991351715617453,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004001726670327621,
      "loss": 4.9485,
      "mean_token_accuracy": 0.23258624374866485,
      "num_tokens": 63596749.0,
      "step": 28485
    },
    {
      "entropy": 5.5381083488464355,
      "epoch": 3.199696748469703,
      "grad_norm": 1.125,
      "learning_rate": 0.0004001393028132933,
      "loss": 4.81,
      "mean_token_accuracy": 0.23113590627908706,
      "num_tokens": 63607629.0,
      "step": 28490
    },
    {
      "entropy": 5.5493487358093265,
      "epoch": 3.2002583253776606,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040010593460934065,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22840212136507035,
      "num_tokens": 63617354.0,
      "step": 28495
    },
    {
      "entropy": 5.577769565582275,
      "epoch": 3.200819902285618,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040007256242196644,
      "loss": 4.9602,
      "mean_token_accuracy": 0.2236603945493698,
      "num_tokens": 63629320.0,
      "step": 28500
    },
    {
      "entropy": 5.625372695922851,
      "epoch": 3.2013814791935755,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040003918625223335,
      "loss": 4.9956,
      "mean_token_accuracy": 0.22030823826789855,
      "num_tokens": 63640453.0,
      "step": 28505
    },
    {
      "entropy": 5.575055027008057,
      "epoch": 3.201943056101533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040000580610120406,
      "loss": 4.9171,
      "mean_token_accuracy": 0.23098643720149994,
      "num_tokens": 63652484.0,
      "step": 28510
    },
    {
      "entropy": 5.543796062469482,
      "epoch": 3.202504633009491,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003999724219699415,
      "loss": 4.9859,
      "mean_token_accuracy": 0.23053297400474548,
      "num_tokens": 63663084.0,
      "step": 28515
    },
    {
      "entropy": 5.543668508529663,
      "epoch": 3.203066209917448,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003999390338595085,
      "loss": 4.949,
      "mean_token_accuracy": 0.22384461313486098,
      "num_tokens": 63674023.0,
      "step": 28520
    },
    {
      "entropy": 5.644239473342895,
      "epoch": 3.2036277868254057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039990564177096817,
      "loss": 4.957,
      "mean_token_accuracy": 0.2296536698937416,
      "num_tokens": 63685370.0,
      "step": 28525
    },
    {
      "entropy": 5.644373273849487,
      "epoch": 3.2041893637333634,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003998722457053838,
      "loss": 5.0363,
      "mean_token_accuracy": 0.22206113040447234,
      "num_tokens": 63696550.0,
      "step": 28530
    },
    {
      "entropy": 5.720673131942749,
      "epoch": 3.204750940641321,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039983884566381866,
      "loss": 5.0715,
      "mean_token_accuracy": 0.2196445047855377,
      "num_tokens": 63708040.0,
      "step": 28535
    },
    {
      "entropy": 5.622500991821289,
      "epoch": 3.2053125175492783,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003998054416473363,
      "loss": 5.0082,
      "mean_token_accuracy": 0.2167461857199669,
      "num_tokens": 63719704.0,
      "step": 28540
    },
    {
      "entropy": 5.568380975723267,
      "epoch": 3.205874094457236,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039977203365700026,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2208381325006485,
      "num_tokens": 63730111.0,
      "step": 28545
    },
    {
      "entropy": 5.629993963241577,
      "epoch": 3.2064356713651936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003997386216938743,
      "loss": 4.9641,
      "mean_token_accuracy": 0.23272769749164582,
      "num_tokens": 63741054.0,
      "step": 28550
    },
    {
      "entropy": 5.664008569717407,
      "epoch": 3.206997248273151,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039970520575902216,
      "loss": 4.9711,
      "mean_token_accuracy": 0.2203924164175987,
      "num_tokens": 63752598.0,
      "step": 28555
    },
    {
      "entropy": 5.600881862640381,
      "epoch": 3.2075588251811085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000399671785853508,
      "loss": 5.044,
      "mean_token_accuracy": 0.22754156589508057,
      "num_tokens": 63763927.0,
      "step": 28560
    },
    {
      "entropy": 5.603454446792602,
      "epoch": 3.208120402089066,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039963836197839584,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22780821472406387,
      "num_tokens": 63775072.0,
      "step": 28565
    },
    {
      "entropy": 5.58297963142395,
      "epoch": 3.208681978997024,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039960493413474977,
      "loss": 4.9496,
      "mean_token_accuracy": 0.225302354991436,
      "num_tokens": 63786044.0,
      "step": 28570
    },
    {
      "entropy": 5.566491651535034,
      "epoch": 3.209243555904981,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039957150232363433,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2282653346657753,
      "num_tokens": 63797889.0,
      "step": 28575
    },
    {
      "entropy": 5.647660875320435,
      "epoch": 3.2098051328129387,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000399538066546114,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2173111230134964,
      "num_tokens": 63809567.0,
      "step": 28580
    },
    {
      "entropy": 5.68712272644043,
      "epoch": 3.2103667097208963,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039950462680325336,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22525574862957,
      "num_tokens": 63819617.0,
      "step": 28585
    },
    {
      "entropy": 5.724773406982422,
      "epoch": 3.210928286628854,
      "grad_norm": 1.125,
      "learning_rate": 0.000399471183096117,
      "loss": 5.1367,
      "mean_token_accuracy": 0.21353906393051147,
      "num_tokens": 63831535.0,
      "step": 28590
    },
    {
      "entropy": 5.669536876678467,
      "epoch": 3.211489863536811,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003994377354257698,
      "loss": 5.056,
      "mean_token_accuracy": 0.22204603701829911,
      "num_tokens": 63842629.0,
      "step": 28595
    },
    {
      "entropy": 5.691928148269653,
      "epoch": 3.212051440444769,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039940428379327703,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2191132366657257,
      "num_tokens": 63853677.0,
      "step": 28600
    },
    {
      "entropy": 5.678381824493409,
      "epoch": 3.2126130173527265,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003993708281997036,
      "loss": 5.0264,
      "mean_token_accuracy": 0.2173156648874283,
      "num_tokens": 63865536.0,
      "step": 28605
    },
    {
      "entropy": 5.632109832763672,
      "epoch": 3.213174594260684,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003993373686461147,
      "loss": 4.9696,
      "mean_token_accuracy": 0.21566594541072845,
      "num_tokens": 63875849.0,
      "step": 28610
    },
    {
      "entropy": 5.616760015487671,
      "epoch": 3.2137361711686414,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003993039051335758,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2271122395992279,
      "num_tokens": 63887222.0,
      "step": 28615
    },
    {
      "entropy": 5.600552940368653,
      "epoch": 3.214297748076599,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003992704376631523,
      "loss": 4.9853,
      "mean_token_accuracy": 0.22412682473659515,
      "num_tokens": 63897969.0,
      "step": 28620
    },
    {
      "entropy": 5.727586364746093,
      "epoch": 3.2148593249845567,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003992369662359099,
      "loss": 5.0942,
      "mean_token_accuracy": 0.22126150578260423,
      "num_tokens": 63908913.0,
      "step": 28625
    },
    {
      "entropy": 5.628938484191894,
      "epoch": 3.2154209018925144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003992034908529143,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22184504568576813,
      "num_tokens": 63919269.0,
      "step": 28630
    },
    {
      "entropy": 5.53924822807312,
      "epoch": 3.2159824788004716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039917001151523144,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22092639058828353,
      "num_tokens": 63930624.0,
      "step": 28635
    },
    {
      "entropy": 5.56453914642334,
      "epoch": 3.2165440557084293,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003991365282239272,
      "loss": 4.9098,
      "mean_token_accuracy": 0.2246415227651596,
      "num_tokens": 63941571.0,
      "step": 28640
    },
    {
      "entropy": 5.647388219833374,
      "epoch": 3.217105632616387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039910304098006775,
      "loss": 5.0376,
      "mean_token_accuracy": 0.22143693417310714,
      "num_tokens": 63953221.0,
      "step": 28645
    },
    {
      "entropy": 5.630496692657471,
      "epoch": 3.217667209524344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003990695497847193,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22409473806619645,
      "num_tokens": 63965076.0,
      "step": 28650
    },
    {
      "entropy": 5.6235198974609375,
      "epoch": 3.218228786432302,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003990360546389484,
      "loss": 4.8886,
      "mean_token_accuracy": 0.2333245947957039,
      "num_tokens": 63976601.0,
      "step": 28655
    },
    {
      "entropy": 5.53252387046814,
      "epoch": 3.2187903633402595,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039900255554382126,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22885023653507233,
      "num_tokens": 63987757.0,
      "step": 28660
    },
    {
      "entropy": 5.612075138092041,
      "epoch": 3.219351940248217,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039896905250040473,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2212228924036026,
      "num_tokens": 64000104.0,
      "step": 28665
    },
    {
      "entropy": 5.6030192375183105,
      "epoch": 3.2199135171561744,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039893554550976545,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22951444387435913,
      "num_tokens": 64010773.0,
      "step": 28670
    },
    {
      "entropy": 5.565099906921387,
      "epoch": 3.220475094064132,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003989020345729705,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2279992401599884,
      "num_tokens": 64022269.0,
      "step": 28675
    },
    {
      "entropy": 5.581518697738647,
      "epoch": 3.2210366709720897,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039886851969108653,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22152754217386245,
      "num_tokens": 64033622.0,
      "step": 28680
    },
    {
      "entropy": 5.550294733047485,
      "epoch": 3.2215982478800473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039883500086518086,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22362757921218873,
      "num_tokens": 64044243.0,
      "step": 28685
    },
    {
      "entropy": 5.681491804122925,
      "epoch": 3.2221598247880046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039880147809632084,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22668851166963577,
      "num_tokens": 64054452.0,
      "step": 28690
    },
    {
      "entropy": 5.6074841022491455,
      "epoch": 3.222721401695962,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039876795138557363,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22375915944576263,
      "num_tokens": 64065743.0,
      "step": 28695
    },
    {
      "entropy": 5.650860977172852,
      "epoch": 3.22328297860392,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003987344207340069,
      "loss": 5.001,
      "mean_token_accuracy": 0.22270847260951995,
      "num_tokens": 64077359.0,
      "step": 28700
    },
    {
      "entropy": 5.58745551109314,
      "epoch": 3.2238445555118775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039870088614268824,
      "loss": 4.9166,
      "mean_token_accuracy": 0.22577347755432128,
      "num_tokens": 64088281.0,
      "step": 28705
    },
    {
      "entropy": 5.582028198242187,
      "epoch": 3.2244061324198348,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039866734761268534,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22200521677732468,
      "num_tokens": 64099600.0,
      "step": 28710
    },
    {
      "entropy": 5.665155363082886,
      "epoch": 3.2249677093277924,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039863380514506614,
      "loss": 4.9561,
      "mean_token_accuracy": 0.22521864473819733,
      "num_tokens": 64110381.0,
      "step": 28715
    },
    {
      "entropy": 5.635094118118286,
      "epoch": 3.22552928623575,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039860025874089865,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2222454681992531,
      "num_tokens": 64120878.0,
      "step": 28720
    },
    {
      "entropy": 5.498745250701904,
      "epoch": 3.2260908631437077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000398566708401251,
      "loss": 4.816,
      "mean_token_accuracy": 0.23892593383789062,
      "num_tokens": 64131427.0,
      "step": 28725
    },
    {
      "entropy": 5.601071977615357,
      "epoch": 3.226652440051665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039853315412719144,
      "loss": 5.0224,
      "mean_token_accuracy": 0.2234024554491043,
      "num_tokens": 64142833.0,
      "step": 28730
    },
    {
      "entropy": 5.534292984008789,
      "epoch": 3.2272140169596226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003984995959197883,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2282778099179268,
      "num_tokens": 64152405.0,
      "step": 28735
    },
    {
      "entropy": 5.626446056365967,
      "epoch": 3.2277755938675803,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003984660337801101,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2192511722445488,
      "num_tokens": 64164398.0,
      "step": 28740
    },
    {
      "entropy": 5.59745225906372,
      "epoch": 3.2283371707755375,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039843246770922557,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22242977768182753,
      "num_tokens": 64175313.0,
      "step": 28745
    },
    {
      "entropy": 5.469136381149292,
      "epoch": 3.228898747683495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003983988977082033,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22637999504804612,
      "num_tokens": 64185866.0,
      "step": 28750
    },
    {
      "entropy": 5.56924467086792,
      "epoch": 3.229460324591453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039836532377811237,
      "loss": 4.927,
      "mean_token_accuracy": 0.2260056659579277,
      "num_tokens": 64197126.0,
      "step": 28755
    },
    {
      "entropy": 5.541614770889282,
      "epoch": 3.2300219014994105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003983317459200217,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22285350710153579,
      "num_tokens": 64209271.0,
      "step": 28760
    },
    {
      "entropy": 5.6782200813293455,
      "epoch": 3.2305834784073677,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039829816413500035,
      "loss": 5.0361,
      "mean_token_accuracy": 0.2234062910079956,
      "num_tokens": 64221200.0,
      "step": 28765
    },
    {
      "entropy": 5.6506760597229,
      "epoch": 3.2311450553153254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003982645784241176,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22013672739267348,
      "num_tokens": 64230835.0,
      "step": 28770
    },
    {
      "entropy": 5.629516983032227,
      "epoch": 3.231706632223283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003982309887884429,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22835978865623474,
      "num_tokens": 64242243.0,
      "step": 28775
    },
    {
      "entropy": 5.609807729721069,
      "epoch": 3.2322682091312407,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039819739522904576,
      "loss": 4.9484,
      "mean_token_accuracy": 0.223925644159317,
      "num_tokens": 64253362.0,
      "step": 28780
    },
    {
      "entropy": 5.574964237213135,
      "epoch": 3.232829786039198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039816379774699573,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2242693930864334,
      "num_tokens": 64264462.0,
      "step": 28785
    },
    {
      "entropy": 5.561899042129516,
      "epoch": 3.2333913629471556,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039813019634336264,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22140322625637054,
      "num_tokens": 64275153.0,
      "step": 28790
    },
    {
      "entropy": 5.556550741195679,
      "epoch": 3.2339529398551132,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003980965910192163,
      "loss": 4.901,
      "mean_token_accuracy": 0.2312205418944359,
      "num_tokens": 64288082.0,
      "step": 28795
    },
    {
      "entropy": 5.574648237228393,
      "epoch": 3.234514516763071,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039806298177562675,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22341010570526124,
      "num_tokens": 64298872.0,
      "step": 28800
    },
    {
      "entropy": 5.646730661392212,
      "epoch": 3.235076093671028,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039802936861366416,
      "loss": 5.0647,
      "mean_token_accuracy": 0.21601880937814713,
      "num_tokens": 64310245.0,
      "step": 28805
    },
    {
      "entropy": 5.646114826202393,
      "epoch": 3.2356376705789858,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003979957515343988,
      "loss": 5.0233,
      "mean_token_accuracy": 0.22137523740530013,
      "num_tokens": 64323681.0,
      "step": 28810
    },
    {
      "entropy": 5.575125694274902,
      "epoch": 3.2361992474869434,
      "grad_norm": 1.0,
      "learning_rate": 0.000397962130538901,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2258685901761055,
      "num_tokens": 64335265.0,
      "step": 28815
    },
    {
      "entropy": 5.684977769851685,
      "epoch": 3.236760824394901,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039792850562824124,
      "loss": 5.0346,
      "mean_token_accuracy": 0.22200240045785904,
      "num_tokens": 64346279.0,
      "step": 28820
    },
    {
      "entropy": 5.646379804611206,
      "epoch": 3.2373224013028583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003978948768034902,
      "loss": 4.9583,
      "mean_token_accuracy": 0.2284569710493088,
      "num_tokens": 64357511.0,
      "step": 28825
    },
    {
      "entropy": 5.585288572311401,
      "epoch": 3.237883978210816,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039786124406571854,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2161759093403816,
      "num_tokens": 64367863.0,
      "step": 28830
    },
    {
      "entropy": 5.559604263305664,
      "epoch": 3.2384455551187736,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039782760741599734,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22239072471857071,
      "num_tokens": 64379746.0,
      "step": 28835
    },
    {
      "entropy": 5.60322151184082,
      "epoch": 3.239007132026731,
      "grad_norm": 1.125,
      "learning_rate": 0.0003977939668553973,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23087994307279586,
      "num_tokens": 64390370.0,
      "step": 28840
    },
    {
      "entropy": 5.642238521575928,
      "epoch": 3.2395687089346885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039776032238498995,
      "loss": 5.0817,
      "mean_token_accuracy": 0.21788007915019988,
      "num_tokens": 64402620.0,
      "step": 28845
    },
    {
      "entropy": 5.628843069076538,
      "epoch": 3.240130285842646,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003977266740058462,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2262054204940796,
      "num_tokens": 64412969.0,
      "step": 28850
    },
    {
      "entropy": 5.594276571273804,
      "epoch": 3.240691862750604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003976930217190376,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22305726259946823,
      "num_tokens": 64423679.0,
      "step": 28855
    },
    {
      "entropy": 5.658161449432373,
      "epoch": 3.241253439658561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003976593655256356,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22480060309171676,
      "num_tokens": 64436124.0,
      "step": 28860
    },
    {
      "entropy": 5.583728790283203,
      "epoch": 3.2418150165665187,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039762570542671176,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22403895109891891,
      "num_tokens": 64446494.0,
      "step": 28865
    },
    {
      "entropy": 5.599643754959106,
      "epoch": 3.2423765934744764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000397592041423338,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2273665577173233,
      "num_tokens": 64456152.0,
      "step": 28870
    },
    {
      "entropy": 5.653568267822266,
      "epoch": 3.242938170382434,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039755837351658605,
      "loss": 5.0095,
      "mean_token_accuracy": 0.2246701017022133,
      "num_tokens": 64468004.0,
      "step": 28875
    },
    {
      "entropy": 5.637023210525513,
      "epoch": 3.2434997472903913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003975247017075279,
      "loss": 4.954,
      "mean_token_accuracy": 0.22552973926067352,
      "num_tokens": 64478506.0,
      "step": 28880
    },
    {
      "entropy": 5.5599085807800295,
      "epoch": 3.244061324198349,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003974910259972358,
      "loss": 4.9246,
      "mean_token_accuracy": 0.22723468542098998,
      "num_tokens": 64489379.0,
      "step": 28885
    },
    {
      "entropy": 5.617467784881592,
      "epoch": 3.2446229011063066,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003974573463867819,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2248564749956131,
      "num_tokens": 64499500.0,
      "step": 28890
    },
    {
      "entropy": 5.547719144821167,
      "epoch": 3.2451844780142642,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039742366287723855,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23481586277484895,
      "num_tokens": 64510869.0,
      "step": 28895
    },
    {
      "entropy": 5.530608749389648,
      "epoch": 3.2457460549222215,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039738997546967834,
      "loss": 4.9867,
      "mean_token_accuracy": 0.2140473335981369,
      "num_tokens": 64521581.0,
      "step": 28900
    },
    {
      "entropy": 5.6370354175567625,
      "epoch": 3.246307631830179,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039735628416517394,
      "loss": 5.0497,
      "mean_token_accuracy": 0.22122990041971208,
      "num_tokens": 64532436.0,
      "step": 28905
    },
    {
      "entropy": 5.661991930007934,
      "epoch": 3.246869208738137,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039732258896479786,
      "loss": 5.0096,
      "mean_token_accuracy": 0.22046742886304854,
      "num_tokens": 64543546.0,
      "step": 28910
    },
    {
      "entropy": 5.695040941238403,
      "epoch": 3.2474307856460944,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003972888898696231,
      "loss": 5.1024,
      "mean_token_accuracy": 0.21267350763082504,
      "num_tokens": 64554964.0,
      "step": 28915
    },
    {
      "entropy": 5.624185991287232,
      "epoch": 3.2479923625540517,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039725518688072276,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22179786413908004,
      "num_tokens": 64564082.0,
      "step": 28920
    },
    {
      "entropy": 5.626805543899536,
      "epoch": 3.2485539394620093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039722147999916976,
      "loss": 5.0375,
      "mean_token_accuracy": 0.22143687307834625,
      "num_tokens": 64575377.0,
      "step": 28925
    },
    {
      "entropy": 5.648537349700928,
      "epoch": 3.249115516369967,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003971877692260375,
      "loss": 4.932,
      "mean_token_accuracy": 0.22759737074375153,
      "num_tokens": 64587686.0,
      "step": 28930
    },
    {
      "entropy": 5.5905115604400635,
      "epoch": 3.249677093277924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039715405456239924,
      "loss": 4.9108,
      "mean_token_accuracy": 0.22152551561594008,
      "num_tokens": 64598501.0,
      "step": 28935
    },
    {
      "entropy": 5.568253707885742,
      "epoch": 3.250238670185882,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003971203360093286,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22281139194965363,
      "num_tokens": 64609341.0,
      "step": 28940
    },
    {
      "entropy": 5.643636465072632,
      "epoch": 3.2508002470938395,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039708661356789896,
      "loss": 5.0059,
      "mean_token_accuracy": 0.22207978665828704,
      "num_tokens": 64621613.0,
      "step": 28945
    },
    {
      "entropy": 5.493619728088379,
      "epoch": 3.251361824001797,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039705288723918415,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22283975929021835,
      "num_tokens": 64633358.0,
      "step": 28950
    },
    {
      "entropy": 5.563966703414917,
      "epoch": 3.2519234009097544,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003970191570242582,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22391816079616547,
      "num_tokens": 64644799.0,
      "step": 28955
    },
    {
      "entropy": 5.5861677646636965,
      "epoch": 3.252484977817712,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000396985422924195,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23028914928436278,
      "num_tokens": 64657028.0,
      "step": 28960
    },
    {
      "entropy": 5.692339372634888,
      "epoch": 3.2530465547256697,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039695168494006857,
      "loss": 5.0505,
      "mean_token_accuracy": 0.21918956637382508,
      "num_tokens": 64668572.0,
      "step": 28965
    },
    {
      "entropy": 5.544000673294067,
      "epoch": 3.2536081316336274,
      "grad_norm": 1.125,
      "learning_rate": 0.0003969179430729531,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22875047475099564,
      "num_tokens": 64679168.0,
      "step": 28970
    },
    {
      "entropy": 5.6800580501556395,
      "epoch": 3.2541697085415846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003968841973239231,
      "loss": 5.0344,
      "mean_token_accuracy": 0.2215167000889778,
      "num_tokens": 64691571.0,
      "step": 28975
    },
    {
      "entropy": 5.680015230178833,
      "epoch": 3.2547312854495423,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000396850447694053,
      "loss": 5.0483,
      "mean_token_accuracy": 0.21842655241489412,
      "num_tokens": 64703659.0,
      "step": 28980
    },
    {
      "entropy": 5.610249328613281,
      "epoch": 3.2552928623575,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003968166941844174,
      "loss": 4.907,
      "mean_token_accuracy": 0.23488358557224273,
      "num_tokens": 64715121.0,
      "step": 28985
    },
    {
      "entropy": 5.583322238922119,
      "epoch": 3.2558544392654576,
      "grad_norm": 1.125,
      "learning_rate": 0.000396782936796091,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2232510045170784,
      "num_tokens": 64726289.0,
      "step": 28990
    },
    {
      "entropy": 5.588436698913574,
      "epoch": 3.256416016173415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003967491755301487,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22780743837356568,
      "num_tokens": 64737378.0,
      "step": 28995
    },
    {
      "entropy": 5.6883563041687015,
      "epoch": 3.2569775930813725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003967154103876654,
      "loss": 5.1335,
      "mean_token_accuracy": 0.21256226003170015,
      "num_tokens": 64749203.0,
      "step": 29000
    },
    {
      "entropy": 5.547046232223511,
      "epoch": 3.25753916998933,
      "grad_norm": 1.125,
      "learning_rate": 0.00039668164136971615,
      "loss": 4.8736,
      "mean_token_accuracy": 0.23236532956361772,
      "num_tokens": 64759514.0,
      "step": 29005
    },
    {
      "entropy": 5.609247350692749,
      "epoch": 3.258100746897288,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003966478684773763,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22604830265045167,
      "num_tokens": 64769370.0,
      "step": 29010
    },
    {
      "entropy": 5.588159894943237,
      "epoch": 3.258662323805245,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039661409171172117,
      "loss": 4.967,
      "mean_token_accuracy": 0.22442339658737182,
      "num_tokens": 64779930.0,
      "step": 29015
    },
    {
      "entropy": 5.579450273513794,
      "epoch": 3.2592239007132027,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039658031107382615,
      "loss": 5.0202,
      "mean_token_accuracy": 0.2163074195384979,
      "num_tokens": 64791726.0,
      "step": 29020
    },
    {
      "entropy": 5.527961874008179,
      "epoch": 3.2597854776211603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039654652656476695,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2375270813703537,
      "num_tokens": 64802778.0,
      "step": 29025
    },
    {
      "entropy": 5.587002754211426,
      "epoch": 3.2603470545291175,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003965127381856191,
      "loss": 5.031,
      "mean_token_accuracy": 0.2196357175707817,
      "num_tokens": 64814664.0,
      "step": 29030
    },
    {
      "entropy": 5.6229386806488035,
      "epoch": 3.260908631437075,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039647894593745865,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22663391828536988,
      "num_tokens": 64825775.0,
      "step": 29035
    },
    {
      "entropy": 5.665863513946533,
      "epoch": 3.261470208345033,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039644514982136136,
      "loss": 5.0633,
      "mean_token_accuracy": 0.22083175629377366,
      "num_tokens": 64837376.0,
      "step": 29040
    },
    {
      "entropy": 5.6541300296783445,
      "epoch": 3.2620317852529905,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039641134983840336,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22685634791851045,
      "num_tokens": 64847941.0,
      "step": 29045
    },
    {
      "entropy": 5.6587823867797855,
      "epoch": 3.2625933621609478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000396377545989661,
      "loss": 5.0542,
      "mean_token_accuracy": 0.2223596915602684,
      "num_tokens": 64858793.0,
      "step": 29050
    },
    {
      "entropy": 5.627451944351196,
      "epoch": 3.2631549390689054,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039634373827621047,
      "loss": 4.9458,
      "mean_token_accuracy": 0.22318787574768068,
      "num_tokens": 64870233.0,
      "step": 29055
    },
    {
      "entropy": 5.61488094329834,
      "epoch": 3.263716515976863,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003963099266991282,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22977886646986007,
      "num_tokens": 64881875.0,
      "step": 29060
    },
    {
      "entropy": 5.554671287536621,
      "epoch": 3.2642780928848207,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039627611125949086,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22426854968070983,
      "num_tokens": 64892837.0,
      "step": 29065
    },
    {
      "entropy": 5.700060176849365,
      "epoch": 3.264839669792778,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039624229195837506,
      "loss": 5.046,
      "mean_token_accuracy": 0.21601335257291793,
      "num_tokens": 64904773.0,
      "step": 29070
    },
    {
      "entropy": 5.618255138397217,
      "epoch": 3.2654012467007356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039620846879685767,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22034273147583008,
      "num_tokens": 64916596.0,
      "step": 29075
    },
    {
      "entropy": 5.581845664978028,
      "epoch": 3.2659628236086933,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039617464177601564,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22576479464769364,
      "num_tokens": 64927736.0,
      "step": 29080
    },
    {
      "entropy": 5.65763258934021,
      "epoch": 3.266524400516651,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039614081089692606,
      "loss": 5.062,
      "mean_token_accuracy": 0.21911101490259172,
      "num_tokens": 64940426.0,
      "step": 29085
    },
    {
      "entropy": 5.5837146759033205,
      "epoch": 3.267085977424608,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039610697616066604,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22637172192335128,
      "num_tokens": 64951034.0,
      "step": 29090
    },
    {
      "entropy": 5.676737976074219,
      "epoch": 3.267647554332566,
      "grad_norm": 1.125,
      "learning_rate": 0.00039607313756831296,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21598758846521376,
      "num_tokens": 64962900.0,
      "step": 29095
    },
    {
      "entropy": 5.584172630310059,
      "epoch": 3.2682091312405235,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039603929512094417,
      "loss": 4.9386,
      "mean_token_accuracy": 0.23082734346389772,
      "num_tokens": 64974820.0,
      "step": 29100
    },
    {
      "entropy": 5.525128602981567,
      "epoch": 3.268770708148481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003960054488196373,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2238277718424797,
      "num_tokens": 64985964.0,
      "step": 29105
    },
    {
      "entropy": 5.602549743652344,
      "epoch": 3.2693322850564384,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039597159866546996,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22121335417032242,
      "num_tokens": 64996772.0,
      "step": 29110
    },
    {
      "entropy": 5.575560188293457,
      "epoch": 3.269893861964396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003959377446595201,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22906988561153413,
      "num_tokens": 65007296.0,
      "step": 29115
    },
    {
      "entropy": 5.537145662307739,
      "epoch": 3.2704554388723537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039590388680286543,
      "loss": 4.937,
      "mean_token_accuracy": 0.23112842589616775,
      "num_tokens": 65018384.0,
      "step": 29120
    },
    {
      "entropy": 5.534000062942505,
      "epoch": 3.271017015780311,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003958700250965842,
      "loss": 4.886,
      "mean_token_accuracy": 0.22674171775579452,
      "num_tokens": 65029042.0,
      "step": 29125
    },
    {
      "entropy": 5.622165060043335,
      "epoch": 3.2715785926882686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003958361595417544,
      "loss": 4.9679,
      "mean_token_accuracy": 0.22443926185369492,
      "num_tokens": 65040792.0,
      "step": 29130
    },
    {
      "entropy": 5.62607102394104,
      "epoch": 3.2721401695962262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003958022901394545,
      "loss": 4.9142,
      "mean_token_accuracy": 0.23209898322820663,
      "num_tokens": 65051702.0,
      "step": 29135
    },
    {
      "entropy": 5.571646356582642,
      "epoch": 3.272701746504184,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003957684168907628,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22886890023946763,
      "num_tokens": 65063288.0,
      "step": 29140
    },
    {
      "entropy": 5.676317262649536,
      "epoch": 3.273263323412141,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039573453979675793,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20668453276157378,
      "num_tokens": 65074913.0,
      "step": 29145
    },
    {
      "entropy": 5.582855367660523,
      "epoch": 3.2738249003200988,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039570065885851844,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21767930090427398,
      "num_tokens": 65085327.0,
      "step": 29150
    },
    {
      "entropy": 5.601827764511109,
      "epoch": 3.2743864772280564,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003956667740771231,
      "loss": 4.938,
      "mean_token_accuracy": 0.2235243871808052,
      "num_tokens": 65095709.0,
      "step": 29155
    },
    {
      "entropy": 5.687465953826904,
      "epoch": 3.274948054136014,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003956328854536509,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22272946536540986,
      "num_tokens": 65106786.0,
      "step": 29160
    },
    {
      "entropy": 5.569165992736816,
      "epoch": 3.2755096310439713,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039559899298918094,
      "loss": 4.962,
      "mean_token_accuracy": 0.223232638835907,
      "num_tokens": 65117506.0,
      "step": 29165
    },
    {
      "entropy": 5.4807679653167725,
      "epoch": 3.276071207951929,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003955650966847922,
      "loss": 4.8402,
      "mean_token_accuracy": 0.2367965593934059,
      "num_tokens": 65127830.0,
      "step": 29170
    },
    {
      "entropy": 5.6102924823760985,
      "epoch": 3.2766327848598866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000395531196541564,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22325972765684127,
      "num_tokens": 65139562.0,
      "step": 29175
    },
    {
      "entropy": 5.608919429779053,
      "epoch": 3.2771943617678443,
      "grad_norm": 1.25,
      "learning_rate": 0.00039549729256057583,
      "loss": 4.9872,
      "mean_token_accuracy": 0.2265869379043579,
      "num_tokens": 65150799.0,
      "step": 29180
    },
    {
      "entropy": 5.550669574737549,
      "epoch": 3.2777559386758015,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000395463384742907,
      "loss": 4.972,
      "mean_token_accuracy": 0.22055285125970842,
      "num_tokens": 65161458.0,
      "step": 29185
    },
    {
      "entropy": 5.590666770935059,
      "epoch": 3.278317515583759,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039542947308963733,
      "loss": 4.9364,
      "mean_token_accuracy": 0.226124569773674,
      "num_tokens": 65172098.0,
      "step": 29190
    },
    {
      "entropy": 5.655611991882324,
      "epoch": 3.278879092491717,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003953955576018466,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22114162147045135,
      "num_tokens": 65183668.0,
      "step": 29195
    },
    {
      "entropy": 5.587381887435913,
      "epoch": 3.2794406693996745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003953616382806145,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22736500352621078,
      "num_tokens": 65193920.0,
      "step": 29200
    },
    {
      "entropy": 5.641961145401001,
      "epoch": 3.2800022463076317,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039532771512702124,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21613470911979676,
      "num_tokens": 65203397.0,
      "step": 29205
    },
    {
      "entropy": 5.584331512451172,
      "epoch": 3.2805638232155894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003952937881421468,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2183898761868477,
      "num_tokens": 65213821.0,
      "step": 29210
    },
    {
      "entropy": 5.540960693359375,
      "epoch": 3.281125400123547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003952598573270716,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22811396569013595,
      "num_tokens": 65224034.0,
      "step": 29215
    },
    {
      "entropy": 5.513869571685791,
      "epoch": 3.2816869770315042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003952259226828759,
      "loss": 4.918,
      "mean_token_accuracy": 0.22574487924575806,
      "num_tokens": 65234879.0,
      "step": 29220
    },
    {
      "entropy": 5.572229242324829,
      "epoch": 3.282248553939462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039519198421064005,
      "loss": 4.9006,
      "mean_token_accuracy": 0.23208923786878585,
      "num_tokens": 65246530.0,
      "step": 29225
    },
    {
      "entropy": 5.546381092071533,
      "epoch": 3.2828101308474196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039515804191144496,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22628300935029982,
      "num_tokens": 65258102.0,
      "step": 29230
    },
    {
      "entropy": 5.58548583984375,
      "epoch": 3.2833717077553772,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039512409578637114,
      "loss": 4.8895,
      "mean_token_accuracy": 0.23770568370819092,
      "num_tokens": 65269396.0,
      "step": 29235
    },
    {
      "entropy": 5.552657175064087,
      "epoch": 3.2839332846633345,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039509014583649947,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22832882404327393,
      "num_tokens": 65280004.0,
      "step": 29240
    },
    {
      "entropy": 5.56945948600769,
      "epoch": 3.284494861571292,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039505619206291106,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22522856146097184,
      "num_tokens": 65291564.0,
      "step": 29245
    },
    {
      "entropy": 5.554083776473999,
      "epoch": 3.2850564384792498,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039502223446668687,
      "loss": 4.8752,
      "mean_token_accuracy": 0.22876468896865845,
      "num_tokens": 65303032.0,
      "step": 29250
    },
    {
      "entropy": 5.515059566497802,
      "epoch": 3.2856180153872074,
      "grad_norm": 1.125,
      "learning_rate": 0.0003949882730489082,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23012297600507736,
      "num_tokens": 65313658.0,
      "step": 29255
    },
    {
      "entropy": 5.690699863433838,
      "epoch": 3.2861795922951647,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039495430781065645,
      "loss": 5.0257,
      "mean_token_accuracy": 0.22814230471849442,
      "num_tokens": 65324593.0,
      "step": 29260
    },
    {
      "entropy": 5.6714927673339846,
      "epoch": 3.2867411692031223,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000394920338753013,
      "loss": 4.9799,
      "mean_token_accuracy": 0.2285405144095421,
      "num_tokens": 65335059.0,
      "step": 29265
    },
    {
      "entropy": 5.671471834182739,
      "epoch": 3.28730274611108,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003948863658770594,
      "loss": 5.0879,
      "mean_token_accuracy": 0.21769069731235505,
      "num_tokens": 65347082.0,
      "step": 29270
    },
    {
      "entropy": 5.484483003616333,
      "epoch": 3.2878643230190376,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003948523891838774,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22367480993270875,
      "num_tokens": 65358644.0,
      "step": 29275
    },
    {
      "entropy": 5.6140124797821045,
      "epoch": 3.288425899926995,
      "grad_norm": 1.125,
      "learning_rate": 0.00039481840867454886,
      "loss": 4.92,
      "mean_token_accuracy": 0.2262679874897003,
      "num_tokens": 65369652.0,
      "step": 29280
    },
    {
      "entropy": 5.649748420715332,
      "epoch": 3.2889874768349525,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003947844243501558,
      "loss": 4.9151,
      "mean_token_accuracy": 0.2327731266617775,
      "num_tokens": 65380196.0,
      "step": 29285
    },
    {
      "entropy": 5.54573540687561,
      "epoch": 3.28954905374291,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039475043621178005,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22683503478765488,
      "num_tokens": 65391135.0,
      "step": 29290
    },
    {
      "entropy": 5.642907333374024,
      "epoch": 3.290110630650868,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003947164442605041,
      "loss": 5.0884,
      "mean_token_accuracy": 0.22051187306642533,
      "num_tokens": 65402330.0,
      "step": 29295
    },
    {
      "entropy": 5.631737852096558,
      "epoch": 3.290672207558825,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003946824484974101,
      "loss": 4.9755,
      "mean_token_accuracy": 0.221550852060318,
      "num_tokens": 65414111.0,
      "step": 29300
    },
    {
      "entropy": 5.599489402770996,
      "epoch": 3.2912337844667827,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003946484489235805,
      "loss": 4.9744,
      "mean_token_accuracy": 0.2305116042494774,
      "num_tokens": 65424760.0,
      "step": 29305
    },
    {
      "entropy": 5.538192844390869,
      "epoch": 3.2917953613747404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039461444554009787,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22642334401607514,
      "num_tokens": 65435213.0,
      "step": 29310
    },
    {
      "entropy": 5.509458065032959,
      "epoch": 3.2923569382826976,
      "grad_norm": 1.078125,
      "learning_rate": 0.000394580438348045,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22075680941343306,
      "num_tokens": 65446141.0,
      "step": 29315
    },
    {
      "entropy": 5.642125463485717,
      "epoch": 3.2929185151906553,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039454642734850447,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22510770261287688,
      "num_tokens": 65457634.0,
      "step": 29320
    },
    {
      "entropy": 5.60284948348999,
      "epoch": 3.293480092098613,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003945124125425594,
      "loss": 5.0076,
      "mean_token_accuracy": 0.22434690445661545,
      "num_tokens": 65468457.0,
      "step": 29325
    },
    {
      "entropy": 5.658034372329712,
      "epoch": 3.2940416690065706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003944783939312928,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2222522497177124,
      "num_tokens": 65479938.0,
      "step": 29330
    },
    {
      "entropy": 5.638342952728271,
      "epoch": 3.294603245914528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003944443715157877,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2181750312447548,
      "num_tokens": 65492030.0,
      "step": 29335
    },
    {
      "entropy": 5.649476146697998,
      "epoch": 3.2951648228224855,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003944103452971275,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22510215491056443,
      "num_tokens": 65503385.0,
      "step": 29340
    },
    {
      "entropy": 5.560481691360474,
      "epoch": 3.295726399730443,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039437631527639553,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22870975583791733,
      "num_tokens": 65514027.0,
      "step": 29345
    },
    {
      "entropy": 5.567655277252197,
      "epoch": 3.296287976638401,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039434228145467554,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22401634007692336,
      "num_tokens": 65525653.0,
      "step": 29350
    },
    {
      "entropy": 5.513604736328125,
      "epoch": 3.296849553546358,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000394308243833051,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22827862799167634,
      "num_tokens": 65536040.0,
      "step": 29355
    },
    {
      "entropy": 5.694618034362793,
      "epoch": 3.2974111304543157,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039427420241260557,
      "loss": 5.0643,
      "mean_token_accuracy": 0.22026854157447814,
      "num_tokens": 65547505.0,
      "step": 29360
    },
    {
      "entropy": 5.67518196105957,
      "epoch": 3.2979727073622733,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039424015719442335,
      "loss": 4.9453,
      "mean_token_accuracy": 0.23142997175455093,
      "num_tokens": 65557565.0,
      "step": 29365
    },
    {
      "entropy": 5.547824382781982,
      "epoch": 3.298534284270231,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039420610817958836,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22313763201236725,
      "num_tokens": 65568729.0,
      "step": 29370
    },
    {
      "entropy": 5.6104645252227785,
      "epoch": 3.299095861178188,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003941720553691846,
      "loss": 5.0451,
      "mean_token_accuracy": 0.22166544795036316,
      "num_tokens": 65580664.0,
      "step": 29375
    },
    {
      "entropy": 5.526612758636475,
      "epoch": 3.299657438086146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003941379987642963,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2298256352543831,
      "num_tokens": 65591348.0,
      "step": 29380
    },
    {
      "entropy": 5.607221269607544,
      "epoch": 3.3002190149941035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000394103938366008,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22742826193571092,
      "num_tokens": 65603553.0,
      "step": 29385
    },
    {
      "entropy": 5.565384292602539,
      "epoch": 3.300780591902061,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003940698741754041,
      "loss": 4.935,
      "mean_token_accuracy": 0.22915679961442947,
      "num_tokens": 65613903.0,
      "step": 29390
    },
    {
      "entropy": 5.599864292144775,
      "epoch": 3.3013421688100184,
      "grad_norm": 1.125,
      "learning_rate": 0.0003940358061935692,
      "loss": 4.9599,
      "mean_token_accuracy": 0.23058982491493224,
      "num_tokens": 65624862.0,
      "step": 29395
    },
    {
      "entropy": 5.545340251922608,
      "epoch": 3.301903745717976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003940017344215881,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22578013986349105,
      "num_tokens": 65636828.0,
      "step": 29400
    },
    {
      "entropy": 5.661517763137818,
      "epoch": 3.3024653226259337,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039396765886054555,
      "loss": 4.985,
      "mean_token_accuracy": 0.22415153235197066,
      "num_tokens": 65648251.0,
      "step": 29405
    },
    {
      "entropy": 5.685822343826294,
      "epoch": 3.303026899533891,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039393357951152667,
      "loss": 5.0954,
      "mean_token_accuracy": 0.21734586358070374,
      "num_tokens": 65659715.0,
      "step": 29410
    },
    {
      "entropy": 5.637775230407715,
      "epoch": 3.3035884764418486,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039389949637561646,
      "loss": 4.9952,
      "mean_token_accuracy": 0.22158850133419036,
      "num_tokens": 65670844.0,
      "step": 29415
    },
    {
      "entropy": 5.543118047714233,
      "epoch": 3.3041500533498063,
      "grad_norm": 1.0,
      "learning_rate": 0.00039386540945390017,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23249766677618028,
      "num_tokens": 65682404.0,
      "step": 29420
    },
    {
      "entropy": 5.592501306533814,
      "epoch": 3.304711630257764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003938313187474631,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22063655257225037,
      "num_tokens": 65692826.0,
      "step": 29425
    },
    {
      "entropy": 5.623502159118653,
      "epoch": 3.3052732071657216,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003937972242573908,
      "loss": 4.9793,
      "mean_token_accuracy": 0.22415825128555297,
      "num_tokens": 65704139.0,
      "step": 29430
    },
    {
      "entropy": 5.552084016799927,
      "epoch": 3.305834784073679,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003937631259847688,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22528549283742905,
      "num_tokens": 65716282.0,
      "step": 29435
    },
    {
      "entropy": 5.6569664478302,
      "epoch": 3.3063963609816365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003937290239306828,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21925278156995773,
      "num_tokens": 65728086.0,
      "step": 29440
    },
    {
      "entropy": 5.66805362701416,
      "epoch": 3.306957937889594,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039369491809621867,
      "loss": 5.0351,
      "mean_token_accuracy": 0.22420178055763246,
      "num_tokens": 65740584.0,
      "step": 29445
    },
    {
      "entropy": 5.591322803497315,
      "epoch": 3.3075195147975514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003936608084824623,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2252998322248459,
      "num_tokens": 65751479.0,
      "step": 29450
    },
    {
      "entropy": 5.55693039894104,
      "epoch": 3.308081091705509,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039362669509049975,
      "loss": 4.9919,
      "mean_token_accuracy": 0.22148360162973404,
      "num_tokens": 65762073.0,
      "step": 29455
    },
    {
      "entropy": 5.551419067382812,
      "epoch": 3.3086426686134667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039359257792141717,
      "loss": 4.9199,
      "mean_token_accuracy": 0.220340596139431,
      "num_tokens": 65772808.0,
      "step": 29460
    },
    {
      "entropy": 5.582326793670655,
      "epoch": 3.3092042455214243,
      "grad_norm": 1.0625,
      "learning_rate": 0.000393558456976301,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22326065301895143,
      "num_tokens": 65783507.0,
      "step": 29465
    },
    {
      "entropy": 5.620807218551636,
      "epoch": 3.3097658224293816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003935243322562375,
      "loss": 4.964,
      "mean_token_accuracy": 0.2256319999694824,
      "num_tokens": 65794250.0,
      "step": 29470
    },
    {
      "entropy": 5.571021318435669,
      "epoch": 3.310327399337339,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003934902037623133,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21984667032957078,
      "num_tokens": 65805487.0,
      "step": 29475
    },
    {
      "entropy": 5.604675531387329,
      "epoch": 3.310888976245297,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039345607149561513,
      "loss": 4.993,
      "mean_token_accuracy": 0.22123419046401976,
      "num_tokens": 65818264.0,
      "step": 29480
    },
    {
      "entropy": 5.7161980152130125,
      "epoch": 3.3114505531532545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039342193545722966,
      "loss": 5.0866,
      "mean_token_accuracy": 0.230594502389431,
      "num_tokens": 65829686.0,
      "step": 29485
    },
    {
      "entropy": 5.592301607131958,
      "epoch": 3.3120121300612118,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003933877956482438,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2341941088438034,
      "num_tokens": 65840800.0,
      "step": 29490
    },
    {
      "entropy": 5.625569772720337,
      "epoch": 3.3125737069691694,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039335365206974465,
      "loss": 4.9774,
      "mean_token_accuracy": 0.22781905382871628,
      "num_tokens": 65853634.0,
      "step": 29495
    },
    {
      "entropy": 5.574265336990356,
      "epoch": 3.313135283877127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003933195047228193,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21915556937456132,
      "num_tokens": 65864754.0,
      "step": 29500
    },
    {
      "entropy": 5.550901174545288,
      "epoch": 3.3136968607850843,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039328535360855503,
      "loss": 4.9903,
      "mean_token_accuracy": 0.21971358954906464,
      "num_tokens": 65876609.0,
      "step": 29505
    },
    {
      "entropy": 5.582277154922485,
      "epoch": 3.314258437693042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039325119872803924,
      "loss": 4.944,
      "mean_token_accuracy": 0.2226081445813179,
      "num_tokens": 65887583.0,
      "step": 29510
    },
    {
      "entropy": 5.617700386047363,
      "epoch": 3.3148200146009996,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039321704008235943,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23450740575790405,
      "num_tokens": 65898131.0,
      "step": 29515
    },
    {
      "entropy": 5.571900081634522,
      "epoch": 3.3153815915089573,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039318287767260316,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22545523643493653,
      "num_tokens": 65909353.0,
      "step": 29520
    },
    {
      "entropy": 5.599009895324707,
      "epoch": 3.315943168416915,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039314871149985834,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22569386959075927,
      "num_tokens": 65920709.0,
      "step": 29525
    },
    {
      "entropy": 5.651408529281616,
      "epoch": 3.316504745324872,
      "grad_norm": 1.0,
      "learning_rate": 0.0003931145415652126,
      "loss": 5.0143,
      "mean_token_accuracy": 0.22709547728300095,
      "num_tokens": 65932231.0,
      "step": 29530
    },
    {
      "entropy": 5.725611019134521,
      "epoch": 3.31706632223283,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039308036786975416,
      "loss": 5.0757,
      "mean_token_accuracy": 0.21580370664596557,
      "num_tokens": 65942629.0,
      "step": 29535
    },
    {
      "entropy": 5.600027751922608,
      "epoch": 3.3176278991407875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003930461904145709,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22972011864185332,
      "num_tokens": 65954385.0,
      "step": 29540
    },
    {
      "entropy": 5.595013570785523,
      "epoch": 3.3181894760487447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003930120092007512,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22538366168737411,
      "num_tokens": 65966283.0,
      "step": 29545
    },
    {
      "entropy": 5.659839487075805,
      "epoch": 3.3187510529567024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003929778242293834,
      "loss": 5.1034,
      "mean_token_accuracy": 0.21716419905424117,
      "num_tokens": 65978594.0,
      "step": 29550
    },
    {
      "entropy": 5.5499207973480225,
      "epoch": 3.31931262986466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003929436355015558,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22540924847126007,
      "num_tokens": 65989398.0,
      "step": 29555
    },
    {
      "entropy": 5.601590299606324,
      "epoch": 3.3198742067726177,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039290944301835723,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22409229278564452,
      "num_tokens": 66000545.0,
      "step": 29560
    },
    {
      "entropy": 5.57735767364502,
      "epoch": 3.320435783680575,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039287524678087617,
      "loss": 4.959,
      "mean_token_accuracy": 0.22355961948633193,
      "num_tokens": 66011638.0,
      "step": 29565
    },
    {
      "entropy": 5.555989408493042,
      "epoch": 3.3209973605885326,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039284104679020154,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22148693352937698,
      "num_tokens": 66023284.0,
      "step": 29570
    },
    {
      "entropy": 5.5745457172393795,
      "epoch": 3.3215589374964902,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003928068430474224,
      "loss": 4.9701,
      "mean_token_accuracy": 0.21495273858308792,
      "num_tokens": 66034099.0,
      "step": 29575
    },
    {
      "entropy": 5.588131093978882,
      "epoch": 3.322120514404448,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003927726355536275,
      "loss": 4.887,
      "mean_token_accuracy": 0.22731123268604278,
      "num_tokens": 66043885.0,
      "step": 29580
    },
    {
      "entropy": 5.644365358352661,
      "epoch": 3.322682091312405,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039273842430990634,
      "loss": 5.0266,
      "mean_token_accuracy": 0.21610798686742783,
      "num_tokens": 66054190.0,
      "step": 29585
    },
    {
      "entropy": 5.607356023788452,
      "epoch": 3.3232436682203628,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003927042093173481,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22645300030708312,
      "num_tokens": 66065254.0,
      "step": 29590
    },
    {
      "entropy": 5.652400732040405,
      "epoch": 3.3238052451283204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003926699905770421,
      "loss": 5.0553,
      "mean_token_accuracy": 0.22079820632934571,
      "num_tokens": 66076582.0,
      "step": 29595
    },
    {
      "entropy": 5.692382383346557,
      "epoch": 3.3243668220362776,
      "grad_norm": 1.140625,
      "learning_rate": 0.000392635768090078,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21992577761411666,
      "num_tokens": 66086410.0,
      "step": 29600
    },
    {
      "entropy": 5.659572315216065,
      "epoch": 3.3249283989442353,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039260154185754545,
      "loss": 5.0344,
      "mean_token_accuracy": 0.2180202767252922,
      "num_tokens": 66098393.0,
      "step": 29605
    },
    {
      "entropy": 5.637894773483277,
      "epoch": 3.325489975852193,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003925673118805341,
      "loss": 4.9885,
      "mean_token_accuracy": 0.223830546438694,
      "num_tokens": 66108077.0,
      "step": 29610
    },
    {
      "entropy": 5.492300939559937,
      "epoch": 3.3260515527601506,
      "grad_norm": 1.109375,
      "learning_rate": 0.000392533078160134,
      "loss": 4.7792,
      "mean_token_accuracy": 0.240425243973732,
      "num_tokens": 66118008.0,
      "step": 29615
    },
    {
      "entropy": 5.607463312149048,
      "epoch": 3.3266131296681083,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039249884069743514,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21901334673166276,
      "num_tokens": 66128697.0,
      "step": 29620
    },
    {
      "entropy": 5.673470211029053,
      "epoch": 3.3271747065760655,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003924645994935276,
      "loss": 4.968,
      "mean_token_accuracy": 0.22654685378074646,
      "num_tokens": 66139800.0,
      "step": 29625
    },
    {
      "entropy": 5.616893291473389,
      "epoch": 3.327736283484023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003924303545495017,
      "loss": 4.8736,
      "mean_token_accuracy": 0.232675601541996,
      "num_tokens": 66151202.0,
      "step": 29630
    },
    {
      "entropy": 5.613529872894287,
      "epoch": 3.328297860391981,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039239610586644765,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22166222035884858,
      "num_tokens": 66162159.0,
      "step": 29635
    },
    {
      "entropy": 5.631685447692871,
      "epoch": 3.328859437299938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003923618534454561,
      "loss": 4.9288,
      "mean_token_accuracy": 0.23267656117677687,
      "num_tokens": 66174492.0,
      "step": 29640
    },
    {
      "entropy": 5.587041568756104,
      "epoch": 3.3294210142078957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003923275972876176,
      "loss": 5.0087,
      "mean_token_accuracy": 0.22501687407493592,
      "num_tokens": 66186617.0,
      "step": 29645
    },
    {
      "entropy": 5.542071342468262,
      "epoch": 3.3299825911158534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039229333739402303,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23222880512475969,
      "num_tokens": 66198001.0,
      "step": 29650
    },
    {
      "entropy": 5.515701246261597,
      "epoch": 3.330544168023811,
      "grad_norm": 1.078125,
      "learning_rate": 0.000392259073765763,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22470787912607193,
      "num_tokens": 66208017.0,
      "step": 29655
    },
    {
      "entropy": 5.542723369598389,
      "epoch": 3.3311057449317683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039222480640392854,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2287622109055519,
      "num_tokens": 66219798.0,
      "step": 29660
    },
    {
      "entropy": 5.625527477264404,
      "epoch": 3.331667321839726,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039219053530961087,
      "loss": 4.9391,
      "mean_token_accuracy": 0.23278995007276534,
      "num_tokens": 66230662.0,
      "step": 29665
    },
    {
      "entropy": 5.611478042602539,
      "epoch": 3.3322288987476836,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039215626048390105,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22718610614538193,
      "num_tokens": 66240787.0,
      "step": 29670
    },
    {
      "entropy": 5.571321821212768,
      "epoch": 3.3327904756556412,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003921219819278905,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22581580877304078,
      "num_tokens": 66250937.0,
      "step": 29675
    },
    {
      "entropy": 5.670702409744263,
      "epoch": 3.3333520525635985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003920876996426706,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21978678852319716,
      "num_tokens": 66261322.0,
      "step": 29680
    },
    {
      "entropy": 5.586051940917969,
      "epoch": 3.333913629471556,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039205341362933294,
      "loss": 4.959,
      "mean_token_accuracy": 0.2251680463552475,
      "num_tokens": 66271307.0,
      "step": 29685
    },
    {
      "entropy": 5.515356731414795,
      "epoch": 3.334475206379514,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003920191238889691,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22960571348667144,
      "num_tokens": 66283067.0,
      "step": 29690
    },
    {
      "entropy": 5.615796518325806,
      "epoch": 3.335036783287471,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003919848304226711,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22390668839216232,
      "num_tokens": 66295853.0,
      "step": 29695
    },
    {
      "entropy": 5.635197162628174,
      "epoch": 3.3355983601954287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003919505332315307,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22944965660572053,
      "num_tokens": 66307573.0,
      "step": 29700
    },
    {
      "entropy": 5.589113855361939,
      "epoch": 3.3361599371033863,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039191623231663985,
      "loss": 4.92,
      "mean_token_accuracy": 0.23356925696134567,
      "num_tokens": 66318100.0,
      "step": 29705
    },
    {
      "entropy": 5.658594083786011,
      "epoch": 3.336721514011344,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000391881927679091,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22305692732334137,
      "num_tokens": 66328918.0,
      "step": 29710
    },
    {
      "entropy": 5.6098065853118895,
      "epoch": 3.3372830909193016,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039184761931997607,
      "loss": 5.0097,
      "mean_token_accuracy": 0.22135841101408005,
      "num_tokens": 66341124.0,
      "step": 29715
    },
    {
      "entropy": 5.594089984893799,
      "epoch": 3.337844667827259,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003918133072403878,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22391465306282043,
      "num_tokens": 66351033.0,
      "step": 29720
    },
    {
      "entropy": 5.5703874111175535,
      "epoch": 3.3384062447352165,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039177899144141835,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2274639829993248,
      "num_tokens": 66363338.0,
      "step": 29725
    },
    {
      "entropy": 5.571627235412597,
      "epoch": 3.338967821643174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003917446719241606,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22717212438583373,
      "num_tokens": 66373716.0,
      "step": 29730
    },
    {
      "entropy": 5.6242577075958256,
      "epoch": 3.3395293985511314,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039171034868970724,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22904265373945237,
      "num_tokens": 66384448.0,
      "step": 29735
    },
    {
      "entropy": 5.530878925323487,
      "epoch": 3.340090975459089,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039167602173915104,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22665576934814452,
      "num_tokens": 66395662.0,
      "step": 29740
    },
    {
      "entropy": 5.499372768402099,
      "epoch": 3.3406525523670467,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003916416910735851,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2302619054913521,
      "num_tokens": 66407856.0,
      "step": 29745
    },
    {
      "entropy": 5.70628719329834,
      "epoch": 3.3412141292750044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039160735669410246,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21395705193281173,
      "num_tokens": 66419324.0,
      "step": 29750
    },
    {
      "entropy": 5.591220140457153,
      "epoch": 3.3417757061829616,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003915730186017964,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22597711086273192,
      "num_tokens": 66430720.0,
      "step": 29755
    },
    {
      "entropy": 5.5861608505249025,
      "epoch": 3.3423372830909193,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039153867679776023,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21388618052005767,
      "num_tokens": 66441736.0,
      "step": 29760
    },
    {
      "entropy": 5.623452138900757,
      "epoch": 3.342898859998877,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003915043312830873,
      "loss": 5.003,
      "mean_token_accuracy": 0.22821534425020218,
      "num_tokens": 66451851.0,
      "step": 29765
    },
    {
      "entropy": 5.715250492095947,
      "epoch": 3.3434604369068346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003914699820588714,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2119002714753151,
      "num_tokens": 66463105.0,
      "step": 29770
    },
    {
      "entropy": 5.683483028411866,
      "epoch": 3.344022013814792,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039143562912620603,
      "loss": 5.0687,
      "mean_token_accuracy": 0.221137535572052,
      "num_tokens": 66474613.0,
      "step": 29775
    },
    {
      "entropy": 5.568648386001587,
      "epoch": 3.3445835907227495,
      "grad_norm": 1.109375,
      "learning_rate": 0.000391401272486185,
      "loss": 5.0077,
      "mean_token_accuracy": 0.22157644778490065,
      "num_tokens": 66486783.0,
      "step": 29780
    },
    {
      "entropy": 5.694684028625488,
      "epoch": 3.345145167630707,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003913669121399024,
      "loss": 5.0468,
      "mean_token_accuracy": 0.22400101125240326,
      "num_tokens": 66498581.0,
      "step": 29785
    },
    {
      "entropy": 5.570183134078979,
      "epoch": 3.3457067445386643,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039133254808845216,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2311703786253929,
      "num_tokens": 66509480.0,
      "step": 29790
    },
    {
      "entropy": 5.5819202899932865,
      "epoch": 3.346268321446622,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039129818033292843,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22586169540882112,
      "num_tokens": 66521831.0,
      "step": 29795
    },
    {
      "entropy": 5.644610500335693,
      "epoch": 3.3468298983545797,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003912638088744256,
      "loss": 4.9775,
      "mean_token_accuracy": 0.22426416873931884,
      "num_tokens": 66533900.0,
      "step": 29800
    },
    {
      "entropy": 5.749560928344726,
      "epoch": 3.3473914752625373,
      "grad_norm": 1.203125,
      "learning_rate": 0.000391229433714038,
      "loss": 5.0782,
      "mean_token_accuracy": 0.21576396375894547,
      "num_tokens": 66543848.0,
      "step": 29805
    },
    {
      "entropy": 5.642170572280884,
      "epoch": 3.347953052170495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003911950548528601,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2233976364135742,
      "num_tokens": 66554113.0,
      "step": 29810
    },
    {
      "entropy": 5.649364852905274,
      "epoch": 3.348514629078452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003911606722919866,
      "loss": 5.0039,
      "mean_token_accuracy": 0.22223274707794188,
      "num_tokens": 66565202.0,
      "step": 29815
    },
    {
      "entropy": 5.644724798202515,
      "epoch": 3.34907620598641,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003911262860325123,
      "loss": 5.039,
      "mean_token_accuracy": 0.21523437798023223,
      "num_tokens": 66575930.0,
      "step": 29820
    },
    {
      "entropy": 5.644399929046631,
      "epoch": 3.3496377828943675,
      "grad_norm": 1.125,
      "learning_rate": 0.0003910918960755319,
      "loss": 4.928,
      "mean_token_accuracy": 0.2250242277979851,
      "num_tokens": 66585422.0,
      "step": 29825
    },
    {
      "entropy": 5.633545589447022,
      "epoch": 3.3501993598023248,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003910575024221406,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22248542606830596,
      "num_tokens": 66597395.0,
      "step": 29830
    },
    {
      "entropy": 5.644780778884888,
      "epoch": 3.3507609367102824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039102310507343335,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22732343524694443,
      "num_tokens": 66608464.0,
      "step": 29835
    },
    {
      "entropy": 5.652032375335693,
      "epoch": 3.35132251361824,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003909887040305055,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21857015639543534,
      "num_tokens": 66619930.0,
      "step": 29840
    },
    {
      "entropy": 5.5503686428070065,
      "epoch": 3.3518840905261977,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003909542992944523,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22794270068407058,
      "num_tokens": 66632833.0,
      "step": 29845
    },
    {
      "entropy": 5.601869249343872,
      "epoch": 3.352445667434155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003909198908663692,
      "loss": 5.006,
      "mean_token_accuracy": 0.2178230404853821,
      "num_tokens": 66643641.0,
      "step": 29850
    },
    {
      "entropy": 5.716343545913697,
      "epoch": 3.3530072443421126,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039088547874735187,
      "loss": 5.0459,
      "mean_token_accuracy": 0.22042033821344376,
      "num_tokens": 66653863.0,
      "step": 29855
    },
    {
      "entropy": 5.551467800140381,
      "epoch": 3.3535688212500703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039085106293849594,
      "loss": 4.8953,
      "mean_token_accuracy": 0.23037908375263214,
      "num_tokens": 66665885.0,
      "step": 29860
    },
    {
      "entropy": 5.625711488723755,
      "epoch": 3.354130398158028,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039081664344089724,
      "loss": 4.97,
      "mean_token_accuracy": 0.22225388288497924,
      "num_tokens": 66676584.0,
      "step": 29865
    },
    {
      "entropy": 5.514528846740722,
      "epoch": 3.354691975065985,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003907822202556516,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23074050098657609,
      "num_tokens": 66687882.0,
      "step": 29870
    },
    {
      "entropy": 5.580749607086181,
      "epoch": 3.355253551973943,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003907477933838553,
      "loss": 4.9362,
      "mean_token_accuracy": 0.22272120118141175,
      "num_tokens": 66699665.0,
      "step": 29875
    },
    {
      "entropy": 5.636850595474243,
      "epoch": 3.3558151288819005,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039071336282660437,
      "loss": 5.0592,
      "mean_token_accuracy": 0.2149189069867134,
      "num_tokens": 66711455.0,
      "step": 29880
    },
    {
      "entropy": 5.556344938278198,
      "epoch": 3.3563767057898577,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039067892858499496,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2338645800948143,
      "num_tokens": 66722369.0,
      "step": 29885
    },
    {
      "entropy": 5.578876399993897,
      "epoch": 3.3569382826978154,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003906444906601237,
      "loss": 5.0134,
      "mean_token_accuracy": 0.22531802654266359,
      "num_tokens": 66733200.0,
      "step": 29890
    },
    {
      "entropy": 5.582014894485473,
      "epoch": 3.357499859605773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000390610049053087,
      "loss": 4.907,
      "mean_token_accuracy": 0.231497922539711,
      "num_tokens": 66744650.0,
      "step": 29895
    },
    {
      "entropy": 5.649638271331787,
      "epoch": 3.3580614365137307,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039057560376498147,
      "loss": 4.9914,
      "mean_token_accuracy": 0.2204744353890419,
      "num_tokens": 66755712.0,
      "step": 29900
    },
    {
      "entropy": 5.672937679290771,
      "epoch": 3.3586230134216883,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003905411547969038,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22682781368494034,
      "num_tokens": 66765855.0,
      "step": 29905
    },
    {
      "entropy": 5.625009202957154,
      "epoch": 3.3591845903296456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003905067021499511,
      "loss": 4.9778,
      "mean_token_accuracy": 0.23037370443344116,
      "num_tokens": 66776870.0,
      "step": 29910
    },
    {
      "entropy": 5.573329210281372,
      "epoch": 3.359746167237603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039047224582522015,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22921680510044098,
      "num_tokens": 66787226.0,
      "step": 29915
    },
    {
      "entropy": 5.566879034042358,
      "epoch": 3.360307744145561,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003904377858238081,
      "loss": 4.9051,
      "mean_token_accuracy": 0.22700815051794052,
      "num_tokens": 66797352.0,
      "step": 29920
    },
    {
      "entropy": 5.6584389209747314,
      "epoch": 3.360869321053518,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003904033221468122,
      "loss": 5.057,
      "mean_token_accuracy": 0.22120561748743056,
      "num_tokens": 66809227.0,
      "step": 29925
    },
    {
      "entropy": 5.647086238861084,
      "epoch": 3.3614308979614758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003903688547953297,
      "loss": 4.9732,
      "mean_token_accuracy": 0.2225823223590851,
      "num_tokens": 66821196.0,
      "step": 29930
    },
    {
      "entropy": 5.649438428878784,
      "epoch": 3.3619924748694334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003903343837704582,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21889761090278625,
      "num_tokens": 66831958.0,
      "step": 29935
    },
    {
      "entropy": 5.630073308944702,
      "epoch": 3.362554051777391,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003902999090732951,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22679290920495987,
      "num_tokens": 66843727.0,
      "step": 29940
    },
    {
      "entropy": 5.569244909286499,
      "epoch": 3.3631156286853483,
      "grad_norm": 1.125,
      "learning_rate": 0.00039026543070493823,
      "loss": 4.9361,
      "mean_token_accuracy": 0.2319529876112938,
      "num_tokens": 66854756.0,
      "step": 29945
    },
    {
      "entropy": 5.539178037643433,
      "epoch": 3.363677205593306,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003902309486664854,
      "loss": 4.9177,
      "mean_token_accuracy": 0.2317902386188507,
      "num_tokens": 66865615.0,
      "step": 29950
    },
    {
      "entropy": 5.585346841812134,
      "epoch": 3.3642387825012636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003901964629590344,
      "loss": 4.9725,
      "mean_token_accuracy": 0.21666838079690934,
      "num_tokens": 66877262.0,
      "step": 29955
    },
    {
      "entropy": 5.71942629814148,
      "epoch": 3.3648003594092213,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003901619735836833,
      "loss": 5.0557,
      "mean_token_accuracy": 0.218070949614048,
      "num_tokens": 66889405.0,
      "step": 29960
    },
    {
      "entropy": 5.620297145843506,
      "epoch": 3.3653619363171785,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039012748054153044,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22968684732913972,
      "num_tokens": 66900013.0,
      "step": 29965
    },
    {
      "entropy": 5.554594659805298,
      "epoch": 3.365923513225136,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003900929838336738,
      "loss": 4.8268,
      "mean_token_accuracy": 0.23168692290782927,
      "num_tokens": 66909731.0,
      "step": 29970
    },
    {
      "entropy": 5.51492018699646,
      "epoch": 3.366485090133094,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003900584834612119,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23270375579595565,
      "num_tokens": 66919731.0,
      "step": 29975
    },
    {
      "entropy": 5.563082647323609,
      "epoch": 3.367046667041051,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003900239794252434,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2206219792366028,
      "num_tokens": 66930169.0,
      "step": 29980
    },
    {
      "entropy": 5.5901227474212645,
      "epoch": 3.3676082439490087,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003899894717268667,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22905222475528716,
      "num_tokens": 66942010.0,
      "step": 29985
    },
    {
      "entropy": 5.63862419128418,
      "epoch": 3.3681698208569664,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003899549603671807,
      "loss": 4.9345,
      "mean_token_accuracy": 0.23436496108770372,
      "num_tokens": 66953972.0,
      "step": 29990
    },
    {
      "entropy": 5.598096656799316,
      "epoch": 3.368731397764924,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038992044534728404,
      "loss": 5.0428,
      "mean_token_accuracy": 0.22241586744785308,
      "num_tokens": 66965548.0,
      "step": 29995
    },
    {
      "entropy": 5.632755279541016,
      "epoch": 3.3692929746728817,
      "grad_norm": 1.125,
      "learning_rate": 0.00038988592666827594,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2229652002453804,
      "num_tokens": 66976414.0,
      "step": 30000
    },
    {
      "epoch": 3.3692929746728817,
      "eval_entropy": 5.383349709883467,
      "eval_loss": 5.082045555114746,
      "eval_mean_token_accuracy": 0.22560365160416473,
      "eval_num_tokens": 66976414.0,
      "eval_runtime": 18.3056,
      "eval_samples_per_second": 1694.183,
      "eval_steps_per_second": 211.793,
      "step": 30000
    },
    {
      "entropy": 5.523261880874633,
      "epoch": 3.369854551580839,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003898514043312553,
      "loss": 4.9223,
      "mean_token_accuracy": 0.23195020258426666,
      "num_tokens": 66987392.0,
      "step": 30005
    },
    {
      "entropy": 5.736445426940918,
      "epoch": 3.3704161284887966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038981687833732146,
      "loss": 5.0985,
      "mean_token_accuracy": 0.21668064296245576,
      "num_tokens": 67000444.0,
      "step": 30010
    },
    {
      "entropy": 5.636292839050293,
      "epoch": 3.3709777053967542,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003897823486875737,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2282344251871109,
      "num_tokens": 67012124.0,
      "step": 30015
    },
    {
      "entropy": 5.725708055496216,
      "epoch": 3.3715392823047114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003897478153831114,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2200545623898506,
      "num_tokens": 67024179.0,
      "step": 30020
    },
    {
      "entropy": 5.5962611675262455,
      "epoch": 3.372100859212669,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038971327842503414,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22684457302093505,
      "num_tokens": 67035876.0,
      "step": 30025
    },
    {
      "entropy": 5.525129699707032,
      "epoch": 3.3726624361206268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038967873781444167,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22383906394243241,
      "num_tokens": 67046499.0,
      "step": 30030
    },
    {
      "entropy": 5.616626930236817,
      "epoch": 3.3732240130285844,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038964419355243365,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22503843754529954,
      "num_tokens": 67058357.0,
      "step": 30035
    },
    {
      "entropy": 5.670000267028809,
      "epoch": 3.3737855899365417,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003896096456401101,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21608024090528488,
      "num_tokens": 67069753.0,
      "step": 30040
    },
    {
      "entropy": 5.592841005325317,
      "epoch": 3.3743471668444993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038957509407857097,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2254795640707016,
      "num_tokens": 67080185.0,
      "step": 30045
    },
    {
      "entropy": 5.5631890296936035,
      "epoch": 3.374908743752457,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038954053886891635,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22584028542041779,
      "num_tokens": 67091382.0,
      "step": 30050
    },
    {
      "entropy": 5.676650905609131,
      "epoch": 3.3754703206604146,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003895059800122466,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20784001648426056,
      "num_tokens": 67102952.0,
      "step": 30055
    },
    {
      "entropy": 5.593445301055908,
      "epoch": 3.376031897568372,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000389471417509662,
      "loss": 4.892,
      "mean_token_accuracy": 0.2209337294101715,
      "num_tokens": 67113498.0,
      "step": 30060
    },
    {
      "entropy": 5.589728784561157,
      "epoch": 3.3765934744763295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000389436851362263,
      "loss": 4.974,
      "mean_token_accuracy": 0.22202499806880951,
      "num_tokens": 67124591.0,
      "step": 30065
    },
    {
      "entropy": 5.608725643157959,
      "epoch": 3.377155051384287,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038940228157115036,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22400722056627273,
      "num_tokens": 67135627.0,
      "step": 30070
    },
    {
      "entropy": 5.592680406570435,
      "epoch": 3.3777166282922444,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038936770813742473,
      "loss": 4.8799,
      "mean_token_accuracy": 0.23156366646289825,
      "num_tokens": 67146084.0,
      "step": 30075
    },
    {
      "entropy": 5.598892498016357,
      "epoch": 3.378278205200202,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038933313106218694,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2316710337996483,
      "num_tokens": 67157749.0,
      "step": 30080
    },
    {
      "entropy": 5.651742219924927,
      "epoch": 3.3788397821081597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038929855034653776,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21888129264116288,
      "num_tokens": 67169386.0,
      "step": 30085
    },
    {
      "entropy": 5.593345069885254,
      "epoch": 3.3794013590161174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038926396599157855,
      "loss": 4.958,
      "mean_token_accuracy": 0.22820251137018205,
      "num_tokens": 67181436.0,
      "step": 30090
    },
    {
      "entropy": 5.615028285980225,
      "epoch": 3.379962935924075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003892293779984103,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2306488886475563,
      "num_tokens": 67192576.0,
      "step": 30095
    },
    {
      "entropy": 5.692945432662964,
      "epoch": 3.3805245128320323,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003891947863681344,
      "loss": 5.0538,
      "mean_token_accuracy": 0.22114433348178864,
      "num_tokens": 67205083.0,
      "step": 30100
    },
    {
      "entropy": 5.480548095703125,
      "epoch": 3.38108608973999,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003891601911018521,
      "loss": 4.8405,
      "mean_token_accuracy": 0.2290043607354164,
      "num_tokens": 67216499.0,
      "step": 30105
    },
    {
      "entropy": 5.528775835037232,
      "epoch": 3.3816476666479476,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003891255922006652,
      "loss": 4.8706,
      "mean_token_accuracy": 0.232944492995739,
      "num_tokens": 67228390.0,
      "step": 30110
    },
    {
      "entropy": 5.6480522632598875,
      "epoch": 3.382209243555905,
      "grad_norm": 1.09375,
      "learning_rate": 0.000389090989665675,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22766149193048477,
      "num_tokens": 67239655.0,
      "step": 30115
    },
    {
      "entropy": 5.639115715026856,
      "epoch": 3.3827708204638625,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003890563834979835,
      "loss": 5.0136,
      "mean_token_accuracy": 0.22221429497003556,
      "num_tokens": 67251712.0,
      "step": 30120
    },
    {
      "entropy": 5.613657760620117,
      "epoch": 3.38333239737182,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003890217736986926,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22915991097688676,
      "num_tokens": 67262494.0,
      "step": 30125
    },
    {
      "entropy": 5.583072805404663,
      "epoch": 3.383893974279778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038898716026890414,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23059938251972198,
      "num_tokens": 67273179.0,
      "step": 30130
    },
    {
      "entropy": 5.605038833618164,
      "epoch": 3.384455551187735,
      "grad_norm": 1.125,
      "learning_rate": 0.0003889525432097203,
      "loss": 4.9748,
      "mean_token_accuracy": 0.23062028586864472,
      "num_tokens": 67284031.0,
      "step": 30135
    },
    {
      "entropy": 5.675262975692749,
      "epoch": 3.3850171280956927,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003889179225222432,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21874882727861406,
      "num_tokens": 67295728.0,
      "step": 30140
    },
    {
      "entropy": 5.61605052947998,
      "epoch": 3.3855787050036503,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038888329820757536,
      "loss": 5.027,
      "mean_token_accuracy": 0.21748437881469726,
      "num_tokens": 67306880.0,
      "step": 30145
    },
    {
      "entropy": 5.5722438335418705,
      "epoch": 3.386140281911608,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038884867026681906,
      "loss": 4.9134,
      "mean_token_accuracy": 0.22826255857944489,
      "num_tokens": 67317849.0,
      "step": 30150
    },
    {
      "entropy": 5.543493461608887,
      "epoch": 3.386701858819565,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003888140387010771,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22740577161312103,
      "num_tokens": 67329421.0,
      "step": 30155
    },
    {
      "entropy": 5.6267987251281735,
      "epoch": 3.387263435727523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038877940351145186,
      "loss": 5.0121,
      "mean_token_accuracy": 0.22620517313480376,
      "num_tokens": 67340072.0,
      "step": 30160
    },
    {
      "entropy": 5.6075108528137205,
      "epoch": 3.3878250126354805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038874476469904633,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22934878170490264,
      "num_tokens": 67351687.0,
      "step": 30165
    },
    {
      "entropy": 5.591091632843018,
      "epoch": 3.3883865895434377,
      "grad_norm": 1.125,
      "learning_rate": 0.0003887101222649634,
      "loss": 4.9202,
      "mean_token_accuracy": 0.2251627713441849,
      "num_tokens": 67362978.0,
      "step": 30170
    },
    {
      "entropy": 5.572110891342163,
      "epoch": 3.3889481664513954,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000388675476210306,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23752855211496354,
      "num_tokens": 67374142.0,
      "step": 30175
    },
    {
      "entropy": 5.618254137039185,
      "epoch": 3.389509743359353,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038864082653617744,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21489768326282502,
      "num_tokens": 67385196.0,
      "step": 30180
    },
    {
      "entropy": 5.590887212753296,
      "epoch": 3.3900713202673107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038860617324368085,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22580941766500473,
      "num_tokens": 67395830.0,
      "step": 30185
    },
    {
      "entropy": 5.589262676239014,
      "epoch": 3.3906328971752684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003885715163339196,
      "loss": 4.964,
      "mean_token_accuracy": 0.22482893466949463,
      "num_tokens": 67405950.0,
      "step": 30190
    },
    {
      "entropy": 5.626977968215942,
      "epoch": 3.3911944740832256,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038853685580799726,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22592855095863343,
      "num_tokens": 67417993.0,
      "step": 30195
    },
    {
      "entropy": 5.637291145324707,
      "epoch": 3.3917560509911833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003885021916670174,
      "loss": 4.937,
      "mean_token_accuracy": 0.23074862062931062,
      "num_tokens": 67429222.0,
      "step": 30200
    },
    {
      "entropy": 5.664545297622681,
      "epoch": 3.392317627899141,
      "grad_norm": 1.125,
      "learning_rate": 0.0003884675239120838,
      "loss": 5.0454,
      "mean_token_accuracy": 0.22051576972007753,
      "num_tokens": 67439870.0,
      "step": 30205
    },
    {
      "entropy": 5.594368124008179,
      "epoch": 3.392879204807098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038843285254430005,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23083623200654985,
      "num_tokens": 67450491.0,
      "step": 30210
    },
    {
      "entropy": 5.645157098770142,
      "epoch": 3.393440781715056,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003883981775647705,
      "loss": 4.9844,
      "mean_token_accuracy": 0.2194953590631485,
      "num_tokens": 67461432.0,
      "step": 30215
    },
    {
      "entropy": 5.635856056213379,
      "epoch": 3.3940023586230135,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003883634989745988,
      "loss": 4.971,
      "mean_token_accuracy": 0.22474358826875687,
      "num_tokens": 67472614.0,
      "step": 30220
    },
    {
      "entropy": 5.629929924011231,
      "epoch": 3.394563935530971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038832881677488944,
      "loss": 4.9884,
      "mean_token_accuracy": 0.2238927513360977,
      "num_tokens": 67484383.0,
      "step": 30225
    },
    {
      "entropy": 5.631404066085816,
      "epoch": 3.3951255124389284,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038829413096674653,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22054000049829484,
      "num_tokens": 67495697.0,
      "step": 30230
    },
    {
      "entropy": 5.570409154891967,
      "epoch": 3.395687089346886,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003882594415512746,
      "loss": 4.9463,
      "mean_token_accuracy": 0.22415576577186586,
      "num_tokens": 67507937.0,
      "step": 30235
    },
    {
      "entropy": 5.574394702911377,
      "epoch": 3.3962486662548437,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038822474852957817,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22794419825077056,
      "num_tokens": 67519649.0,
      "step": 30240
    },
    {
      "entropy": 5.726794958114624,
      "epoch": 3.3968102431628013,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003881900519027618,
      "loss": 5.06,
      "mean_token_accuracy": 0.22045824378728868,
      "num_tokens": 67532301.0,
      "step": 30245
    },
    {
      "entropy": 5.614274597167968,
      "epoch": 3.3973718200707586,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003881553516719303,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22654667794704436,
      "num_tokens": 67545106.0,
      "step": 30250
    },
    {
      "entropy": 5.631371355056762,
      "epoch": 3.397933396978716,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038812064783818847,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2213209480047226,
      "num_tokens": 67556188.0,
      "step": 30255
    },
    {
      "entropy": 5.5762636184692385,
      "epoch": 3.398494973886674,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038808594040264127,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22803838104009627,
      "num_tokens": 67567707.0,
      "step": 30260
    },
    {
      "entropy": 5.576560163497925,
      "epoch": 3.399056550794631,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000388051229366394,
      "loss": 4.959,
      "mean_token_accuracy": 0.22167426943778992,
      "num_tokens": 67578334.0,
      "step": 30265
    },
    {
      "entropy": 5.5212408065795895,
      "epoch": 3.3996181277025888,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038801651473055167,
      "loss": 4.7809,
      "mean_token_accuracy": 0.22954143434762955,
      "num_tokens": 67588896.0,
      "step": 30270
    },
    {
      "entropy": 5.606796836853027,
      "epoch": 3.4001797046105464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003879817964962197,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2206254005432129,
      "num_tokens": 67599032.0,
      "step": 30275
    },
    {
      "entropy": 5.6068977355957035,
      "epoch": 3.400741281518504,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003879470746645035,
      "loss": 5.0311,
      "mean_token_accuracy": 0.21556340008974076,
      "num_tokens": 67610095.0,
      "step": 30280
    },
    {
      "entropy": 5.647648429870605,
      "epoch": 3.4013028584264617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038791234923650866,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21646194010972977,
      "num_tokens": 67622039.0,
      "step": 30285
    },
    {
      "entropy": 5.668623256683349,
      "epoch": 3.401864435334419,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038787762021334077,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2208823397755623,
      "num_tokens": 67634123.0,
      "step": 30290
    },
    {
      "entropy": 5.610450887680054,
      "epoch": 3.4024260122423766,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003878428875961057,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22458212822675705,
      "num_tokens": 67644696.0,
      "step": 30295
    },
    {
      "entropy": 5.634199810028076,
      "epoch": 3.4029875891503343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038780815138590935,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2165103405714035,
      "num_tokens": 67656193.0,
      "step": 30300
    },
    {
      "entropy": 5.550748491287232,
      "epoch": 3.4035491660582915,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003877734115838577,
      "loss": 4.8936,
      "mean_token_accuracy": 0.22753604203462602,
      "num_tokens": 67668261.0,
      "step": 30305
    },
    {
      "entropy": 5.620476818084716,
      "epoch": 3.404110742966249,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038773866819105684,
      "loss": 4.9896,
      "mean_token_accuracy": 0.21673968136310579,
      "num_tokens": 67678264.0,
      "step": 30310
    },
    {
      "entropy": 5.64462103843689,
      "epoch": 3.404672319874207,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003877039212086131,
      "loss": 4.9967,
      "mean_token_accuracy": 0.2271535038948059,
      "num_tokens": 67688455.0,
      "step": 30315
    },
    {
      "entropy": 5.562869596481323,
      "epoch": 3.4052338967821645,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038766917063763287,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2282581925392151,
      "num_tokens": 67700970.0,
      "step": 30320
    },
    {
      "entropy": 5.611657333374024,
      "epoch": 3.4057954736901217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038763441647922246,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22381629794836044,
      "num_tokens": 67711835.0,
      "step": 30325
    },
    {
      "entropy": 5.547625112533569,
      "epoch": 3.4063570505980794,
      "grad_norm": 1.125,
      "learning_rate": 0.0003875996587344885,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2260254964232445,
      "num_tokens": 67722915.0,
      "step": 30330
    },
    {
      "entropy": 5.650733947753906,
      "epoch": 3.406918627506037,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038756489740453785,
      "loss": 5.001,
      "mean_token_accuracy": 0.2248174801468849,
      "num_tokens": 67734301.0,
      "step": 30335
    },
    {
      "entropy": 5.615321636199951,
      "epoch": 3.4074802044139947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038753013249047706,
      "loss": 4.998,
      "mean_token_accuracy": 0.22540899962186814,
      "num_tokens": 67745763.0,
      "step": 30340
    },
    {
      "entropy": 5.615873718261719,
      "epoch": 3.408041781321952,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003874953639934134,
      "loss": 4.963,
      "mean_token_accuracy": 0.22985192239284516,
      "num_tokens": 67757241.0,
      "step": 30345
    },
    {
      "entropy": 5.607564640045166,
      "epoch": 3.4086033582299096,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003874605919144536,
      "loss": 4.9724,
      "mean_token_accuracy": 0.22419131845235823,
      "num_tokens": 67768231.0,
      "step": 30350
    },
    {
      "entropy": 5.652108812332154,
      "epoch": 3.4091649351378672,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003874258162547049,
      "loss": 5.085,
      "mean_token_accuracy": 0.21392460018396378,
      "num_tokens": 67780735.0,
      "step": 30355
    },
    {
      "entropy": 5.589298391342163,
      "epoch": 3.4097265120458244,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003873910370152747,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2312761664390564,
      "num_tokens": 67791817.0,
      "step": 30360
    },
    {
      "entropy": 5.659060096740722,
      "epoch": 3.410288088953782,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003873562541972702,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2316821336746216,
      "num_tokens": 67803591.0,
      "step": 30365
    },
    {
      "entropy": 5.626767921447754,
      "epoch": 3.4108496658617398,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003873214678017991,
      "loss": 4.987,
      "mean_token_accuracy": 0.22309007346630097,
      "num_tokens": 67813250.0,
      "step": 30370
    },
    {
      "entropy": 5.538013219833374,
      "epoch": 3.4114112427696974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003872866778299688,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22684767246246337,
      "num_tokens": 67825042.0,
      "step": 30375
    },
    {
      "entropy": 5.647594213485718,
      "epoch": 3.411972819677655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003872518842828872,
      "loss": 5.0286,
      "mean_token_accuracy": 0.21833957135677337,
      "num_tokens": 67837535.0,
      "step": 30380
    },
    {
      "entropy": 5.69045991897583,
      "epoch": 3.4125343965856123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038721708716166203,
      "loss": 5.0402,
      "mean_token_accuracy": 0.22459879219532014,
      "num_tokens": 67848950.0,
      "step": 30385
    },
    {
      "entropy": 5.609149503707886,
      "epoch": 3.41309597349357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003871822864674013,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21604367196559907,
      "num_tokens": 67861488.0,
      "step": 30390
    },
    {
      "entropy": 5.606183958053589,
      "epoch": 3.4136575504015276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038714748220121305,
      "loss": 5.0767,
      "mean_token_accuracy": 0.22150315046310426,
      "num_tokens": 67872627.0,
      "step": 30395
    },
    {
      "entropy": 5.610053205490113,
      "epoch": 3.414219127309485,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003871126743642055,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2172175779938698,
      "num_tokens": 67884033.0,
      "step": 30400
    },
    {
      "entropy": 5.615303039550781,
      "epoch": 3.4147807042174425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038707786295748693,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22159097641706466,
      "num_tokens": 67895696.0,
      "step": 30405
    },
    {
      "entropy": 5.679406690597534,
      "epoch": 3.4153422811254,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038704304798216566,
      "loss": 4.9463,
      "mean_token_accuracy": 0.2313348263502121,
      "num_tokens": 67906328.0,
      "step": 30410
    },
    {
      "entropy": 5.58122706413269,
      "epoch": 3.415903858033358,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003870082294393503,
      "loss": 4.8704,
      "mean_token_accuracy": 0.23074278533458709,
      "num_tokens": 67917258.0,
      "step": 30415
    },
    {
      "entropy": 5.549253797531128,
      "epoch": 3.416465434941315,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003869734073301495,
      "loss": 4.9123,
      "mean_token_accuracy": 0.22260607033967972,
      "num_tokens": 67927431.0,
      "step": 30420
    },
    {
      "entropy": 5.67120132446289,
      "epoch": 3.4170270118492727,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038693858165567195,
      "loss": 5.0166,
      "mean_token_accuracy": 0.22327926456928254,
      "num_tokens": 67938801.0,
      "step": 30425
    },
    {
      "entropy": 5.606635236740113,
      "epoch": 3.4175885887572304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003869037524170266,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22749864161014557,
      "num_tokens": 67949690.0,
      "step": 30430
    },
    {
      "entropy": 5.559078788757324,
      "epoch": 3.418150165665188,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038686891961532227,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22701652050018312,
      "num_tokens": 67959542.0,
      "step": 30435
    },
    {
      "entropy": 5.542855262756348,
      "epoch": 3.4187117425731453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003868340832516683,
      "loss": 4.8625,
      "mean_token_accuracy": 0.229389950633049,
      "num_tokens": 67969759.0,
      "step": 30440
    },
    {
      "entropy": 5.527825212478637,
      "epoch": 3.419273319481103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038679924332717363,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2263535439968109,
      "num_tokens": 67982302.0,
      "step": 30445
    },
    {
      "entropy": 5.619446229934693,
      "epoch": 3.4198348963890606,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038676439984294776,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21991921365261077,
      "num_tokens": 67994452.0,
      "step": 30450
    },
    {
      "entropy": 5.648626852035522,
      "epoch": 3.420396473297018,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038672955280009996,
      "loss": 5.006,
      "mean_token_accuracy": 0.21889083832502365,
      "num_tokens": 68006445.0,
      "step": 30455
    },
    {
      "entropy": 5.661879825592041,
      "epoch": 3.4209580502049755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038669470219973996,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21750044375658034,
      "num_tokens": 68017720.0,
      "step": 30460
    },
    {
      "entropy": 5.643873310089111,
      "epoch": 3.421519627112933,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038665984804297734,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22378359884023666,
      "num_tokens": 68028199.0,
      "step": 30465
    },
    {
      "entropy": 5.63177227973938,
      "epoch": 3.4220812040208908,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003866249903309218,
      "loss": 5.0167,
      "mean_token_accuracy": 0.21584415137767793,
      "num_tokens": 68040171.0,
      "step": 30470
    },
    {
      "entropy": 5.6669450283050535,
      "epoch": 3.4226427809288484,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003865901290646833,
      "loss": 4.981,
      "mean_token_accuracy": 0.2225454032421112,
      "num_tokens": 68050484.0,
      "step": 30475
    },
    {
      "entropy": 5.641538906097412,
      "epoch": 3.4232043578368057,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038655526424537175,
      "loss": 5.0132,
      "mean_token_accuracy": 0.22421416491270066,
      "num_tokens": 68062110.0,
      "step": 30480
    },
    {
      "entropy": 5.638208389282227,
      "epoch": 3.4237659347447633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038652039587409734,
      "loss": 5.0328,
      "mean_token_accuracy": 0.22318679988384246,
      "num_tokens": 68074890.0,
      "step": 30485
    },
    {
      "entropy": 5.604484939575196,
      "epoch": 3.424327511652721,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003864855239519703,
      "loss": 4.9799,
      "mean_token_accuracy": 0.21798235177993774,
      "num_tokens": 68084909.0,
      "step": 30490
    },
    {
      "entropy": 5.549176502227783,
      "epoch": 3.424889088560678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038645064848010093,
      "loss": 4.8675,
      "mean_token_accuracy": 0.23197941482067108,
      "num_tokens": 68097114.0,
      "step": 30495
    },
    {
      "entropy": 5.677799940109253,
      "epoch": 3.425450665468636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038641576945959964,
      "loss": 4.9899,
      "mean_token_accuracy": 0.23333112448453902,
      "num_tokens": 68108377.0,
      "step": 30500
    },
    {
      "entropy": 5.5909992218017575,
      "epoch": 3.4260122423765935,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038638088689157713,
      "loss": 4.9814,
      "mean_token_accuracy": 0.2246622234582901,
      "num_tokens": 68118859.0,
      "step": 30505
    },
    {
      "entropy": 5.562845087051391,
      "epoch": 3.426573819284551,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038634600077714395,
      "loss": 4.9083,
      "mean_token_accuracy": 0.2359718143939972,
      "num_tokens": 68130151.0,
      "step": 30510
    },
    {
      "entropy": 5.566322088241577,
      "epoch": 3.4271353961925084,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003863111111174109,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22171176075935364,
      "num_tokens": 68142514.0,
      "step": 30515
    },
    {
      "entropy": 5.654617786407471,
      "epoch": 3.427696973100466,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038627621791348886,
      "loss": 5.0579,
      "mean_token_accuracy": 0.2203414782881737,
      "num_tokens": 68153751.0,
      "step": 30520
    },
    {
      "entropy": 5.675266122817993,
      "epoch": 3.4282585500084237,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038624132116648897,
      "loss": 4.9676,
      "mean_token_accuracy": 0.23085785508155823,
      "num_tokens": 68164885.0,
      "step": 30525
    },
    {
      "entropy": 5.554609107971191,
      "epoch": 3.4288201269163814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003862064208775223,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2309204876422882,
      "num_tokens": 68176320.0,
      "step": 30530
    },
    {
      "entropy": 5.607319688796997,
      "epoch": 3.4293817038243386,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038617151704769994,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22862659990787507,
      "num_tokens": 68186665.0,
      "step": 30535
    },
    {
      "entropy": 5.655731582641602,
      "epoch": 3.4299432807322963,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003861366096781334,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2175727903842926,
      "num_tokens": 68198011.0,
      "step": 30540
    },
    {
      "entropy": 5.686902046203613,
      "epoch": 3.430504857640254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003861016987699342,
      "loss": 5.0773,
      "mean_token_accuracy": 0.2191153347492218,
      "num_tokens": 68209659.0,
      "step": 30545
    },
    {
      "entropy": 5.619765520095825,
      "epoch": 3.431066434548211,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038606678432421366,
      "loss": 4.901,
      "mean_token_accuracy": 0.23331335335969924,
      "num_tokens": 68220424.0,
      "step": 30550
    },
    {
      "entropy": 5.591808128356933,
      "epoch": 3.431628011456169,
      "grad_norm": 1.125,
      "learning_rate": 0.0003860318663420837,
      "loss": 4.9285,
      "mean_token_accuracy": 0.22478584498167037,
      "num_tokens": 68231413.0,
      "step": 30555
    },
    {
      "entropy": 5.617114543914795,
      "epoch": 3.4321895883641265,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038599694482465606,
      "loss": 4.8956,
      "mean_token_accuracy": 0.23373379707336425,
      "num_tokens": 68242435.0,
      "step": 30560
    },
    {
      "entropy": 5.623822164535523,
      "epoch": 3.432751165272084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038596201977304263,
      "loss": 4.9479,
      "mean_token_accuracy": 0.23218359649181367,
      "num_tokens": 68254462.0,
      "step": 30565
    },
    {
      "entropy": 5.633166980743408,
      "epoch": 3.433312742180042,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038592709118835547,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21927962452173233,
      "num_tokens": 68265842.0,
      "step": 30570
    },
    {
      "entropy": 5.598307847976685,
      "epoch": 3.433874319087999,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038589215907170664,
      "loss": 4.9348,
      "mean_token_accuracy": 0.2250130593776703,
      "num_tokens": 68277117.0,
      "step": 30575
    },
    {
      "entropy": 5.625154161453247,
      "epoch": 3.4344358959959567,
      "grad_norm": 1.25,
      "learning_rate": 0.0003858572234242086,
      "loss": 5.0318,
      "mean_token_accuracy": 0.22632357478141785,
      "num_tokens": 68287869.0,
      "step": 30580
    },
    {
      "entropy": 5.640726137161255,
      "epoch": 3.4349974729039143,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038582228424697347,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22249376326799392,
      "num_tokens": 68300711.0,
      "step": 30585
    },
    {
      "entropy": 5.671656560897827,
      "epoch": 3.4355590498118715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003857873415411138,
      "loss": 5.0653,
      "mean_token_accuracy": 0.22192825973033906,
      "num_tokens": 68311997.0,
      "step": 30590
    },
    {
      "entropy": 5.5805213928222654,
      "epoch": 3.436120626719829,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003857523953077423,
      "loss": 4.9265,
      "mean_token_accuracy": 0.22621593177318572,
      "num_tokens": 68322129.0,
      "step": 30595
    },
    {
      "entropy": 5.548124837875366,
      "epoch": 3.436682203627787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038571744554797156,
      "loss": 4.8874,
      "mean_token_accuracy": 0.23027300238609313,
      "num_tokens": 68334202.0,
      "step": 30600
    },
    {
      "entropy": 5.594081544876099,
      "epoch": 3.4372437805357445,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038568249226291433,
      "loss": 4.9474,
      "mean_token_accuracy": 0.23337027579545974,
      "num_tokens": 68345370.0,
      "step": 30605
    },
    {
      "entropy": 5.544433164596557,
      "epoch": 3.4378053574437017,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038564753545368375,
      "loss": 4.8256,
      "mean_token_accuracy": 0.2306428372859955,
      "num_tokens": 68355703.0,
      "step": 30610
    },
    {
      "entropy": 5.5922520637512205,
      "epoch": 3.4383669343516594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038561257512139264,
      "loss": 4.9838,
      "mean_token_accuracy": 0.21718621850013733,
      "num_tokens": 68367412.0,
      "step": 30615
    },
    {
      "entropy": 5.609319019317627,
      "epoch": 3.438928511259617,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038557761126715426,
      "loss": 4.9366,
      "mean_token_accuracy": 0.22822362333536148,
      "num_tokens": 68378074.0,
      "step": 30620
    },
    {
      "entropy": 5.628255319595337,
      "epoch": 3.4394900881675747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003855426438920818,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21959949284791946,
      "num_tokens": 68388730.0,
      "step": 30625
    },
    {
      "entropy": 5.573291969299317,
      "epoch": 3.440051665075532,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038550767299728875,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22583104074001312,
      "num_tokens": 68399172.0,
      "step": 30630
    },
    {
      "entropy": 5.627045059204102,
      "epoch": 3.4406132419834896,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038547269858388853,
      "loss": 4.996,
      "mean_token_accuracy": 0.22877585589885713,
      "num_tokens": 68411704.0,
      "step": 30635
    },
    {
      "entropy": 5.57098159790039,
      "epoch": 3.4411748188914473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003854377206529947,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22018378973007202,
      "num_tokens": 68422774.0,
      "step": 30640
    },
    {
      "entropy": 5.585470485687256,
      "epoch": 3.4417363957994045,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000385402739205721,
      "loss": 4.8591,
      "mean_token_accuracy": 0.234435798227787,
      "num_tokens": 68432964.0,
      "step": 30645
    },
    {
      "entropy": 5.595256996154785,
      "epoch": 3.442297972707362,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003853677542431813,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2178899183869362,
      "num_tokens": 68445875.0,
      "step": 30650
    },
    {
      "entropy": 5.684937381744385,
      "epoch": 3.44285954961532,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003853327657664895,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2193182408809662,
      "num_tokens": 68457972.0,
      "step": 30655
    },
    {
      "entropy": 5.592076253890991,
      "epoch": 3.4434211265232775,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003852977737767596,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2276830717921257,
      "num_tokens": 68468825.0,
      "step": 30660
    },
    {
      "entropy": 5.548013973236084,
      "epoch": 3.443982703431235,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003852627782751059,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23588606268167495,
      "num_tokens": 68479427.0,
      "step": 30665
    },
    {
      "entropy": 5.5307403087615965,
      "epoch": 3.4445442803391924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003852277792626425,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2283538356423378,
      "num_tokens": 68490542.0,
      "step": 30670
    },
    {
      "entropy": 5.512075614929199,
      "epoch": 3.44510585724715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003851927767404839,
      "loss": 4.9124,
      "mean_token_accuracy": 0.23161577433347702,
      "num_tokens": 68501396.0,
      "step": 30675
    },
    {
      "entropy": 5.683462142944336,
      "epoch": 3.4456674341551077,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003851577707097445,
      "loss": 5.0883,
      "mean_token_accuracy": 0.2106744185090065,
      "num_tokens": 68512710.0,
      "step": 30680
    },
    {
      "entropy": 5.601498651504516,
      "epoch": 3.446229011063065,
      "grad_norm": 1.03125,
      "learning_rate": 0.000385122761171539,
      "loss": 4.94,
      "mean_token_accuracy": 0.222114859521389,
      "num_tokens": 68524289.0,
      "step": 30685
    },
    {
      "entropy": 5.581991910934448,
      "epoch": 3.4467905879710226,
      "grad_norm": 1.0,
      "learning_rate": 0.00038508774812698207,
      "loss": 4.9488,
      "mean_token_accuracy": 0.22357377409934998,
      "num_tokens": 68536783.0,
      "step": 30690
    },
    {
      "entropy": 5.622279024124145,
      "epoch": 3.44735216487898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003850527315771885,
      "loss": 5.0115,
      "mean_token_accuracy": 0.22019148170948027,
      "num_tokens": 68548113.0,
      "step": 30695
    },
    {
      "entropy": 5.60187726020813,
      "epoch": 3.447913741786938,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038501771152327333,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22881070226430894,
      "num_tokens": 68560234.0,
      "step": 30700
    },
    {
      "entropy": 5.571246099472046,
      "epoch": 3.448475318694895,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038498268796635155,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2290831670165062,
      "num_tokens": 68571712.0,
      "step": 30705
    },
    {
      "entropy": 5.540252780914306,
      "epoch": 3.4490368956028528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038494766090753824,
      "loss": 4.912,
      "mean_token_accuracy": 0.22731430679559708,
      "num_tokens": 68583166.0,
      "step": 30710
    },
    {
      "entropy": 5.553392028808593,
      "epoch": 3.4495984725108104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038491263034794886,
      "loss": 4.8267,
      "mean_token_accuracy": 0.24211188703775405,
      "num_tokens": 68594026.0,
      "step": 30715
    },
    {
      "entropy": 5.535852432250977,
      "epoch": 3.450160049418768,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003848775962886987,
      "loss": 4.9118,
      "mean_token_accuracy": 0.22372815012931824,
      "num_tokens": 68604046.0,
      "step": 30720
    },
    {
      "entropy": 5.5290789127349855,
      "epoch": 3.4507216263267253,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003848425587309032,
      "loss": 4.8724,
      "mean_token_accuracy": 0.23336535841226577,
      "num_tokens": 68615575.0,
      "step": 30725
    },
    {
      "entropy": 5.632025146484375,
      "epoch": 3.451283203234683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038480751767567807,
      "loss": 5.052,
      "mean_token_accuracy": 0.21808125227689742,
      "num_tokens": 68626805.0,
      "step": 30730
    },
    {
      "entropy": 5.7174746036529545,
      "epoch": 3.4518447801426406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000384772473124139,
      "loss": 5.0565,
      "mean_token_accuracy": 0.2179803356528282,
      "num_tokens": 68638802.0,
      "step": 30735
    },
    {
      "entropy": 5.61834306716919,
      "epoch": 3.452406357050598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003847374250774018,
      "loss": 5.0433,
      "mean_token_accuracy": 0.22289520055055617,
      "num_tokens": 68651137.0,
      "step": 30740
    },
    {
      "entropy": 5.544924736022949,
      "epoch": 3.4529679339585555,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038470237353658237,
      "loss": 4.9032,
      "mean_token_accuracy": 0.23421748727560043,
      "num_tokens": 68661268.0,
      "step": 30745
    },
    {
      "entropy": 5.489525604248047,
      "epoch": 3.453529510866513,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038466731850279683,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22640616297721863,
      "num_tokens": 68672334.0,
      "step": 30750
    },
    {
      "entropy": 5.590087127685547,
      "epoch": 3.454091087774471,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038463225997716137,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22504039704799653,
      "num_tokens": 68683765.0,
      "step": 30755
    },
    {
      "entropy": 5.754266834259033,
      "epoch": 3.4546526646824285,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038459719796079227,
      "loss": 5.0857,
      "mean_token_accuracy": 0.2190423384308815,
      "num_tokens": 68694969.0,
      "step": 30760
    },
    {
      "entropy": 5.656019592285157,
      "epoch": 3.4552142415903857,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003845621324548059,
      "loss": 5.0462,
      "mean_token_accuracy": 0.21791984289884567,
      "num_tokens": 68708586.0,
      "step": 30765
    },
    {
      "entropy": 5.659211015701294,
      "epoch": 3.4557758184983434,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003845270634603186,
      "loss": 4.9815,
      "mean_token_accuracy": 0.22382639944553376,
      "num_tokens": 68719649.0,
      "step": 30770
    },
    {
      "entropy": 5.646788883209228,
      "epoch": 3.456337395406301,
      "grad_norm": 1.125,
      "learning_rate": 0.00038449199097844724,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2217692628502846,
      "num_tokens": 68729491.0,
      "step": 30775
    },
    {
      "entropy": 5.599238395690918,
      "epoch": 3.4568989723142582,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003844569150103084,
      "loss": 4.9264,
      "mean_token_accuracy": 0.22615962773561477,
      "num_tokens": 68741087.0,
      "step": 30780
    },
    {
      "entropy": 5.59586877822876,
      "epoch": 3.457460549222216,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003844218355570189,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23099910616874694,
      "num_tokens": 68752699.0,
      "step": 30785
    },
    {
      "entropy": 5.6122785091400145,
      "epoch": 3.4580221261301736,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003843867526196957,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22426115721464157,
      "num_tokens": 68763066.0,
      "step": 30790
    },
    {
      "entropy": 5.606549215316773,
      "epoch": 3.4585837030381312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003843516661994559,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22673625648021697,
      "num_tokens": 68773878.0,
      "step": 30795
    },
    {
      "entropy": 5.55983190536499,
      "epoch": 3.4591452799460884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003843165762974167,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22978467345237732,
      "num_tokens": 68783864.0,
      "step": 30800
    },
    {
      "entropy": 5.575194787979126,
      "epoch": 3.459706856854046,
      "grad_norm": 1.125,
      "learning_rate": 0.00038428148291469515,
      "loss": 4.9016,
      "mean_token_accuracy": 0.23124938160181047,
      "num_tokens": 68794280.0,
      "step": 30805
    },
    {
      "entropy": 5.674701118469239,
      "epoch": 3.4602684337620038,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000384246386052409,
      "loss": 5.0448,
      "mean_token_accuracy": 0.22247038334608077,
      "num_tokens": 68806125.0,
      "step": 30810
    },
    {
      "entropy": 5.633944368362426,
      "epoch": 3.4608300106699614,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003842112857116754,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2255101501941681,
      "num_tokens": 68816742.0,
      "step": 30815
    },
    {
      "entropy": 5.55916395187378,
      "epoch": 3.4613915875779186,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003841761818936122,
      "loss": 4.9558,
      "mean_token_accuracy": 0.22513847500085832,
      "num_tokens": 68828504.0,
      "step": 30820
    },
    {
      "entropy": 5.690720415115356,
      "epoch": 3.4619531644858763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000384141074599337,
      "loss": 5.0302,
      "mean_token_accuracy": 0.2162155032157898,
      "num_tokens": 68839993.0,
      "step": 30825
    },
    {
      "entropy": 5.552586221694947,
      "epoch": 3.462514741393834,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038410596382996764,
      "loss": 4.8645,
      "mean_token_accuracy": 0.23417493253946303,
      "num_tokens": 68850395.0,
      "step": 30830
    },
    {
      "entropy": 5.5869636058807375,
      "epoch": 3.463076318301791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003840708495866221,
      "loss": 4.8714,
      "mean_token_accuracy": 0.2252185046672821,
      "num_tokens": 68860621.0,
      "step": 30835
    },
    {
      "entropy": 5.583223581314087,
      "epoch": 3.463637895209749,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003840357318704183,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22129344940185547,
      "num_tokens": 68872838.0,
      "step": 30840
    },
    {
      "entropy": 5.6169548511505125,
      "epoch": 3.4641994721177065,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003840006106824746,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2283323422074318,
      "num_tokens": 68883372.0,
      "step": 30845
    },
    {
      "entropy": 5.626774311065674,
      "epoch": 3.464761049025664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003839654860239091,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2286691576242447,
      "num_tokens": 68894148.0,
      "step": 30850
    },
    {
      "entropy": 5.657864332199097,
      "epoch": 3.465322625933622,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003839303578958404,
      "loss": 4.9823,
      "mean_token_accuracy": 0.22892193198204042,
      "num_tokens": 68904381.0,
      "step": 30855
    },
    {
      "entropy": 5.619052505493164,
      "epoch": 3.465884202841579,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038389522629938676,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21839626729488373,
      "num_tokens": 68916354.0,
      "step": 30860
    },
    {
      "entropy": 5.561387634277343,
      "epoch": 3.4664457797495367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003838600912356669,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22960263043642043,
      "num_tokens": 68927551.0,
      "step": 30865
    },
    {
      "entropy": 5.692802000045776,
      "epoch": 3.4670073566574944,
      "grad_norm": 1.125,
      "learning_rate": 0.0003838249527057995,
      "loss": 5.0355,
      "mean_token_accuracy": 0.22505109310150145,
      "num_tokens": 68939528.0,
      "step": 30870
    },
    {
      "entropy": 5.596079349517822,
      "epoch": 3.4675689335654516,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038378981071090346,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22722901105880738,
      "num_tokens": 68951634.0,
      "step": 30875
    },
    {
      "entropy": 5.60067777633667,
      "epoch": 3.4681305104734093,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038375466525209764,
      "loss": 4.9115,
      "mean_token_accuracy": 0.23839885145425796,
      "num_tokens": 68962259.0,
      "step": 30880
    },
    {
      "entropy": 5.5662683010101315,
      "epoch": 3.468692087381367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038371951633050106,
      "loss": 5.0229,
      "mean_token_accuracy": 0.22648354023694992,
      "num_tokens": 68973185.0,
      "step": 30885
    },
    {
      "entropy": 5.6693190574646,
      "epoch": 3.4692536642893246,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038368436394723294,
      "loss": 5.0165,
      "mean_token_accuracy": 0.21942865550518037,
      "num_tokens": 68984052.0,
      "step": 30890
    },
    {
      "entropy": 5.628211307525635,
      "epoch": 3.469815241197282,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038364920810341255,
      "loss": 5.012,
      "mean_token_accuracy": 0.22026114612817765,
      "num_tokens": 68993730.0,
      "step": 30895
    },
    {
      "entropy": 5.6757289409637455,
      "epoch": 3.4703768181052395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003836140488001592,
      "loss": 4.9988,
      "mean_token_accuracy": 0.22100814133882524,
      "num_tokens": 69006229.0,
      "step": 30900
    },
    {
      "entropy": 5.591998720169068,
      "epoch": 3.470938395013197,
      "grad_norm": 1.125,
      "learning_rate": 0.00038357888603859236,
      "loss": 5.0015,
      "mean_token_accuracy": 0.22301123589277266,
      "num_tokens": 69018518.0,
      "step": 30905
    },
    {
      "entropy": 5.622909879684448,
      "epoch": 3.471499971921155,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038354371981983175,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21956004798412324,
      "num_tokens": 69031113.0,
      "step": 30910
    },
    {
      "entropy": 5.704088115692139,
      "epoch": 3.472061548829112,
      "grad_norm": 1.125,
      "learning_rate": 0.0003835085501449969,
      "loss": 5.0966,
      "mean_token_accuracy": 0.21573621332645415,
      "num_tokens": 69042038.0,
      "step": 30915
    },
    {
      "entropy": 5.560885238647461,
      "epoch": 3.4726231257370697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038347337701520785,
      "loss": 4.8821,
      "mean_token_accuracy": 0.2385395035147667,
      "num_tokens": 69053675.0,
      "step": 30920
    },
    {
      "entropy": 5.61664309501648,
      "epoch": 3.4731847026450273,
      "grad_norm": 1.125,
      "learning_rate": 0.00038343820043158436,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21817353516817092,
      "num_tokens": 69064475.0,
      "step": 30925
    },
    {
      "entropy": 5.574958562850952,
      "epoch": 3.4737462795529845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003834030203952465,
      "loss": 4.9479,
      "mean_token_accuracy": 0.22585614621639252,
      "num_tokens": 69076069.0,
      "step": 30930
    },
    {
      "entropy": 5.653940868377686,
      "epoch": 3.474307856460942,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003833678369073144,
      "loss": 5.0465,
      "mean_token_accuracy": 0.2200354442000389,
      "num_tokens": 69087144.0,
      "step": 30935
    },
    {
      "entropy": 5.620909309387207,
      "epoch": 3.4748694333689,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003833326499689084,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22005306482315062,
      "num_tokens": 69098035.0,
      "step": 30940
    },
    {
      "entropy": 5.66155595779419,
      "epoch": 3.4754310102768575,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003832974595811488,
      "loss": 4.9878,
      "mean_token_accuracy": 0.23329222053289414,
      "num_tokens": 69111722.0,
      "step": 30945
    },
    {
      "entropy": 5.63422999382019,
      "epoch": 3.475992587184815,
      "grad_norm": 1.09375,
      "learning_rate": 0.000383262265745156,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22805160284042358,
      "num_tokens": 69122228.0,
      "step": 30950
    },
    {
      "entropy": 5.61802716255188,
      "epoch": 3.4765541640927724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038322706846205073,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22922816574573518,
      "num_tokens": 69134150.0,
      "step": 30955
    },
    {
      "entropy": 5.533202123641968,
      "epoch": 3.47711574100073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038319186773295364,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22085291743278504,
      "num_tokens": 69145757.0,
      "step": 30960
    },
    {
      "entropy": 5.7020762920379635,
      "epoch": 3.4776773179086877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003831566635589856,
      "loss": 5.0556,
      "mean_token_accuracy": 0.22399376779794694,
      "num_tokens": 69157545.0,
      "step": 30965
    },
    {
      "entropy": 5.638968801498413,
      "epoch": 3.478238894816645,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003831214559412674,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22580737322568895,
      "num_tokens": 69169070.0,
      "step": 30970
    },
    {
      "entropy": 5.604268980026245,
      "epoch": 3.4788004717246026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038308624488092005,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22743292152881622,
      "num_tokens": 69178903.0,
      "step": 30975
    },
    {
      "entropy": 5.56746768951416,
      "epoch": 3.4793620486325603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003830510303790647,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22851233333349227,
      "num_tokens": 69190489.0,
      "step": 30980
    },
    {
      "entropy": 5.524733161926269,
      "epoch": 3.479923625540518,
      "grad_norm": 1.125,
      "learning_rate": 0.00038301581243682263,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22671681344509126,
      "num_tokens": 69200574.0,
      "step": 30985
    },
    {
      "entropy": 5.55325231552124,
      "epoch": 3.480485202448475,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003829805910553153,
      "loss": 4.8666,
      "mean_token_accuracy": 0.23640053868293762,
      "num_tokens": 69210597.0,
      "step": 30990
    },
    {
      "entropy": 5.669768810272217,
      "epoch": 3.481046779356433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038294536623566407,
      "loss": 4.9977,
      "mean_token_accuracy": 0.22450327575206758,
      "num_tokens": 69221392.0,
      "step": 30995
    },
    {
      "entropy": 5.50048189163208,
      "epoch": 3.4816083562643905,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003829101379789905,
      "loss": 4.8767,
      "mean_token_accuracy": 0.22482903748750688,
      "num_tokens": 69232825.0,
      "step": 31000
    },
    {
      "entropy": 5.593155002593994,
      "epoch": 3.482169933172348,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003828749062864162,
      "loss": 4.9625,
      "mean_token_accuracy": 0.2242489740252495,
      "num_tokens": 69243989.0,
      "step": 31005
    },
    {
      "entropy": 5.638734006881714,
      "epoch": 3.4827315100803053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003828396711590631,
      "loss": 5.032,
      "mean_token_accuracy": 0.21798980236053467,
      "num_tokens": 69255828.0,
      "step": 31010
    },
    {
      "entropy": 5.627222156524658,
      "epoch": 3.483293086988263,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038280443259805307,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22470954805612564,
      "num_tokens": 69266496.0,
      "step": 31015
    },
    {
      "entropy": 5.5796669006347654,
      "epoch": 3.4838546638962207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003827691906045081,
      "loss": 5.0368,
      "mean_token_accuracy": 0.2177361950278282,
      "num_tokens": 69277884.0,
      "step": 31020
    },
    {
      "entropy": 5.681703996658325,
      "epoch": 3.484416240804178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003827339451795503,
      "loss": 5.1185,
      "mean_token_accuracy": 0.21563642621040344,
      "num_tokens": 69290660.0,
      "step": 31025
    },
    {
      "entropy": 5.730710792541504,
      "epoch": 3.4849778177121356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003826986963243018,
      "loss": 5.0842,
      "mean_token_accuracy": 0.22975359559059144,
      "num_tokens": 69302310.0,
      "step": 31030
    },
    {
      "entropy": 5.674553251266479,
      "epoch": 3.485539394620093,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003826634440398852,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2225155711174011,
      "num_tokens": 69313804.0,
      "step": 31035
    },
    {
      "entropy": 5.581120491027832,
      "epoch": 3.486100971528051,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038262818832742276,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22438670247793197,
      "num_tokens": 69325372.0,
      "step": 31040
    },
    {
      "entropy": 5.589874219894409,
      "epoch": 3.4866625484360085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038259292918803696,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22330953925848007,
      "num_tokens": 69335835.0,
      "step": 31045
    },
    {
      "entropy": 5.596391010284424,
      "epoch": 3.4872241253439658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003825576666228506,
      "loss": 4.9632,
      "mean_token_accuracy": 0.21886592209339142,
      "num_tokens": 69346895.0,
      "step": 31050
    },
    {
      "entropy": 5.602743530273438,
      "epoch": 3.4877857022519234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003825224006329864,
      "loss": 4.9619,
      "mean_token_accuracy": 0.2291029006242752,
      "num_tokens": 69357649.0,
      "step": 31055
    },
    {
      "entropy": 5.645297718048096,
      "epoch": 3.488347279159881,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003824871312195673,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2229963943362236,
      "num_tokens": 69367973.0,
      "step": 31060
    },
    {
      "entropy": 5.650474691390992,
      "epoch": 3.4889088560678383,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003824518583837163,
      "loss": 4.9174,
      "mean_token_accuracy": 0.23468296080827714,
      "num_tokens": 69379411.0,
      "step": 31065
    },
    {
      "entropy": 5.655237197875977,
      "epoch": 3.489470432975796,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003824165821265564,
      "loss": 4.9953,
      "mean_token_accuracy": 0.22361326217651367,
      "num_tokens": 69391261.0,
      "step": 31070
    },
    {
      "entropy": 5.549314880371094,
      "epoch": 3.4900320098837536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003823813024492108,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23097708821296692,
      "num_tokens": 69401712.0,
      "step": 31075
    },
    {
      "entropy": 5.528225421905518,
      "epoch": 3.4905935867917113,
      "grad_norm": 1.03125,
      "learning_rate": 0.000382346019352803,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2363419771194458,
      "num_tokens": 69412612.0,
      "step": 31080
    },
    {
      "entropy": 5.62422285079956,
      "epoch": 3.4911551636996685,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003823107328384562,
      "loss": 5.0109,
      "mean_token_accuracy": 0.22640868723392488,
      "num_tokens": 69424042.0,
      "step": 31085
    },
    {
      "entropy": 5.623110723495484,
      "epoch": 3.491716740607626,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003822754429072942,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2318141132593155,
      "num_tokens": 69434667.0,
      "step": 31090
    },
    {
      "entropy": 5.689328908920288,
      "epoch": 3.492278317515584,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003822401495604403,
      "loss": 5.0613,
      "mean_token_accuracy": 0.2218056544661522,
      "num_tokens": 69447015.0,
      "step": 31095
    },
    {
      "entropy": 5.573739242553711,
      "epoch": 3.4928398944235415,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038220485279901855,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22298658341169358,
      "num_tokens": 69457585.0,
      "step": 31100
    },
    {
      "entropy": 5.57650990486145,
      "epoch": 3.4934014713314987,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038216955262415273,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2197430357336998,
      "num_tokens": 69467751.0,
      "step": 31105
    },
    {
      "entropy": 5.67989330291748,
      "epoch": 3.4939630482394564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038213424903696673,
      "loss": 5.0696,
      "mean_token_accuracy": 0.22185358554124832,
      "num_tokens": 69478898.0,
      "step": 31110
    },
    {
      "entropy": 5.672914600372314,
      "epoch": 3.494524625147414,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038209894203858477,
      "loss": 5.0112,
      "mean_token_accuracy": 0.2177964061498642,
      "num_tokens": 69488367.0,
      "step": 31115
    },
    {
      "entropy": 5.595298099517822,
      "epoch": 3.4950862020553712,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038206363163013093,
      "loss": 4.9771,
      "mean_token_accuracy": 0.23080790787935257,
      "num_tokens": 69500446.0,
      "step": 31120
    },
    {
      "entropy": 5.615807104110718,
      "epoch": 3.495647778963329,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038202831781272944,
      "loss": 5.0253,
      "mean_token_accuracy": 0.22346668392419816,
      "num_tokens": 69512383.0,
      "step": 31125
    },
    {
      "entropy": 5.59432315826416,
      "epoch": 3.4962093558712866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038199300058750494,
      "loss": 4.9237,
      "mean_token_accuracy": 0.2330954849720001,
      "num_tokens": 69523238.0,
      "step": 31130
    },
    {
      "entropy": 5.7544474601745605,
      "epoch": 3.4967709327792442,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038195767995558174,
      "loss": 5.0832,
      "mean_token_accuracy": 0.22064117789268495,
      "num_tokens": 69535196.0,
      "step": 31135
    },
    {
      "entropy": 5.641109561920166,
      "epoch": 3.497332509687202,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003819223559180845,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22799086719751357,
      "num_tokens": 69546896.0,
      "step": 31140
    },
    {
      "entropy": 5.5590136528015135,
      "epoch": 3.497894086595159,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000381887028476138,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2252562314271927,
      "num_tokens": 69558230.0,
      "step": 31145
    },
    {
      "entropy": 5.532695770263672,
      "epoch": 3.4984556635031168,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038185169763086703,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23073282539844514,
      "num_tokens": 69568891.0,
      "step": 31150
    },
    {
      "entropy": 5.624896144866943,
      "epoch": 3.4990172404110744,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003818163633833966,
      "loss": 5.021,
      "mean_token_accuracy": 0.22178518772125244,
      "num_tokens": 69579988.0,
      "step": 31155
    },
    {
      "entropy": 5.601309108734131,
      "epoch": 3.4995788173190316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038178102573485173,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22643479853868484,
      "num_tokens": 69590714.0,
      "step": 31160
    },
    {
      "entropy": 5.561829376220703,
      "epoch": 3.5001403942269893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003817456846863575,
      "loss": 4.9112,
      "mean_token_accuracy": 0.237107352912426,
      "num_tokens": 69601755.0,
      "step": 31165
    },
    {
      "entropy": 5.569378089904785,
      "epoch": 3.500701971134947,
      "grad_norm": 1.125,
      "learning_rate": 0.00038171034023903937,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2284553736448288,
      "num_tokens": 69612862.0,
      "step": 31170
    },
    {
      "entropy": 5.725499677658081,
      "epoch": 3.5012635480429046,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003816749923940226,
      "loss": 5.0117,
      "mean_token_accuracy": 0.2228134647011757,
      "num_tokens": 69624746.0,
      "step": 31175
    },
    {
      "entropy": 5.6124366283416744,
      "epoch": 3.5018251249508623,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038163964115243276,
      "loss": 4.8482,
      "mean_token_accuracy": 0.23332529962062837,
      "num_tokens": 69634477.0,
      "step": 31180
    },
    {
      "entropy": 5.617247676849365,
      "epoch": 3.5023867018588195,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038160428651539535,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21609114855527878,
      "num_tokens": 69645913.0,
      "step": 31185
    },
    {
      "entropy": 5.56106014251709,
      "epoch": 3.502948278766777,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038156892848403606,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22692551761865615,
      "num_tokens": 69657727.0,
      "step": 31190
    },
    {
      "entropy": 5.52779803276062,
      "epoch": 3.503509855674735,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003815335670594808,
      "loss": 4.8127,
      "mean_token_accuracy": 0.2532953187823296,
      "num_tokens": 69668649.0,
      "step": 31195
    },
    {
      "entropy": 5.622902202606201,
      "epoch": 3.504071432582692,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038149820224285546,
      "loss": 5.0288,
      "mean_token_accuracy": 0.22101829946041107,
      "num_tokens": 69680286.0,
      "step": 31200
    },
    {
      "entropy": 5.602732086181641,
      "epoch": 3.5046330094906497,
      "grad_norm": 1.3125,
      "learning_rate": 0.000381462834035286,
      "loss": 4.9523,
      "mean_token_accuracy": 0.22258362919092178,
      "num_tokens": 69689801.0,
      "step": 31205
    },
    {
      "entropy": 5.630863380432129,
      "epoch": 3.5051945863986074,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003814274624378986,
      "loss": 4.9616,
      "mean_token_accuracy": 0.22988025099039078,
      "num_tokens": 69700898.0,
      "step": 31210
    },
    {
      "entropy": 5.614192724227905,
      "epoch": 3.5057561633065646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003813920874518196,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23314825892448426,
      "num_tokens": 69711969.0,
      "step": 31215
    },
    {
      "entropy": 5.611251544952393,
      "epoch": 3.5063177402145222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003813567090781752,
      "loss": 4.983,
      "mean_token_accuracy": 0.22602288275957108,
      "num_tokens": 69723458.0,
      "step": 31220
    },
    {
      "entropy": 5.5658571243286135,
      "epoch": 3.50687931712248,
      "grad_norm": 1.015625,
      "learning_rate": 0.000381321327318092,
      "loss": 4.9364,
      "mean_token_accuracy": 0.2269906744360924,
      "num_tokens": 69735527.0,
      "step": 31225
    },
    {
      "entropy": 5.658479690551758,
      "epoch": 3.5074408940304376,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003812859421726964,
      "loss": 5.0359,
      "mean_token_accuracy": 0.2267757087945938,
      "num_tokens": 69748829.0,
      "step": 31230
    },
    {
      "entropy": 5.626160144805908,
      "epoch": 3.5080024709383952,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038125055364311526,
      "loss": 4.849,
      "mean_token_accuracy": 0.22813363373279572,
      "num_tokens": 69758970.0,
      "step": 31235
    },
    {
      "entropy": 5.642369508743286,
      "epoch": 3.5085640478463525,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003812151617304753,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21927174627780915,
      "num_tokens": 69770082.0,
      "step": 31240
    },
    {
      "entropy": 5.602598237991333,
      "epoch": 3.50912562475431,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003811797664359033,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22263611555099488,
      "num_tokens": 69781386.0,
      "step": 31245
    },
    {
      "entropy": 5.660308218002319,
      "epoch": 3.5096872016622678,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038114436776052644,
      "loss": 4.9934,
      "mean_token_accuracy": 0.22116755396127702,
      "num_tokens": 69792571.0,
      "step": 31250
    },
    {
      "entropy": 5.6746570587158205,
      "epoch": 3.510248778570225,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003811089657054717,
      "loss": 5.0053,
      "mean_token_accuracy": 0.22229277044534684,
      "num_tokens": 69804737.0,
      "step": 31255
    },
    {
      "entropy": 5.649727535247803,
      "epoch": 3.5108103554781827,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003810735602718664,
      "loss": 5.0538,
      "mean_token_accuracy": 0.2304052457213402,
      "num_tokens": 69816584.0,
      "step": 31260
    },
    {
      "entropy": 5.5366734027862545,
      "epoch": 3.5113719323861403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003810381514608378,
      "loss": 4.834,
      "mean_token_accuracy": 0.24419038295745848,
      "num_tokens": 69828657.0,
      "step": 31265
    },
    {
      "entropy": 5.625572872161865,
      "epoch": 3.511933509294098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003810027392735133,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22615618109703065,
      "num_tokens": 69839780.0,
      "step": 31270
    },
    {
      "entropy": 5.587393474578858,
      "epoch": 3.5124950862020556,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003809673237110204,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22946233302354813,
      "num_tokens": 69850452.0,
      "step": 31275
    },
    {
      "entropy": 5.614347743988037,
      "epoch": 3.513056663110013,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003809319047744869,
      "loss": 4.9942,
      "mean_token_accuracy": 0.22104823738336563,
      "num_tokens": 69861840.0,
      "step": 31280
    },
    {
      "entropy": 5.66679573059082,
      "epoch": 3.5136182400179705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038089648246504045,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22608171850442887,
      "num_tokens": 69872653.0,
      "step": 31285
    },
    {
      "entropy": 5.596135902404785,
      "epoch": 3.514179816925928,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038086105678380896,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22242544293403627,
      "num_tokens": 69883381.0,
      "step": 31290
    },
    {
      "entropy": 5.686621522903442,
      "epoch": 3.5147413938338854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038082562773192026,
      "loss": 5.0598,
      "mean_token_accuracy": 0.222451551258564,
      "num_tokens": 69894732.0,
      "step": 31295
    },
    {
      "entropy": 5.73718523979187,
      "epoch": 3.515302970741843,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038079019531050257,
      "loss": 5.0099,
      "mean_token_accuracy": 0.22654856145381927,
      "num_tokens": 69905645.0,
      "step": 31300
    },
    {
      "entropy": 5.692336988449097,
      "epoch": 3.5158645476498007,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038075475952068405,
      "loss": 4.9856,
      "mean_token_accuracy": 0.2206431284546852,
      "num_tokens": 69917176.0,
      "step": 31305
    },
    {
      "entropy": 5.635174751281738,
      "epoch": 3.516426124557758,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000380719320363593,
      "loss": 5.065,
      "mean_token_accuracy": 0.22432641834020614,
      "num_tokens": 69928923.0,
      "step": 31310
    },
    {
      "entropy": 5.581644201278687,
      "epoch": 3.5169877014657156,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038068387784035774,
      "loss": 4.988,
      "mean_token_accuracy": 0.22067814469337463,
      "num_tokens": 69938981.0,
      "step": 31315
    },
    {
      "entropy": 5.567266845703125,
      "epoch": 3.5175492783736733,
      "grad_norm": 1.125,
      "learning_rate": 0.0003806484319521069,
      "loss": 4.9173,
      "mean_token_accuracy": 0.22734318673610687,
      "num_tokens": 69949265.0,
      "step": 31320
    },
    {
      "entropy": 5.505045366287232,
      "epoch": 3.518110855281631,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038061298269996893,
      "loss": 4.836,
      "mean_token_accuracy": 0.2298795148730278,
      "num_tokens": 69961399.0,
      "step": 31325
    },
    {
      "entropy": 5.627475357055664,
      "epoch": 3.5186724321895886,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003805775300850727,
      "loss": 4.9508,
      "mean_token_accuracy": 0.2237937018275261,
      "num_tokens": 69971601.0,
      "step": 31330
    },
    {
      "entropy": 5.582835626602173,
      "epoch": 3.519234009097546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003805420741085469,
      "loss": 4.9599,
      "mean_token_accuracy": 0.21393674612045288,
      "num_tokens": 69982375.0,
      "step": 31335
    },
    {
      "entropy": 5.615369033813477,
      "epoch": 3.5197955860055035,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003805066147715205,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2316150337457657,
      "num_tokens": 69994787.0,
      "step": 31340
    },
    {
      "entropy": 5.718700408935547,
      "epoch": 3.520357162913461,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003804711520751226,
      "loss": 5.1407,
      "mean_token_accuracy": 0.2144143044948578,
      "num_tokens": 70006077.0,
      "step": 31345
    },
    {
      "entropy": 5.644205951690674,
      "epoch": 3.5209187398214183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038043568602048227,
      "loss": 4.9407,
      "mean_token_accuracy": 0.22394158840179443,
      "num_tokens": 70016655.0,
      "step": 31350
    },
    {
      "entropy": 5.6333780765533445,
      "epoch": 3.521480316729376,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003804002166087289,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22858455330133437,
      "num_tokens": 70026857.0,
      "step": 31355
    },
    {
      "entropy": 5.568215084075928,
      "epoch": 3.5220418936373337,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038036474384099174,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22835718244314193,
      "num_tokens": 70038311.0,
      "step": 31360
    },
    {
      "entropy": 5.588877582550049,
      "epoch": 3.5226034705452913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003803292677184002,
      "loss": 4.941,
      "mean_token_accuracy": 0.22754995226860047,
      "num_tokens": 70048613.0,
      "step": 31365
    },
    {
      "entropy": 5.716201496124268,
      "epoch": 3.523165047453249,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038029378824208394,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22881992757320405,
      "num_tokens": 70058686.0,
      "step": 31370
    },
    {
      "entropy": 5.661733913421631,
      "epoch": 3.523726624361206,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003802583054131727,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22336457073688507,
      "num_tokens": 70068845.0,
      "step": 31375
    },
    {
      "entropy": 5.53814115524292,
      "epoch": 3.524288201269164,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003802228192327961,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21937679946422578,
      "num_tokens": 70080223.0,
      "step": 31380
    },
    {
      "entropy": 5.62872347831726,
      "epoch": 3.5248497781771215,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038018732970208414,
      "loss": 5.0271,
      "mean_token_accuracy": 0.22265533059835435,
      "num_tokens": 70091729.0,
      "step": 31385
    },
    {
      "entropy": 5.566092014312744,
      "epoch": 3.5254113550850787,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003801518368221668,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22206970453262329,
      "num_tokens": 70103172.0,
      "step": 31390
    },
    {
      "entropy": 5.638320112228394,
      "epoch": 3.5259729319930364,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003801163405941743,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2168669730424881,
      "num_tokens": 70114303.0,
      "step": 31395
    },
    {
      "entropy": 5.586967945098877,
      "epoch": 3.526534508900994,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038008084101923665,
      "loss": 4.9381,
      "mean_token_accuracy": 0.22384575605392457,
      "num_tokens": 70126802.0,
      "step": 31400
    },
    {
      "entropy": 5.714948844909668,
      "epoch": 3.5270960858089513,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038004533809848435,
      "loss": 4.9613,
      "mean_token_accuracy": 0.23401091396808624,
      "num_tokens": 70138768.0,
      "step": 31405
    },
    {
      "entropy": 5.58281307220459,
      "epoch": 3.527657662716909,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038000983183304765,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22047585248947144,
      "num_tokens": 70148411.0,
      "step": 31410
    },
    {
      "entropy": 5.5222841739654545,
      "epoch": 3.5282192396248666,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003799743222240572,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22821857929229736,
      "num_tokens": 70159788.0,
      "step": 31415
    },
    {
      "entropy": 5.609221029281616,
      "epoch": 3.5287808165328243,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003799388092726436,
      "loss": 4.8848,
      "mean_token_accuracy": 0.22796582728624343,
      "num_tokens": 70169707.0,
      "step": 31420
    },
    {
      "entropy": 5.585494756698608,
      "epoch": 3.529342393440782,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037990329297993763,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22530307471752167,
      "num_tokens": 70180958.0,
      "step": 31425
    },
    {
      "entropy": 5.543320894241333,
      "epoch": 3.529903970348739,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003798677733470701,
      "loss": 4.913,
      "mean_token_accuracy": 0.2211833715438843,
      "num_tokens": 70192848.0,
      "step": 31430
    },
    {
      "entropy": 5.556037187576294,
      "epoch": 3.530465547256697,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037983225037517207,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22758328765630723,
      "num_tokens": 70204034.0,
      "step": 31435
    },
    {
      "entropy": 5.689043045043945,
      "epoch": 3.5310271241646545,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037979672406537445,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22452564686536788,
      "num_tokens": 70214274.0,
      "step": 31440
    },
    {
      "entropy": 5.676802301406861,
      "epoch": 3.5315887010726117,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003797611944188086,
      "loss": 4.9711,
      "mean_token_accuracy": 0.22418046295642852,
      "num_tokens": 70223890.0,
      "step": 31445
    },
    {
      "entropy": 5.653112363815308,
      "epoch": 3.5321502779805694,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037972566143660566,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21666655093431472,
      "num_tokens": 70234529.0,
      "step": 31450
    },
    {
      "entropy": 5.528748607635498,
      "epoch": 3.532711854888527,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003796901251198969,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22886830866336821,
      "num_tokens": 70246900.0,
      "step": 31455
    },
    {
      "entropy": 5.690800189971924,
      "epoch": 3.5332734317964847,
      "grad_norm": 1.140625,
      "learning_rate": 0.000379654585469814,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2155233785510063,
      "num_tokens": 70257942.0,
      "step": 31460
    },
    {
      "entropy": 5.6803053855896,
      "epoch": 3.5338350087044423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003796190424874886,
      "loss": 5.0318,
      "mean_token_accuracy": 0.22396439760923387,
      "num_tokens": 70270297.0,
      "step": 31465
    },
    {
      "entropy": 5.678803396224976,
      "epoch": 3.5343965856123996,
      "grad_norm": 1.125,
      "learning_rate": 0.0003795834961740523,
      "loss": 5.0419,
      "mean_token_accuracy": 0.22263674139976503,
      "num_tokens": 70281542.0,
      "step": 31470
    },
    {
      "entropy": 5.622856330871582,
      "epoch": 3.534958162520357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037954794653063695,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22027012556791306,
      "num_tokens": 70290957.0,
      "step": 31475
    },
    {
      "entropy": 5.640298557281494,
      "epoch": 3.535519739428315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003795123935583743,
      "loss": 4.9781,
      "mean_token_accuracy": 0.23542795330286026,
      "num_tokens": 70302796.0,
      "step": 31480
    },
    {
      "entropy": 5.589291620254516,
      "epoch": 3.536081316336272,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003794768372583966,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2370254099369049,
      "num_tokens": 70313709.0,
      "step": 31485
    },
    {
      "entropy": 5.617755556106568,
      "epoch": 3.5366428932442298,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037944127763183587,
      "loss": 5.0055,
      "mean_token_accuracy": 0.22545716613531114,
      "num_tokens": 70324651.0,
      "step": 31490
    },
    {
      "entropy": 5.698909902572632,
      "epoch": 3.5372044701521874,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037940571467982436,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21290592551231385,
      "num_tokens": 70335810.0,
      "step": 31495
    },
    {
      "entropy": 5.589028263092041,
      "epoch": 3.5377660470601446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003793701484034944,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2304742231965065,
      "num_tokens": 70347171.0,
      "step": 31500
    },
    {
      "entropy": 5.578693675994873,
      "epoch": 3.5383276239681023,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037933457880397843,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2342184841632843,
      "num_tokens": 70359122.0,
      "step": 31505
    },
    {
      "entropy": 5.475494432449341,
      "epoch": 3.53888920087606,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037929900588240897,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2317228451371193,
      "num_tokens": 70369691.0,
      "step": 31510
    },
    {
      "entropy": 5.610324335098267,
      "epoch": 3.5394507777840176,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003792634296399187,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22314298450946807,
      "num_tokens": 70379602.0,
      "step": 31515
    },
    {
      "entropy": 5.638565969467163,
      "epoch": 3.5400123546919753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037922785007764043,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21807505339384078,
      "num_tokens": 70391819.0,
      "step": 31520
    },
    {
      "entropy": 5.703551292419434,
      "epoch": 3.5405739315999325,
      "grad_norm": 1.140625,
      "learning_rate": 0.000379192267196707,
      "loss": 5.0166,
      "mean_token_accuracy": 0.22423942685127257,
      "num_tokens": 70402078.0,
      "step": 31525
    },
    {
      "entropy": 5.567209196090698,
      "epoch": 3.54113550850789,
      "grad_norm": 1.0,
      "learning_rate": 0.00037915668099825135,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22463220953941346,
      "num_tokens": 70413829.0,
      "step": 31530
    },
    {
      "entropy": 5.465787315368653,
      "epoch": 3.541697085415848,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037912109148340663,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22249214947223664,
      "num_tokens": 70426221.0,
      "step": 31535
    },
    {
      "entropy": 5.6704014301300045,
      "epoch": 3.542258662323805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037908549865330596,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21451297104358674,
      "num_tokens": 70437215.0,
      "step": 31540
    },
    {
      "entropy": 5.646689176559448,
      "epoch": 3.5428202392317627,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037904990250908256,
      "loss": 4.8839,
      "mean_token_accuracy": 0.230146387219429,
      "num_tokens": 70448918.0,
      "step": 31545
    },
    {
      "entropy": 5.621914482116699,
      "epoch": 3.5433818161397204,
      "grad_norm": 1.125,
      "learning_rate": 0.00037901430305187,
      "loss": 4.9882,
      "mean_token_accuracy": 0.21666447520256044,
      "num_tokens": 70460713.0,
      "step": 31550
    },
    {
      "entropy": 5.565734004974365,
      "epoch": 3.543943393047678,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037897870028280167,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2232869505882263,
      "num_tokens": 70471460.0,
      "step": 31555
    },
    {
      "entropy": 5.596219110488891,
      "epoch": 3.5445049699556357,
      "grad_norm": 1.125,
      "learning_rate": 0.0003789430942030112,
      "loss": 4.9809,
      "mean_token_accuracy": 0.2226853847503662,
      "num_tokens": 70482359.0,
      "step": 31560
    },
    {
      "entropy": 5.670266771316529,
      "epoch": 3.545066546863593,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037890748481363216,
      "loss": 5.0284,
      "mean_token_accuracy": 0.22041308879852295,
      "num_tokens": 70493272.0,
      "step": 31565
    },
    {
      "entropy": 5.591851806640625,
      "epoch": 3.5456281237715506,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003788718721157987,
      "loss": 4.9504,
      "mean_token_accuracy": 0.2259292334318161,
      "num_tokens": 70504118.0,
      "step": 31570
    },
    {
      "entropy": 5.674164295196533,
      "epoch": 3.5461897006795082,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003788362561106445,
      "loss": 5.068,
      "mean_token_accuracy": 0.21998519003391265,
      "num_tokens": 70515379.0,
      "step": 31575
    },
    {
      "entropy": 5.635763216018677,
      "epoch": 3.5467512775874654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037880063679930355,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22388467490673064,
      "num_tokens": 70526047.0,
      "step": 31580
    },
    {
      "entropy": 5.655713129043579,
      "epoch": 3.547312854495423,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037876501418291015,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22829454839229585,
      "num_tokens": 70537614.0,
      "step": 31585
    },
    {
      "entropy": 5.626202964782715,
      "epoch": 3.5478744314033808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003787293882625984,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22851701080799103,
      "num_tokens": 70548048.0,
      "step": 31590
    },
    {
      "entropy": 5.568056440353393,
      "epoch": 3.548436008311338,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003786937590395027,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23282091468572616,
      "num_tokens": 70558668.0,
      "step": 31595
    },
    {
      "entropy": 5.589310646057129,
      "epoch": 3.5489975852192956,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003786581265147575,
      "loss": 5.0448,
      "mean_token_accuracy": 0.21595223844051362,
      "num_tokens": 70570044.0,
      "step": 31600
    },
    {
      "entropy": 5.615332078933716,
      "epoch": 3.5495591621272533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003786224906894974,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2333023190498352,
      "num_tokens": 70580857.0,
      "step": 31605
    },
    {
      "entropy": 5.537058258056641,
      "epoch": 3.550120739035211,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000378586851564857,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23336283266544341,
      "num_tokens": 70592476.0,
      "step": 31610
    },
    {
      "entropy": 5.584992837905884,
      "epoch": 3.5506823159431686,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000378551209141971,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22787859588861464,
      "num_tokens": 70603614.0,
      "step": 31615
    },
    {
      "entropy": 5.639754724502564,
      "epoch": 3.551243892851126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003785155634219744,
      "loss": 5.0047,
      "mean_token_accuracy": 0.21388435065746308,
      "num_tokens": 70615239.0,
      "step": 31620
    },
    {
      "entropy": 5.643446016311645,
      "epoch": 3.5518054697590835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003784799144060021,
      "loss": 4.9536,
      "mean_token_accuracy": 0.224198417365551,
      "num_tokens": 70625785.0,
      "step": 31625
    },
    {
      "entropy": 5.65679440498352,
      "epoch": 3.552367046667041,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037844426209518916,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22294149696826934,
      "num_tokens": 70636942.0,
      "step": 31630
    },
    {
      "entropy": 5.570975589752197,
      "epoch": 3.5529286235749984,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037840860649067084,
      "loss": 4.8401,
      "mean_token_accuracy": 0.22785678654909133,
      "num_tokens": 70648106.0,
      "step": 31635
    },
    {
      "entropy": 5.576294374465943,
      "epoch": 3.553490200482956,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037837294759358235,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22461603879928588,
      "num_tokens": 70658850.0,
      "step": 31640
    },
    {
      "entropy": 5.718483591079712,
      "epoch": 3.5540517773909137,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037833728540505914,
      "loss": 5.0265,
      "mean_token_accuracy": 0.217880417406559,
      "num_tokens": 70671060.0,
      "step": 31645
    },
    {
      "entropy": 5.56511607170105,
      "epoch": 3.5546133542988714,
      "grad_norm": 1.125,
      "learning_rate": 0.0003783016199262367,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23147940188646315,
      "num_tokens": 70682186.0,
      "step": 31650
    },
    {
      "entropy": 5.576669406890869,
      "epoch": 3.555174931206829,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003782659511582505,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22736668288707734,
      "num_tokens": 70692622.0,
      "step": 31655
    },
    {
      "entropy": 5.5418150424957275,
      "epoch": 3.5557365081147863,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003782302791022365,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23182711452245713,
      "num_tokens": 70703293.0,
      "step": 31660
    },
    {
      "entropy": 5.550937080383301,
      "epoch": 3.556298085022744,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037819460375933034,
      "loss": 4.8516,
      "mean_token_accuracy": 0.23216301202774048,
      "num_tokens": 70714961.0,
      "step": 31665
    },
    {
      "entropy": 5.6709370613098145,
      "epoch": 3.5568596619307016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003781589251306679,
      "loss": 5.025,
      "mean_token_accuracy": 0.2204990193247795,
      "num_tokens": 70727030.0,
      "step": 31670
    },
    {
      "entropy": 5.642950010299683,
      "epoch": 3.557421238838659,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003781232432173853,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22451940327882766,
      "num_tokens": 70738331.0,
      "step": 31675
    },
    {
      "entropy": 5.621253585815429,
      "epoch": 3.5579828157466165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037808755802061874,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21528009474277496,
      "num_tokens": 70748347.0,
      "step": 31680
    },
    {
      "entropy": 5.618359041213989,
      "epoch": 3.558544392654574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003780518695415043,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22806454449892044,
      "num_tokens": 70759905.0,
      "step": 31685
    },
    {
      "entropy": 5.713841676712036,
      "epoch": 3.5591059695625313,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037801617778117835,
      "loss": 5.0319,
      "mean_token_accuracy": 0.22582800984382628,
      "num_tokens": 70770538.0,
      "step": 31690
    },
    {
      "entropy": 5.631787824630737,
      "epoch": 3.559667546470489,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037798048274077735,
      "loss": 5.0276,
      "mean_token_accuracy": 0.22458020001649856,
      "num_tokens": 70781881.0,
      "step": 31695
    },
    {
      "entropy": 5.528191328048706,
      "epoch": 3.5602291233784467,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037794478442143786,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2306903824210167,
      "num_tokens": 70792299.0,
      "step": 31700
    },
    {
      "entropy": 5.619390058517456,
      "epoch": 3.5607907002864043,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037790908282429647,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22749179601669312,
      "num_tokens": 70802470.0,
      "step": 31705
    },
    {
      "entropy": 5.643369197845459,
      "epoch": 3.561352277194362,
      "grad_norm": 1.125,
      "learning_rate": 0.00037787337795049,
      "loss": 5.0398,
      "mean_token_accuracy": 0.22824373990297317,
      "num_tokens": 70813502.0,
      "step": 31710
    },
    {
      "entropy": 5.595929718017578,
      "epoch": 3.561913854102319,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037783766980115525,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22956487834453582,
      "num_tokens": 70823917.0,
      "step": 31715
    },
    {
      "entropy": 5.61225209236145,
      "epoch": 3.562475431010277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037780195837742927,
      "loss": 5.0504,
      "mean_token_accuracy": 0.21776285469532014,
      "num_tokens": 70834503.0,
      "step": 31720
    },
    {
      "entropy": 5.627058458328247,
      "epoch": 3.5630370079182345,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037776624368044895,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2287063330411911,
      "num_tokens": 70844954.0,
      "step": 31725
    },
    {
      "entropy": 5.695714521408081,
      "epoch": 3.5635985848261917,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037773052571135157,
      "loss": 5.0722,
      "mean_token_accuracy": 0.21557333171367646,
      "num_tokens": 70856142.0,
      "step": 31730
    },
    {
      "entropy": 5.635737800598145,
      "epoch": 3.5641601617341494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003776948044712745,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2276559516787529,
      "num_tokens": 70866880.0,
      "step": 31735
    },
    {
      "entropy": 5.59807596206665,
      "epoch": 3.564721738642107,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000377659079961355,
      "loss": 4.906,
      "mean_token_accuracy": 0.23266281187534332,
      "num_tokens": 70876836.0,
      "step": 31740
    },
    {
      "entropy": 5.580051469802856,
      "epoch": 3.5652833155500647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037762335218273057,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22543539851903915,
      "num_tokens": 70889430.0,
      "step": 31745
    },
    {
      "entropy": 5.572664785385132,
      "epoch": 3.5658448924580224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003775876211365388,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22964335978031158,
      "num_tokens": 70899291.0,
      "step": 31750
    },
    {
      "entropy": 5.652202415466308,
      "epoch": 3.5664064693659796,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003775518868239174,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22714188247919082,
      "num_tokens": 70910596.0,
      "step": 31755
    },
    {
      "entropy": 5.659063243865967,
      "epoch": 3.5669680462739373,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037751614924600417,
      "loss": 4.973,
      "mean_token_accuracy": 0.22753164023160935,
      "num_tokens": 70921969.0,
      "step": 31760
    },
    {
      "entropy": 5.596659183502197,
      "epoch": 3.567529623181895,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000377480408403937,
      "loss": 4.9243,
      "mean_token_accuracy": 0.2244533807039261,
      "num_tokens": 70933372.0,
      "step": 31765
    },
    {
      "entropy": 5.705514526367187,
      "epoch": 3.568091200089852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003774446642988538,
      "loss": 5.0771,
      "mean_token_accuracy": 0.2196504145860672,
      "num_tokens": 70944989.0,
      "step": 31770
    },
    {
      "entropy": 5.617773962020874,
      "epoch": 3.56865277699781,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037740891693189286,
      "loss": 5.0002,
      "mean_token_accuracy": 0.2203492760658264,
      "num_tokens": 70957516.0,
      "step": 31775
    },
    {
      "entropy": 5.703213310241699,
      "epoch": 3.5692143539057675,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003773731663041922,
      "loss": 5.0451,
      "mean_token_accuracy": 0.22140483260154725,
      "num_tokens": 70969485.0,
      "step": 31780
    },
    {
      "entropy": 5.630150651931762,
      "epoch": 3.5697759308137247,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037733741241689024,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22720604240894318,
      "num_tokens": 70980335.0,
      "step": 31785
    },
    {
      "entropy": 5.5520843982696535,
      "epoch": 3.5703375077216823,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003773016552711254,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22730220407247542,
      "num_tokens": 70991870.0,
      "step": 31790
    },
    {
      "entropy": 5.617423582077026,
      "epoch": 3.57089908462964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037726589486803617,
      "loss": 5.0666,
      "mean_token_accuracy": 0.21999655067920684,
      "num_tokens": 71003906.0,
      "step": 31795
    },
    {
      "entropy": 5.578919506072998,
      "epoch": 3.5714606615375977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037723013120876113,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2237373024225235,
      "num_tokens": 71015280.0,
      "step": 31800
    },
    {
      "entropy": 5.64251217842102,
      "epoch": 3.5720222384455553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037719436429443915,
      "loss": 4.9881,
      "mean_token_accuracy": 0.2249878540635109,
      "num_tokens": 71026299.0,
      "step": 31805
    },
    {
      "entropy": 5.704422330856323,
      "epoch": 3.5725838153535125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037715859412620895,
      "loss": 5.044,
      "mean_token_accuracy": 0.22291786819696427,
      "num_tokens": 71037704.0,
      "step": 31810
    },
    {
      "entropy": 5.647889089584351,
      "epoch": 3.57314539226147,
      "grad_norm": 1.125,
      "learning_rate": 0.0003771228207052095,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2314022809267044,
      "num_tokens": 71048359.0,
      "step": 31815
    },
    {
      "entropy": 5.584072542190552,
      "epoch": 3.573706969169428,
      "grad_norm": 1.125,
      "learning_rate": 0.00037708704403257974,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22978124618530274,
      "num_tokens": 71059723.0,
      "step": 31820
    },
    {
      "entropy": 5.509850406646729,
      "epoch": 3.574268546077385,
      "grad_norm": 1.078125,
      "learning_rate": 0.000377051264109459,
      "loss": 4.8672,
      "mean_token_accuracy": 0.233322311937809,
      "num_tokens": 71071437.0,
      "step": 31825
    },
    {
      "entropy": 5.567746925354004,
      "epoch": 3.5748301229853428,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003770154809369864,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2372191444039345,
      "num_tokens": 71082544.0,
      "step": 31830
    },
    {
      "entropy": 5.613438653945923,
      "epoch": 3.5753916998933004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037697969451630135,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22692109346389772,
      "num_tokens": 71092429.0,
      "step": 31835
    },
    {
      "entropy": 5.56281008720398,
      "epoch": 3.575953276801258,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037694390484854323,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22255572825670242,
      "num_tokens": 71103154.0,
      "step": 31840
    },
    {
      "entropy": 5.569131565093994,
      "epoch": 3.5765148537092157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003769081119348517,
      "loss": 4.8954,
      "mean_token_accuracy": 0.23493303954601288,
      "num_tokens": 71112921.0,
      "step": 31845
    },
    {
      "entropy": 5.6330688953399655,
      "epoch": 3.577076430617173,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037687231577636626,
      "loss": 4.9613,
      "mean_token_accuracy": 0.23138314932584764,
      "num_tokens": 71123281.0,
      "step": 31850
    },
    {
      "entropy": 5.674470281600952,
      "epoch": 3.5776380075251306,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003768365163742269,
      "loss": 5.0763,
      "mean_token_accuracy": 0.21671097576618195,
      "num_tokens": 71134358.0,
      "step": 31855
    },
    {
      "entropy": 5.654243516921997,
      "epoch": 3.5781995844330883,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037680071372957327,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22346639186143874,
      "num_tokens": 71146805.0,
      "step": 31860
    },
    {
      "entropy": 5.637287664413452,
      "epoch": 3.5787611613410455,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037676490784354543,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22315446883440018,
      "num_tokens": 71157388.0,
      "step": 31865
    },
    {
      "entropy": 5.500623989105224,
      "epoch": 3.579322738249003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003767290987172835,
      "loss": 4.7801,
      "mean_token_accuracy": 0.23470230996608735,
      "num_tokens": 71168601.0,
      "step": 31870
    },
    {
      "entropy": 5.596658992767334,
      "epoch": 3.579884315156961,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037669328635192757,
      "loss": 4.9694,
      "mean_token_accuracy": 0.22655526399612427,
      "num_tokens": 71180796.0,
      "step": 31875
    },
    {
      "entropy": 5.596127080917358,
      "epoch": 3.580445892064918,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000376657470748618,
      "loss": 4.9152,
      "mean_token_accuracy": 0.23616715520620346,
      "num_tokens": 71191686.0,
      "step": 31880
    },
    {
      "entropy": 5.546095371246338,
      "epoch": 3.5810074689728757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003766216519084952,
      "loss": 4.7887,
      "mean_token_accuracy": 0.23263462632894516,
      "num_tokens": 71202454.0,
      "step": 31885
    },
    {
      "entropy": 5.62254056930542,
      "epoch": 3.5815690458808334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003765858298326994,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23478156924247742,
      "num_tokens": 71213729.0,
      "step": 31890
    },
    {
      "entropy": 5.740342664718628,
      "epoch": 3.582130622788791,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037655000452237154,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2177546963095665,
      "num_tokens": 71226810.0,
      "step": 31895
    },
    {
      "entropy": 5.56725378036499,
      "epoch": 3.5826921996967487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003765141759786521,
      "loss": 4.8807,
      "mean_token_accuracy": 0.23608027547597885,
      "num_tokens": 71237233.0,
      "step": 31900
    },
    {
      "entropy": 5.61645770072937,
      "epoch": 3.583253776604706,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037647834420268195,
      "loss": 4.9353,
      "mean_token_accuracy": 0.23654620349407196,
      "num_tokens": 71248678.0,
      "step": 31905
    },
    {
      "entropy": 5.530041933059692,
      "epoch": 3.5838153535126636,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003764425091956019,
      "loss": 4.8857,
      "mean_token_accuracy": 0.23413211852312088,
      "num_tokens": 71259098.0,
      "step": 31910
    },
    {
      "entropy": 5.567439985275269,
      "epoch": 3.584376930420621,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037640667095855307,
      "loss": 4.8797,
      "mean_token_accuracy": 0.23361638486385344,
      "num_tokens": 71270082.0,
      "step": 31915
    },
    {
      "entropy": 5.645912027359008,
      "epoch": 3.5849385073285784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037637082949267654,
      "loss": 4.9674,
      "mean_token_accuracy": 0.2330487459897995,
      "num_tokens": 71280975.0,
      "step": 31920
    },
    {
      "entropy": 5.600361156463623,
      "epoch": 3.585500084236536,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003763349847991134,
      "loss": 4.9472,
      "mean_token_accuracy": 0.23377084732055664,
      "num_tokens": 71292552.0,
      "step": 31925
    },
    {
      "entropy": 5.644791221618652,
      "epoch": 3.5860616611444938,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000376299136879005,
      "loss": 5.0535,
      "mean_token_accuracy": 0.2157470241189003,
      "num_tokens": 71304653.0,
      "step": 31930
    },
    {
      "entropy": 5.598940706253051,
      "epoch": 3.5866232380524514,
      "grad_norm": 1.125,
      "learning_rate": 0.00037626328573349296,
      "loss": 4.9867,
      "mean_token_accuracy": 0.22070961445569992,
      "num_tokens": 71316734.0,
      "step": 31935
    },
    {
      "entropy": 5.735214185714722,
      "epoch": 3.587184814960409,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003762274313637185,
      "loss": 5.1146,
      "mean_token_accuracy": 0.21587740778923034,
      "num_tokens": 71328228.0,
      "step": 31940
    },
    {
      "entropy": 5.606037521362305,
      "epoch": 3.5877463918683663,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003761915737708234,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2248762995004654,
      "num_tokens": 71339921.0,
      "step": 31945
    },
    {
      "entropy": 5.621533584594727,
      "epoch": 3.588307968776324,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037615571295594936,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22260644882917405,
      "num_tokens": 71350995.0,
      "step": 31950
    },
    {
      "entropy": 5.550945234298706,
      "epoch": 3.5888695456842816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003761198489202382,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22220165580511092,
      "num_tokens": 71362546.0,
      "step": 31955
    },
    {
      "entropy": 5.648637533187866,
      "epoch": 3.589431122592239,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003760839816648317,
      "loss": 5.0396,
      "mean_token_accuracy": 0.22617257982492447,
      "num_tokens": 71373431.0,
      "step": 31960
    },
    {
      "entropy": 5.675836753845215,
      "epoch": 3.5899926995001965,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037604811119087213,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2326497569680214,
      "num_tokens": 71384488.0,
      "step": 31965
    },
    {
      "entropy": 5.639913558959961,
      "epoch": 3.590554276408154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003760122374995015,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2231338232755661,
      "num_tokens": 71396558.0,
      "step": 31970
    },
    {
      "entropy": 5.538418531417847,
      "epoch": 3.5911158533161114,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003759763605918621,
      "loss": 4.8854,
      "mean_token_accuracy": 0.2315576568245888,
      "num_tokens": 71407107.0,
      "step": 31975
    },
    {
      "entropy": 5.678438806533814,
      "epoch": 3.591677430224069,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037594048046909616,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2238185554742813,
      "num_tokens": 71419801.0,
      "step": 31980
    },
    {
      "entropy": 5.721470546722412,
      "epoch": 3.5922390071320267,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003759045971323462,
      "loss": 5.0986,
      "mean_token_accuracy": 0.21571807861328124,
      "num_tokens": 71431740.0,
      "step": 31985
    },
    {
      "entropy": 5.620782136917114,
      "epoch": 3.5928005840399844,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037586871058275476,
      "loss": 4.9742,
      "mean_token_accuracy": 0.22530414313077926,
      "num_tokens": 71443514.0,
      "step": 31990
    },
    {
      "entropy": 5.613966464996338,
      "epoch": 3.593362160947942,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037583282082146434,
      "loss": 4.968,
      "mean_token_accuracy": 0.22724979221820832,
      "num_tokens": 71454240.0,
      "step": 31995
    },
    {
      "entropy": 5.595543670654297,
      "epoch": 3.5939237378558992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037579692784961785,
      "loss": 4.917,
      "mean_token_accuracy": 0.22649177610874177,
      "num_tokens": 71464707.0,
      "step": 32000
    },
    {
      "entropy": 5.658187007904052,
      "epoch": 3.594485314763857,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003757610316683581,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21893186569213868,
      "num_tokens": 71475931.0,
      "step": 32005
    },
    {
      "entropy": 5.657009506225586,
      "epoch": 3.5950468916718146,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037572513227882807,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2200138196349144,
      "num_tokens": 71487851.0,
      "step": 32010
    },
    {
      "entropy": 5.658991765975952,
      "epoch": 3.595608468579772,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037568922968217065,
      "loss": 4.9995,
      "mean_token_accuracy": 0.22542385905981063,
      "num_tokens": 71498144.0,
      "step": 32015
    },
    {
      "entropy": 5.618118810653686,
      "epoch": 3.5961700454877294,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037565332387952914,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22366383224725722,
      "num_tokens": 71508074.0,
      "step": 32020
    },
    {
      "entropy": 5.6122602939605715,
      "epoch": 3.596731622395687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037561741487204684,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22750355005264283,
      "num_tokens": 71519609.0,
      "step": 32025
    },
    {
      "entropy": 5.565634822845459,
      "epoch": 3.5972931993036448,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003755815026608669,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23492198437452316,
      "num_tokens": 71530256.0,
      "step": 32030
    },
    {
      "entropy": 5.705154514312744,
      "epoch": 3.5978547762116024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037554558724713305,
      "loss": 4.9738,
      "mean_token_accuracy": 0.22052489072084427,
      "num_tokens": 71540383.0,
      "step": 32035
    },
    {
      "entropy": 5.646464681625366,
      "epoch": 3.5984163531195597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003755096686319886,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2246050640940666,
      "num_tokens": 71551739.0,
      "step": 32040
    },
    {
      "entropy": 5.640783643722534,
      "epoch": 3.5989779300275173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037547374681657736,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22811383306980132,
      "num_tokens": 71562429.0,
      "step": 32045
    },
    {
      "entropy": 5.684541511535644,
      "epoch": 3.599539506935475,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037543782180204303,
      "loss": 5.0089,
      "mean_token_accuracy": 0.22057738602161409,
      "num_tokens": 71573179.0,
      "step": 32050
    },
    {
      "entropy": 5.635150527954101,
      "epoch": 3.600101083843432,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037540189358952955,
      "loss": 5.0115,
      "mean_token_accuracy": 0.22204408198595046,
      "num_tokens": 71584223.0,
      "step": 32055
    },
    {
      "entropy": 5.598528528213501,
      "epoch": 3.60066266075139,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003753659621801808,
      "loss": 4.9683,
      "mean_token_accuracy": 0.2299070402979851,
      "num_tokens": 71595166.0,
      "step": 32060
    },
    {
      "entropy": 5.626034355163574,
      "epoch": 3.6012242376593475,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037533002757514095,
      "loss": 5.005,
      "mean_token_accuracy": 0.22104533910751342,
      "num_tokens": 71606290.0,
      "step": 32065
    },
    {
      "entropy": 5.622946643829346,
      "epoch": 3.6017858145673047,
      "grad_norm": 1.171875,
      "learning_rate": 0.000375294089775554,
      "loss": 4.9842,
      "mean_token_accuracy": 0.22561465501785277,
      "num_tokens": 71617545.0,
      "step": 32070
    },
    {
      "entropy": 5.62714114189148,
      "epoch": 3.6023473914752624,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003752581487825644,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22899961620569229,
      "num_tokens": 71627949.0,
      "step": 32075
    },
    {
      "entropy": 5.660893201828003,
      "epoch": 3.60290896838322,
      "grad_norm": 1.125,
      "learning_rate": 0.00037522220459731644,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2280524656176567,
      "num_tokens": 71639017.0,
      "step": 32080
    },
    {
      "entropy": 5.569404602050781,
      "epoch": 3.6034705452911777,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003751862572209546,
      "loss": 4.8405,
      "mean_token_accuracy": 0.2353963002562523,
      "num_tokens": 71649189.0,
      "step": 32085
    },
    {
      "entropy": 5.551142692565918,
      "epoch": 3.6040321221991354,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037515030665462345,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2317319169640541,
      "num_tokens": 71660083.0,
      "step": 32090
    },
    {
      "entropy": 5.549918556213379,
      "epoch": 3.6045936991070926,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037511435289946775,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2297135129570961,
      "num_tokens": 71670874.0,
      "step": 32095
    },
    {
      "entropy": 5.669686222076416,
      "epoch": 3.6051552760150503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003750783959566322,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2234621524810791,
      "num_tokens": 71681479.0,
      "step": 32100
    },
    {
      "entropy": 5.632257413864136,
      "epoch": 3.605716852923008,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003750424358272617,
      "loss": 4.9536,
      "mean_token_accuracy": 0.2289967119693756,
      "num_tokens": 71693865.0,
      "step": 32105
    },
    {
      "entropy": 5.588266324996948,
      "epoch": 3.606278429830965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003750064725125013,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22741254717111586,
      "num_tokens": 71704513.0,
      "step": 32110
    },
    {
      "entropy": 5.543050861358642,
      "epoch": 3.606840006738923,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037497050601349605,
      "loss": 4.96,
      "mean_token_accuracy": 0.223631651699543,
      "num_tokens": 71715444.0,
      "step": 32115
    },
    {
      "entropy": 5.615339040756226,
      "epoch": 3.6074015836468805,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037493453633139103,
      "loss": 4.8859,
      "mean_token_accuracy": 0.22972326278686522,
      "num_tokens": 71725452.0,
      "step": 32120
    },
    {
      "entropy": 5.651366233825684,
      "epoch": 3.607963160554838,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037489856346733163,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22265108823776245,
      "num_tokens": 71736536.0,
      "step": 32125
    },
    {
      "entropy": 5.587708759307861,
      "epoch": 3.608524737462796,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003748625874224632,
      "loss": 4.9512,
      "mean_token_accuracy": 0.2274011805653572,
      "num_tokens": 71748420.0,
      "step": 32130
    },
    {
      "entropy": 5.5871296405792235,
      "epoch": 3.609086314370753,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037482660819793134,
      "loss": 4.9353,
      "mean_token_accuracy": 0.23183376342058182,
      "num_tokens": 71758779.0,
      "step": 32135
    },
    {
      "entropy": 5.735920000076294,
      "epoch": 3.6096478912787107,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003747906257948814,
      "loss": 5.12,
      "mean_token_accuracy": 0.2136550650000572,
      "num_tokens": 71771042.0,
      "step": 32140
    },
    {
      "entropy": 5.631336355209351,
      "epoch": 3.6102094681866683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003747546402144594,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22944415658712386,
      "num_tokens": 71782234.0,
      "step": 32145
    },
    {
      "entropy": 5.585093069076538,
      "epoch": 3.6107710450946255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037471865145781094,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23412672728300093,
      "num_tokens": 71793115.0,
      "step": 32150
    },
    {
      "entropy": 5.5366302013397215,
      "epoch": 3.611332622002583,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037468265952608196,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23714719414711,
      "num_tokens": 71804466.0,
      "step": 32155
    },
    {
      "entropy": 5.652343988418579,
      "epoch": 3.611894198910541,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003746466644204183,
      "loss": 5.0202,
      "mean_token_accuracy": 0.21680717319250106,
      "num_tokens": 71815248.0,
      "step": 32160
    },
    {
      "entropy": 5.689828777313233,
      "epoch": 3.612455775818498,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003746106661419663,
      "loss": 4.9991,
      "mean_token_accuracy": 0.2288440629839897,
      "num_tokens": 71825816.0,
      "step": 32165
    },
    {
      "entropy": 5.625261354446411,
      "epoch": 3.6130173527264557,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000374574664691872,
      "loss": 4.9682,
      "mean_token_accuracy": 0.22346384078264236,
      "num_tokens": 71837918.0,
      "step": 32170
    },
    {
      "entropy": 5.665679597854615,
      "epoch": 3.6135789296344134,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037453866007128176,
      "loss": 5.0384,
      "mean_token_accuracy": 0.22590884268283845,
      "num_tokens": 71848736.0,
      "step": 32175
    },
    {
      "entropy": 5.5440247535705565,
      "epoch": 3.614140506542371,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037450265228134203,
      "loss": 4.8771,
      "mean_token_accuracy": 0.23174545019865037,
      "num_tokens": 71859903.0,
      "step": 32180
    },
    {
      "entropy": 5.5968467712402346,
      "epoch": 3.6147020834503287,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037446664132319925,
      "loss": 4.9801,
      "mean_token_accuracy": 0.2193368062376976,
      "num_tokens": 71871656.0,
      "step": 32185
    },
    {
      "entropy": 5.66455340385437,
      "epoch": 3.615263660358286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000374430627198,
      "loss": 4.9632,
      "mean_token_accuracy": 0.2235743522644043,
      "num_tokens": 71883282.0,
      "step": 32190
    },
    {
      "entropy": 5.664215087890625,
      "epoch": 3.6158252372662436,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037439460990689106,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22470613121986388,
      "num_tokens": 71894473.0,
      "step": 32195
    },
    {
      "entropy": 5.634539461135864,
      "epoch": 3.6163868141742013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003743585894510192,
      "loss": 5.0689,
      "mean_token_accuracy": 0.21530965715646744,
      "num_tokens": 71905555.0,
      "step": 32200
    },
    {
      "entropy": 5.558196878433227,
      "epoch": 3.6169483910821585,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037432256583153127,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2196456879377365,
      "num_tokens": 71916599.0,
      "step": 32205
    },
    {
      "entropy": 5.664769983291626,
      "epoch": 3.617509967990116,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037428653904957433,
      "loss": 4.954,
      "mean_token_accuracy": 0.22054607272148133,
      "num_tokens": 71926173.0,
      "step": 32210
    },
    {
      "entropy": 5.663511133193969,
      "epoch": 3.618071544898074,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003742505091062955,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2253687635064125,
      "num_tokens": 71937389.0,
      "step": 32215
    },
    {
      "entropy": 5.625271940231324,
      "epoch": 3.6186331218060315,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003742144760028419,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22591810524463654,
      "num_tokens": 71949216.0,
      "step": 32220
    },
    {
      "entropy": 5.665780305862427,
      "epoch": 3.619194698713989,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000374178439740361,
      "loss": 5.0314,
      "mean_token_accuracy": 0.216696834564209,
      "num_tokens": 71959861.0,
      "step": 32225
    },
    {
      "entropy": 5.603806495666504,
      "epoch": 3.6197562756219464,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003741424003200001,
      "loss": 4.9984,
      "mean_token_accuracy": 0.2228460654616356,
      "num_tokens": 71970043.0,
      "step": 32230
    },
    {
      "entropy": 5.648821306228638,
      "epoch": 3.620317852529904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037410635774290667,
      "loss": 4.9416,
      "mean_token_accuracy": 0.2303069680929184,
      "num_tokens": 71981096.0,
      "step": 32235
    },
    {
      "entropy": 5.620360231399536,
      "epoch": 3.6208794294378617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003740703120102284,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22745125144720077,
      "num_tokens": 71992759.0,
      "step": 32240
    },
    {
      "entropy": 5.60250825881958,
      "epoch": 3.621441006345819,
      "grad_norm": 1.15625,
      "learning_rate": 0.000374034263123113,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2300302803516388,
      "num_tokens": 72003187.0,
      "step": 32245
    },
    {
      "entropy": 5.732370138168335,
      "epoch": 3.6220025832537766,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037399821108270823,
      "loss": 5.1073,
      "mean_token_accuracy": 0.21590884178876876,
      "num_tokens": 72013712.0,
      "step": 32250
    },
    {
      "entropy": 5.70269365310669,
      "epoch": 3.622564160161734,
      "grad_norm": 1.125,
      "learning_rate": 0.000373962155890162,
      "loss": 5.084,
      "mean_token_accuracy": 0.21226473450660704,
      "num_tokens": 72024693.0,
      "step": 32255
    },
    {
      "entropy": 5.624772405624389,
      "epoch": 3.6231257370696914,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003739260975466223,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2297005608677864,
      "num_tokens": 72036578.0,
      "step": 32260
    },
    {
      "entropy": 5.664010810852051,
      "epoch": 3.623687313977649,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037389003605323737,
      "loss": 5.0322,
      "mean_token_accuracy": 0.225683730840683,
      "num_tokens": 72047574.0,
      "step": 32265
    },
    {
      "entropy": 5.6457596778869625,
      "epoch": 3.6242488908856068,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037385397141115535,
      "loss": 4.954,
      "mean_token_accuracy": 0.2300664469599724,
      "num_tokens": 72059998.0,
      "step": 32270
    },
    {
      "entropy": 5.694514799118042,
      "epoch": 3.6248104677935644,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037381790362152453,
      "loss": 5.0564,
      "mean_token_accuracy": 0.2208966851234436,
      "num_tokens": 72070753.0,
      "step": 32275
    },
    {
      "entropy": 5.605670404434204,
      "epoch": 3.625372044701522,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037378183268549326,
      "loss": 4.9002,
      "mean_token_accuracy": 0.22993775606155395,
      "num_tokens": 72081718.0,
      "step": 32280
    },
    {
      "entropy": 5.613668155670166,
      "epoch": 3.6259336216094793,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003737457586042102,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22842640578746795,
      "num_tokens": 72092403.0,
      "step": 32285
    },
    {
      "entropy": 5.5910422801971436,
      "epoch": 3.626495198517437,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003737096813788238,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22817525863647461,
      "num_tokens": 72103711.0,
      "step": 32290
    },
    {
      "entropy": 5.693708610534668,
      "epoch": 3.6270567754253946,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003736736010104829,
      "loss": 5.025,
      "mean_token_accuracy": 0.2207014963030815,
      "num_tokens": 72115095.0,
      "step": 32295
    },
    {
      "entropy": 5.671240520477295,
      "epoch": 3.627618352333352,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003736375175003363,
      "loss": 5.0279,
      "mean_token_accuracy": 0.22200235724449158,
      "num_tokens": 72126559.0,
      "step": 32300
    },
    {
      "entropy": 5.633226776123047,
      "epoch": 3.6281799292413095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037360143084953285,
      "loss": 4.998,
      "mean_token_accuracy": 0.22451096773147583,
      "num_tokens": 72137752.0,
      "step": 32305
    },
    {
      "entropy": 5.55942850112915,
      "epoch": 3.628741506149267,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003735653410592217,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2267162710428238,
      "num_tokens": 72148888.0,
      "step": 32310
    },
    {
      "entropy": 5.606093025207519,
      "epoch": 3.629303083057225,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037352924813055185,
      "loss": 4.9551,
      "mean_token_accuracy": 0.23307398110628127,
      "num_tokens": 72159607.0,
      "step": 32315
    },
    {
      "entropy": 5.666454601287842,
      "epoch": 3.6298646599651825,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003734931520646724,
      "loss": 5.0034,
      "mean_token_accuracy": 0.22803719639778136,
      "num_tokens": 72171204.0,
      "step": 32320
    },
    {
      "entropy": 5.592747449874878,
      "epoch": 3.6304262368731397,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037345705286273285,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22418228089809417,
      "num_tokens": 72182125.0,
      "step": 32325
    },
    {
      "entropy": 5.631580448150634,
      "epoch": 3.6309878137810974,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003734209505258825,
      "loss": 4.965,
      "mean_token_accuracy": 0.22529291212558747,
      "num_tokens": 72193093.0,
      "step": 32330
    },
    {
      "entropy": 5.702210187911987,
      "epoch": 3.631549390689055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037338484505527103,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21815684586763381,
      "num_tokens": 72204235.0,
      "step": 32335
    },
    {
      "entropy": 5.61715784072876,
      "epoch": 3.6321109675970122,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037334873645204785,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22255922853946686,
      "num_tokens": 72215160.0,
      "step": 32340
    },
    {
      "entropy": 5.610084819793701,
      "epoch": 3.63267254450497,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037331262471736285,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22318883389234542,
      "num_tokens": 72227517.0,
      "step": 32345
    },
    {
      "entropy": 5.712872934341431,
      "epoch": 3.6332341214129276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037327650985236564,
      "loss": 5.0413,
      "mean_token_accuracy": 0.22069282978773117,
      "num_tokens": 72238616.0,
      "step": 32350
    },
    {
      "entropy": 5.6946300029754635,
      "epoch": 3.633795698320885,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003732403918582063,
      "loss": 4.9896,
      "mean_token_accuracy": 0.2252253919839859,
      "num_tokens": 72250252.0,
      "step": 32355
    },
    {
      "entropy": 5.694857120513916,
      "epoch": 3.6343572752288424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037320427073603475,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21867020130157472,
      "num_tokens": 72261197.0,
      "step": 32360
    },
    {
      "entropy": 5.555647563934326,
      "epoch": 3.6349188521368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037316814648700113,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23026449531316756,
      "num_tokens": 72273838.0,
      "step": 32365
    },
    {
      "entropy": 5.57886700630188,
      "epoch": 3.6354804290447578,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003731320191122557,
      "loss": 4.9309,
      "mean_token_accuracy": 0.23526440113782882,
      "num_tokens": 72285047.0,
      "step": 32370
    },
    {
      "entropy": 5.622797012329102,
      "epoch": 3.6360420059527154,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003730958886129486,
      "loss": 5.0301,
      "mean_token_accuracy": 0.22264543771743775,
      "num_tokens": 72295883.0,
      "step": 32375
    },
    {
      "entropy": 5.696510696411133,
      "epoch": 3.6366035828606726,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037305975499023046,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22744131535291673,
      "num_tokens": 72307586.0,
      "step": 32380
    },
    {
      "entropy": 5.600030899047852,
      "epoch": 3.6371651597686303,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003730236182452517,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22583484947681426,
      "num_tokens": 72318501.0,
      "step": 32385
    },
    {
      "entropy": 5.6231354713439945,
      "epoch": 3.637726736676588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003729874783791629,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22714998871088027,
      "num_tokens": 72330178.0,
      "step": 32390
    },
    {
      "entropy": 5.6730725288391115,
      "epoch": 3.638288313584545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003729513353931148,
      "loss": 5.0557,
      "mean_token_accuracy": 0.221679425239563,
      "num_tokens": 72340304.0,
      "step": 32395
    },
    {
      "entropy": 5.67783408164978,
      "epoch": 3.638849890492503,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003729151892882581,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22930595874786378,
      "num_tokens": 72350353.0,
      "step": 32400
    },
    {
      "entropy": 5.620272159576416,
      "epoch": 3.6394114674004605,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003728790400657439,
      "loss": 4.913,
      "mean_token_accuracy": 0.2360897719860077,
      "num_tokens": 72361293.0,
      "step": 32405
    },
    {
      "entropy": 5.58805251121521,
      "epoch": 3.639973044308418,
      "grad_norm": 1.25,
      "learning_rate": 0.00037284288772672315,
      "loss": 4.9271,
      "mean_token_accuracy": 0.23252396732568742,
      "num_tokens": 72372013.0,
      "step": 32410
    },
    {
      "entropy": 5.600398588180542,
      "epoch": 3.640534621216376,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037280673227234687,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2202797070145607,
      "num_tokens": 72382611.0,
      "step": 32415
    },
    {
      "entropy": 5.620935249328613,
      "epoch": 3.641096198124333,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003727705737037662,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22879421412944795,
      "num_tokens": 72393701.0,
      "step": 32420
    },
    {
      "entropy": 5.750066089630127,
      "epoch": 3.6416577750322907,
      "grad_norm": 1.25,
      "learning_rate": 0.0003727344120221326,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21626118421554566,
      "num_tokens": 72404168.0,
      "step": 32425
    },
    {
      "entropy": 5.676728868484497,
      "epoch": 3.6422193519402484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037269824722859743,
      "loss": 4.958,
      "mean_token_accuracy": 0.22739725261926652,
      "num_tokens": 72415184.0,
      "step": 32430
    },
    {
      "entropy": 5.618797636032104,
      "epoch": 3.6427809288482056,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003726620793243121,
      "loss": 4.9539,
      "mean_token_accuracy": 0.22523107528686523,
      "num_tokens": 72426189.0,
      "step": 32435
    },
    {
      "entropy": 5.645012092590332,
      "epoch": 3.6433425057561633,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003726259083104284,
      "loss": 5.0021,
      "mean_token_accuracy": 0.22670815885066986,
      "num_tokens": 72436741.0,
      "step": 32440
    },
    {
      "entropy": 5.688428783416748,
      "epoch": 3.643904082664121,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037258973418809784,
      "loss": 4.9872,
      "mean_token_accuracy": 0.22785927802324296,
      "num_tokens": 72448610.0,
      "step": 32445
    },
    {
      "entropy": 5.5868669033050535,
      "epoch": 3.644465659572078,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003725535569584723,
      "loss": 4.9646,
      "mean_token_accuracy": 0.225261789560318,
      "num_tokens": 72460789.0,
      "step": 32450
    },
    {
      "entropy": 5.594180679321289,
      "epoch": 3.645027236480036,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037251737662270367,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22226393222808838,
      "num_tokens": 72471065.0,
      "step": 32455
    },
    {
      "entropy": 5.628716516494751,
      "epoch": 3.6455888133879935,
      "grad_norm": 1.171875,
      "learning_rate": 0.000372481193181944,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22842729091644287,
      "num_tokens": 72481788.0,
      "step": 32460
    },
    {
      "entropy": 5.645275259017945,
      "epoch": 3.646150390295951,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037244500663734524,
      "loss": 4.9125,
      "mean_token_accuracy": 0.23052416741847992,
      "num_tokens": 72493840.0,
      "step": 32465
    },
    {
      "entropy": 5.591303825378418,
      "epoch": 3.6467119672039088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003724088169900597,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2231943726539612,
      "num_tokens": 72505002.0,
      "step": 32470
    },
    {
      "entropy": 5.531902933120728,
      "epoch": 3.647273544111866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037237262424123975,
      "loss": 4.867,
      "mean_token_accuracy": 0.2275007963180542,
      "num_tokens": 72515917.0,
      "step": 32475
    },
    {
      "entropy": 5.666153383255005,
      "epoch": 3.6478351210198237,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037233642839203763,
      "loss": 5.0,
      "mean_token_accuracy": 0.22394780814647675,
      "num_tokens": 72527500.0,
      "step": 32480
    },
    {
      "entropy": 5.5438805103302,
      "epoch": 3.6483966979277813,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003723002294436059,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23115599304437637,
      "num_tokens": 72538402.0,
      "step": 32485
    },
    {
      "entropy": 5.620054864883423,
      "epoch": 3.6489582748357385,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037226402739709704,
      "loss": 4.9735,
      "mean_token_accuracy": 0.219794924557209,
      "num_tokens": 72550260.0,
      "step": 32490
    },
    {
      "entropy": 5.72707896232605,
      "epoch": 3.649519851743696,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003722278222536639,
      "loss": 5.029,
      "mean_token_accuracy": 0.21957138627767564,
      "num_tokens": 72562335.0,
      "step": 32495
    },
    {
      "entropy": 5.620959758758545,
      "epoch": 3.650081428651654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037219161401445914,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22683589905500412,
      "num_tokens": 72572620.0,
      "step": 32500
    },
    {
      "entropy": 5.571025371551514,
      "epoch": 3.6506430055596115,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037215540268063573,
      "loss": 5.0221,
      "mean_token_accuracy": 0.22288970053195953,
      "num_tokens": 72584487.0,
      "step": 32505
    },
    {
      "entropy": 5.637506437301636,
      "epoch": 3.651204582467569,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003721191882533466,
      "loss": 5.0191,
      "mean_token_accuracy": 0.22073039412498474,
      "num_tokens": 72596191.0,
      "step": 32510
    },
    {
      "entropy": 5.689265966415405,
      "epoch": 3.6517661593755264,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003720829707337449,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21881598085165024,
      "num_tokens": 72605901.0,
      "step": 32515
    },
    {
      "entropy": 5.647314596176147,
      "epoch": 3.652327736283484,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003720467501229836,
      "loss": 4.9625,
      "mean_token_accuracy": 0.226349937915802,
      "num_tokens": 72617231.0,
      "step": 32520
    },
    {
      "entropy": 5.7026166915893555,
      "epoch": 3.6528893131914417,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003720105264222163,
      "loss": 5.0294,
      "mean_token_accuracy": 0.22198336720466613,
      "num_tokens": 72628777.0,
      "step": 32525
    },
    {
      "entropy": 5.636749315261841,
      "epoch": 3.653450890099399,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037197429963259615,
      "loss": 5.0036,
      "mean_token_accuracy": 0.2251940369606018,
      "num_tokens": 72639670.0,
      "step": 32530
    },
    {
      "entropy": 5.60249285697937,
      "epoch": 3.6540124670073566,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003719380697552767,
      "loss": 4.8557,
      "mean_token_accuracy": 0.23149244636297225,
      "num_tokens": 72651010.0,
      "step": 32535
    },
    {
      "entropy": 5.769377374649048,
      "epoch": 3.6545740439153143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003719018367914115,
      "loss": 5.0707,
      "mean_token_accuracy": 0.21734413206577302,
      "num_tokens": 72662900.0,
      "step": 32540
    },
    {
      "entropy": 5.659744739532471,
      "epoch": 3.6551356208232715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003718656007421542,
      "loss": 5.0601,
      "mean_token_accuracy": 0.2240633875131607,
      "num_tokens": 72674348.0,
      "step": 32545
    },
    {
      "entropy": 5.640695953369141,
      "epoch": 3.655697197731229,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037182936160865866,
      "loss": 4.8619,
      "mean_token_accuracy": 0.23180197328329086,
      "num_tokens": 72685771.0,
      "step": 32550
    },
    {
      "entropy": 5.6190235137939455,
      "epoch": 3.656258774639187,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003717931193920788,
      "loss": 4.8806,
      "mean_token_accuracy": 0.23211172372102737,
      "num_tokens": 72697121.0,
      "step": 32555
    },
    {
      "entropy": 5.576670122146607,
      "epoch": 3.6568203515471445,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003717568740935683,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2246488079428673,
      "num_tokens": 72707245.0,
      "step": 32560
    },
    {
      "entropy": 5.587843227386474,
      "epoch": 3.657381928455102,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003717206257142815,
      "loss": 4.8988,
      "mean_token_accuracy": 0.23345900177955628,
      "num_tokens": 72718328.0,
      "step": 32565
    },
    {
      "entropy": 5.710547733306885,
      "epoch": 3.6579435053630593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003716843742553724,
      "loss": 5.0476,
      "mean_token_accuracy": 0.22006697058677674,
      "num_tokens": 72729192.0,
      "step": 32570
    },
    {
      "entropy": 5.616707611083984,
      "epoch": 3.658505082271017,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003716481197179953,
      "loss": 4.9893,
      "mean_token_accuracy": 0.22216222286224366,
      "num_tokens": 72740213.0,
      "step": 32575
    },
    {
      "entropy": 5.566905403137207,
      "epoch": 3.6590666591789747,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037161186210330464,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22942349016666413,
      "num_tokens": 72752477.0,
      "step": 32580
    },
    {
      "entropy": 5.675377607345581,
      "epoch": 3.659628236086932,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037157560141245485,
      "loss": 4.9309,
      "mean_token_accuracy": 0.23458940386772156,
      "num_tokens": 72763926.0,
      "step": 32585
    },
    {
      "entropy": 5.63490047454834,
      "epoch": 3.6601898129948895,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037153933764660034,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2243576854467392,
      "num_tokens": 72774922.0,
      "step": 32590
    },
    {
      "entropy": 5.603993225097656,
      "epoch": 3.660751389902847,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000371503070806896,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22568070143461227,
      "num_tokens": 72786362.0,
      "step": 32595
    },
    {
      "entropy": 5.613710117340088,
      "epoch": 3.661312966810805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037146680089449643,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22355927377939225,
      "num_tokens": 72798337.0,
      "step": 32600
    },
    {
      "entropy": 5.635991334915161,
      "epoch": 3.6618745437187625,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037143052791055654,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23198408484458924,
      "num_tokens": 72808923.0,
      "step": 32605
    },
    {
      "entropy": 5.579428625106812,
      "epoch": 3.6624361206267197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003713942518562311,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22751136720180512,
      "num_tokens": 72820148.0,
      "step": 32610
    },
    {
      "entropy": 5.612331676483154,
      "epoch": 3.6629976975346774,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003713579727326754,
      "loss": 4.9974,
      "mean_token_accuracy": 0.22342761904001235,
      "num_tokens": 72830412.0,
      "step": 32615
    },
    {
      "entropy": 5.612610101699829,
      "epoch": 3.663559274442635,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037132169054104446,
      "loss": 4.89,
      "mean_token_accuracy": 0.2322948917746544,
      "num_tokens": 72840857.0,
      "step": 32620
    },
    {
      "entropy": 5.586479425430298,
      "epoch": 3.6641208513505923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037128540528249355,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2258547991514206,
      "num_tokens": 72852269.0,
      "step": 32625
    },
    {
      "entropy": 5.633088827133179,
      "epoch": 3.66468242825855,
      "grad_norm": 1.09375,
      "learning_rate": 0.000371249116958178,
      "loss": 5.0072,
      "mean_token_accuracy": 0.22400410622358322,
      "num_tokens": 72863536.0,
      "step": 32630
    },
    {
      "entropy": 5.5760619163513185,
      "epoch": 3.6652440051665076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037121282556925326,
      "loss": 4.952,
      "mean_token_accuracy": 0.22775647789239883,
      "num_tokens": 72873835.0,
      "step": 32635
    },
    {
      "entropy": 5.668719005584717,
      "epoch": 3.665805582074465,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003711765311168748,
      "loss": 4.8768,
      "mean_token_accuracy": 0.22957945317029954,
      "num_tokens": 72883916.0,
      "step": 32640
    },
    {
      "entropy": 5.660086965560913,
      "epoch": 3.6663671589824225,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037114023360219824,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21641875207424163,
      "num_tokens": 72895928.0,
      "step": 32645
    },
    {
      "entropy": 5.580229759216309,
      "epoch": 3.66692873589038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037110393302637937,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2263745442032814,
      "num_tokens": 72908138.0,
      "step": 32650
    },
    {
      "entropy": 5.653624820709228,
      "epoch": 3.667490312798338,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037106762939057404,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22223068475723268,
      "num_tokens": 72918326.0,
      "step": 32655
    },
    {
      "entropy": 5.683623743057251,
      "epoch": 3.6680518897062955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003710313226959381,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2224065124988556,
      "num_tokens": 72929140.0,
      "step": 32660
    },
    {
      "entropy": 5.6565269947052,
      "epoch": 3.6686134666142527,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003709950129436276,
      "loss": 4.9856,
      "mean_token_accuracy": 0.23015641868114473,
      "num_tokens": 72941355.0,
      "step": 32665
    },
    {
      "entropy": 5.649455738067627,
      "epoch": 3.6691750435222104,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003709587001347987,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2281857028603554,
      "num_tokens": 72951841.0,
      "step": 32670
    },
    {
      "entropy": 5.535343360900879,
      "epoch": 3.669736620430168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037092238427060756,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2261653408408165,
      "num_tokens": 72963843.0,
      "step": 32675
    },
    {
      "entropy": 5.63853440284729,
      "epoch": 3.6702981973381252,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003708860653522104,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22783015966415404,
      "num_tokens": 72973709.0,
      "step": 32680
    },
    {
      "entropy": 5.575225830078125,
      "epoch": 3.670859774246083,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003708497433807639,
      "loss": 4.7868,
      "mean_token_accuracy": 0.23835694640874863,
      "num_tokens": 72984658.0,
      "step": 32685
    },
    {
      "entropy": 5.561942291259766,
      "epoch": 3.6714213511540406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037081341835742433,
      "loss": 5.0152,
      "mean_token_accuracy": 0.22239467054605483,
      "num_tokens": 72995745.0,
      "step": 32690
    },
    {
      "entropy": 5.5920392036437985,
      "epoch": 3.671982928061998,
      "grad_norm": 1.25,
      "learning_rate": 0.0003707770902833483,
      "loss": 5.0076,
      "mean_token_accuracy": 0.22170147746801377,
      "num_tokens": 73007619.0,
      "step": 32695
    },
    {
      "entropy": 5.61634783744812,
      "epoch": 3.672544504969956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003707407591596926,
      "loss": 4.8817,
      "mean_token_accuracy": 0.22970870584249498,
      "num_tokens": 73017779.0,
      "step": 32700
    },
    {
      "entropy": 5.676770973205566,
      "epoch": 3.673106081877913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000370704424987614,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22696149051189424,
      "num_tokens": 73028246.0,
      "step": 32705
    },
    {
      "entropy": 5.648423480987549,
      "epoch": 3.6736676587858708,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003706680877682694,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21824499517679213,
      "num_tokens": 73038896.0,
      "step": 32710
    },
    {
      "entropy": 5.679548978805542,
      "epoch": 3.6742292356938284,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037063174750281573,
      "loss": 5.043,
      "mean_token_accuracy": 0.21369494497776031,
      "num_tokens": 73050960.0,
      "step": 32715
    },
    {
      "entropy": 5.6898548126220705,
      "epoch": 3.6747908126017856,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003705954041924102,
      "loss": 4.9265,
      "mean_token_accuracy": 0.23419596999883652,
      "num_tokens": 73061537.0,
      "step": 32720
    },
    {
      "entropy": 5.668991327285767,
      "epoch": 3.6753523895097433,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037055905783821,
      "loss": 5.0622,
      "mean_token_accuracy": 0.2212167963385582,
      "num_tokens": 73072357.0,
      "step": 32725
    },
    {
      "entropy": 5.600331449508667,
      "epoch": 3.675913966417701,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037052270844137216,
      "loss": 4.8895,
      "mean_token_accuracy": 0.2337587684392929,
      "num_tokens": 73083527.0,
      "step": 32730
    },
    {
      "entropy": 5.587911081314087,
      "epoch": 3.676475543325658,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003704863560030543,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22716978043317795,
      "num_tokens": 73095068.0,
      "step": 32735
    },
    {
      "entropy": 5.545595502853393,
      "epoch": 3.677037120233616,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003704500005244139,
      "loss": 4.8578,
      "mean_token_accuracy": 0.22877768725156783,
      "num_tokens": 73106318.0,
      "step": 32740
    },
    {
      "entropy": 5.606224775314331,
      "epoch": 3.6775986971415735,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037041364200660835,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22624971866607665,
      "num_tokens": 73118433.0,
      "step": 32745
    },
    {
      "entropy": 5.649015283584594,
      "epoch": 3.678160274049531,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037037728045079553,
      "loss": 4.9824,
      "mean_token_accuracy": 0.217895570397377,
      "num_tokens": 73129887.0,
      "step": 32750
    },
    {
      "entropy": 5.57633867263794,
      "epoch": 3.678721850957489,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000370340915858133,
      "loss": 4.901,
      "mean_token_accuracy": 0.22723153084516526,
      "num_tokens": 73141529.0,
      "step": 32755
    },
    {
      "entropy": 5.607232570648193,
      "epoch": 3.679283427865446,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003703045482297788,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22469446808099747,
      "num_tokens": 73151317.0,
      "step": 32760
    },
    {
      "entropy": 5.5180588722229,
      "epoch": 3.6798450047734037,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003702681775668908,
      "loss": 4.8333,
      "mean_token_accuracy": 0.22966392040252687,
      "num_tokens": 73162382.0,
      "step": 32765
    },
    {
      "entropy": 5.659786748886108,
      "epoch": 3.6804065816813614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003702318038706272,
      "loss": 5.0447,
      "mean_token_accuracy": 0.2234057679772377,
      "num_tokens": 73173142.0,
      "step": 32770
    },
    {
      "entropy": 5.596737241744995,
      "epoch": 3.6809681585893186,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003701954271421459,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2256189301609993,
      "num_tokens": 73185074.0,
      "step": 32775
    },
    {
      "entropy": 5.693842077255249,
      "epoch": 3.6815297354972762,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037015904738260533,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2279267355799675,
      "num_tokens": 73195806.0,
      "step": 32780
    },
    {
      "entropy": 5.73114857673645,
      "epoch": 3.682091312405234,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003701226645931637,
      "loss": 5.1074,
      "mean_token_accuracy": 0.21142646968364714,
      "num_tokens": 73207213.0,
      "step": 32785
    },
    {
      "entropy": 5.6061944484710695,
      "epoch": 3.6826528893131916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003700862787749796,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22768742293119432,
      "num_tokens": 73219715.0,
      "step": 32790
    },
    {
      "entropy": 5.629995441436767,
      "epoch": 3.6832144662211492,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003700498899292115,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22897828072309495,
      "num_tokens": 73229868.0,
      "step": 32795
    },
    {
      "entropy": 5.640066289901734,
      "epoch": 3.6837760431291064,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003700134980570181,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22238052636384964,
      "num_tokens": 73240347.0,
      "step": 32800
    },
    {
      "entropy": 5.64123420715332,
      "epoch": 3.684337620037064,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003699771031595579,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2291148766875267,
      "num_tokens": 73250165.0,
      "step": 32805
    },
    {
      "entropy": 5.6947023391723635,
      "epoch": 3.6848991969450218,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036994070523799005,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22855819314718245,
      "num_tokens": 73261320.0,
      "step": 32810
    },
    {
      "entropy": 5.6389869213104244,
      "epoch": 3.685460773852979,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036990430429347334,
      "loss": 5.0245,
      "mean_token_accuracy": 0.2180490091443062,
      "num_tokens": 73272342.0,
      "step": 32815
    },
    {
      "entropy": 5.679329347610474,
      "epoch": 3.6860223507609367,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036986790032716673,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2282141327857971,
      "num_tokens": 73283790.0,
      "step": 32820
    },
    {
      "entropy": 5.579628801345825,
      "epoch": 3.6865839276688943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003698314933402292,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22824359238147734,
      "num_tokens": 73295869.0,
      "step": 32825
    },
    {
      "entropy": 5.548293113708496,
      "epoch": 3.6871455045768515,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003697950833338204,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22213429361581802,
      "num_tokens": 73307518.0,
      "step": 32830
    },
    {
      "entropy": 5.628569555282593,
      "epoch": 3.687707081484809,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036975867030909933,
      "loss": 4.9622,
      "mean_token_accuracy": 0.223396597802639,
      "num_tokens": 73318419.0,
      "step": 32835
    },
    {
      "entropy": 5.6460541725158695,
      "epoch": 3.688268658392767,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036972225426722545,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22315593957901,
      "num_tokens": 73328094.0,
      "step": 32840
    },
    {
      "entropy": 5.649577808380127,
      "epoch": 3.6888302353007245,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003696858352093582,
      "loss": 4.9419,
      "mean_token_accuracy": 0.2358602911233902,
      "num_tokens": 73339156.0,
      "step": 32845
    },
    {
      "entropy": 5.625084972381591,
      "epoch": 3.689391812208682,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036964941313665715,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23757439255714416,
      "num_tokens": 73350740.0,
      "step": 32850
    },
    {
      "entropy": 5.6820227146148685,
      "epoch": 3.6899533891166394,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003696129880502823,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22765185534954072,
      "num_tokens": 73360745.0,
      "step": 32855
    },
    {
      "entropy": 5.670301294326782,
      "epoch": 3.690514966024597,
      "grad_norm": 1.1875,
      "learning_rate": 0.000369576559951393,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22643517404794694,
      "num_tokens": 73372580.0,
      "step": 32860
    },
    {
      "entropy": 5.651171731948852,
      "epoch": 3.6910765429325547,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036954012884114946,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2265417292714119,
      "num_tokens": 73383970.0,
      "step": 32865
    },
    {
      "entropy": 5.616010951995849,
      "epoch": 3.691638119840512,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003695036947207115,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2319076418876648,
      "num_tokens": 73394472.0,
      "step": 32870
    },
    {
      "entropy": 5.607106399536133,
      "epoch": 3.6921996967484696,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036946725759123936,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21791497766971588,
      "num_tokens": 73405415.0,
      "step": 32875
    },
    {
      "entropy": 5.667752408981324,
      "epoch": 3.6927612736564273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003694308174538929,
      "loss": 4.9399,
      "mean_token_accuracy": 0.22347363233566284,
      "num_tokens": 73416583.0,
      "step": 32880
    },
    {
      "entropy": 5.702135515213013,
      "epoch": 3.693322850564385,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036939437430983274,
      "loss": 5.0459,
      "mean_token_accuracy": 0.22230492830276488,
      "num_tokens": 73427408.0,
      "step": 32885
    },
    {
      "entropy": 5.600741577148438,
      "epoch": 3.6938844274723426,
      "grad_norm": 1.125,
      "learning_rate": 0.00036935792816021905,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2372066169977188,
      "num_tokens": 73438509.0,
      "step": 32890
    },
    {
      "entropy": 5.61663122177124,
      "epoch": 3.6944460043803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036932147900621234,
      "loss": 4.9284,
      "mean_token_accuracy": 0.23072850108146667,
      "num_tokens": 73449197.0,
      "step": 32895
    },
    {
      "entropy": 5.661715412139893,
      "epoch": 3.6950075812882575,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036928502684897314,
      "loss": 5.0143,
      "mean_token_accuracy": 0.21789206564426422,
      "num_tokens": 73460777.0,
      "step": 32900
    },
    {
      "entropy": 5.719080543518066,
      "epoch": 3.695569158196215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036924857168966216,
      "loss": 4.9903,
      "mean_token_accuracy": 0.22456386387348176,
      "num_tokens": 73471747.0,
      "step": 32905
    },
    {
      "entropy": 5.520402956008911,
      "epoch": 3.6961307351041723,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003692121135294401,
      "loss": 4.8064,
      "mean_token_accuracy": 0.23506252318620682,
      "num_tokens": 73481258.0,
      "step": 32910
    },
    {
      "entropy": 5.600296974182129,
      "epoch": 3.69669231201213,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036917565236946776,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2275716930627823,
      "num_tokens": 73492073.0,
      "step": 32915
    },
    {
      "entropy": 5.596223306655884,
      "epoch": 3.6972538889200877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036913918821090616,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21960689276456832,
      "num_tokens": 73502813.0,
      "step": 32920
    },
    {
      "entropy": 5.680485677719116,
      "epoch": 3.697815465828045,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036910272105491625,
      "loss": 4.9785,
      "mean_token_accuracy": 0.23192473798990249,
      "num_tokens": 73514417.0,
      "step": 32925
    },
    {
      "entropy": 5.665118455886841,
      "epoch": 3.6983770427360025,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036906625090265925,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22893057018518448,
      "num_tokens": 73525636.0,
      "step": 32930
    },
    {
      "entropy": 5.605863046646118,
      "epoch": 3.69893861964396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036902977775529636,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22283730804920196,
      "num_tokens": 73536557.0,
      "step": 32935
    },
    {
      "entropy": 5.596952629089356,
      "epoch": 3.699500196551918,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003689933016139888,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2372545838356018,
      "num_tokens": 73547910.0,
      "step": 32940
    },
    {
      "entropy": 5.535828733444214,
      "epoch": 3.7000617734598755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003689568224798981,
      "loss": 4.8723,
      "mean_token_accuracy": 0.23502295166254045,
      "num_tokens": 73557583.0,
      "step": 32945
    },
    {
      "entropy": 5.596096515655518,
      "epoch": 3.7006233503678327,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003689203403541857,
      "loss": 4.9556,
      "mean_token_accuracy": 0.222739277780056,
      "num_tokens": 73567300.0,
      "step": 32950
    },
    {
      "entropy": 5.674807786941528,
      "epoch": 3.7011849272757904,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036888385523801325,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2200673907995224,
      "num_tokens": 73577387.0,
      "step": 32955
    },
    {
      "entropy": 5.74084095954895,
      "epoch": 3.701746504183748,
      "grad_norm": 1.125,
      "learning_rate": 0.00036884736713254247,
      "loss": 5.0239,
      "mean_token_accuracy": 0.22086202204227448,
      "num_tokens": 73588340.0,
      "step": 32960
    },
    {
      "entropy": 5.67548246383667,
      "epoch": 3.7023080810917053,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036881087603893515,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22282915711402893,
      "num_tokens": 73599650.0,
      "step": 32965
    },
    {
      "entropy": 5.72567400932312,
      "epoch": 3.702869657999663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036877438195835307,
      "loss": 5.0534,
      "mean_token_accuracy": 0.22417525500059127,
      "num_tokens": 73611279.0,
      "step": 32970
    },
    {
      "entropy": 5.568027114868164,
      "epoch": 3.7034312349076206,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003687378848919584,
      "loss": 4.9638,
      "mean_token_accuracy": 0.22294799685478212,
      "num_tokens": 73623279.0,
      "step": 32975
    },
    {
      "entropy": 5.615837717056275,
      "epoch": 3.7039928118155783,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036870138484091297,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21895844489336014,
      "num_tokens": 73633756.0,
      "step": 32980
    },
    {
      "entropy": 5.723199367523193,
      "epoch": 3.704554388723536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036866488180637926,
      "loss": 5.072,
      "mean_token_accuracy": 0.21554359793663025,
      "num_tokens": 73646033.0,
      "step": 32985
    },
    {
      "entropy": 5.63420352935791,
      "epoch": 3.705115965631493,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003686283757895193,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2218160629272461,
      "num_tokens": 73655833.0,
      "step": 32990
    },
    {
      "entropy": 5.711342716217041,
      "epoch": 3.705677542539451,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036859186679149555,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21596214771270753,
      "num_tokens": 73667755.0,
      "step": 32995
    },
    {
      "entropy": 5.712758016586304,
      "epoch": 3.7062391194474085,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003685553548134705,
      "loss": 5.0652,
      "mean_token_accuracy": 0.22402504533529283,
      "num_tokens": 73680537.0,
      "step": 33000
    },
    {
      "epoch": 3.7062391194474085,
      "eval_entropy": 5.39406407823383,
      "eval_loss": 5.0710320472717285,
      "eval_mean_token_accuracy": 0.22827773395159948,
      "eval_num_tokens": 73680537.0,
      "eval_runtime": 18.5138,
      "eval_samples_per_second": 1675.128,
      "eval_steps_per_second": 209.411,
      "step": 33000
    },
    {
      "entropy": 5.658633279800415,
      "epoch": 3.7068006963553657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003685188398566066,
      "loss": 4.916,
      "mean_token_accuracy": 0.23173293620347976,
      "num_tokens": 73692446.0,
      "step": 33005
    },
    {
      "entropy": 5.593331718444825,
      "epoch": 3.7073622732633233,
      "grad_norm": 1.125,
      "learning_rate": 0.0003684823219220666,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2232112169265747,
      "num_tokens": 73704059.0,
      "step": 33010
    },
    {
      "entropy": 5.584496355056762,
      "epoch": 3.707923850171281,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003684458010110133,
      "loss": 4.9577,
      "mean_token_accuracy": 0.22661312967538833,
      "num_tokens": 73715234.0,
      "step": 33015
    },
    {
      "entropy": 5.674624824523926,
      "epoch": 3.7084854270792382,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003684092771246094,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22135655879974364,
      "num_tokens": 73726930.0,
      "step": 33020
    },
    {
      "entropy": 5.67268214225769,
      "epoch": 3.709047003987196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003683727502640178,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2238362818956375,
      "num_tokens": 73738107.0,
      "step": 33025
    },
    {
      "entropy": 5.682896471023559,
      "epoch": 3.7096085808951536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003683362204304017,
      "loss": 5.0796,
      "mean_token_accuracy": 0.21642850935459138,
      "num_tokens": 73749165.0,
      "step": 33030
    },
    {
      "entropy": 5.642362880706787,
      "epoch": 3.710170157803111,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003682996876249241,
      "loss": 4.9282,
      "mean_token_accuracy": 0.23243032544851303,
      "num_tokens": 73760921.0,
      "step": 33035
    },
    {
      "entropy": 5.622722053527832,
      "epoch": 3.710731734711069,
      "grad_norm": 1.125,
      "learning_rate": 0.0003682631518487483,
      "loss": 4.981,
      "mean_token_accuracy": 0.22621459513902664,
      "num_tokens": 73772045.0,
      "step": 33040
    },
    {
      "entropy": 5.615737390518189,
      "epoch": 3.711293311619026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003682266131030375,
      "loss": 4.907,
      "mean_token_accuracy": 0.23363373130559922,
      "num_tokens": 73782698.0,
      "step": 33045
    },
    {
      "entropy": 5.540624618530273,
      "epoch": 3.7118548885269838,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036819007138895525,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22691967338323593,
      "num_tokens": 73793758.0,
      "step": 33050
    },
    {
      "entropy": 5.631842708587646,
      "epoch": 3.7124164654349414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000368153526707665,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2176589250564575,
      "num_tokens": 73805479.0,
      "step": 33055
    },
    {
      "entropy": 5.628508710861206,
      "epoch": 3.7129780423428986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003681169790603303,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2206690415740013,
      "num_tokens": 73817361.0,
      "step": 33060
    },
    {
      "entropy": 5.6222373962402346,
      "epoch": 3.7135396192508563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036808042844811484,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2252813160419464,
      "num_tokens": 73829651.0,
      "step": 33065
    },
    {
      "entropy": 5.590271377563477,
      "epoch": 3.714101196158814,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036804387487218254,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22958937287330627,
      "num_tokens": 73839608.0,
      "step": 33070
    },
    {
      "entropy": 5.629891443252563,
      "epoch": 3.7146627730667716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000368007318333697,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22549425959587097,
      "num_tokens": 73850663.0,
      "step": 33075
    },
    {
      "entropy": 5.659318685531616,
      "epoch": 3.7152243499747293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003679707588338225,
      "loss": 4.9908,
      "mean_token_accuracy": 0.22957434952259065,
      "num_tokens": 73862441.0,
      "step": 33080
    },
    {
      "entropy": 5.641404390335083,
      "epoch": 3.7157859268826865,
      "grad_norm": 1.171875,
      "learning_rate": 0.000367934196373723,
      "loss": 4.9432,
      "mean_token_accuracy": 0.23049639612436296,
      "num_tokens": 73873549.0,
      "step": 33085
    },
    {
      "entropy": 5.611599969863891,
      "epoch": 3.716347503790644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036789763095456264,
      "loss": 4.926,
      "mean_token_accuracy": 0.22918736785650254,
      "num_tokens": 73884872.0,
      "step": 33090
    },
    {
      "entropy": 5.700418853759766,
      "epoch": 3.716909080698602,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003678610625775056,
      "loss": 5.1397,
      "mean_token_accuracy": 0.210645492374897,
      "num_tokens": 73896137.0,
      "step": 33095
    },
    {
      "entropy": 5.569653797149658,
      "epoch": 3.717470657606559,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036782449124371643,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22630729973316194,
      "num_tokens": 73906135.0,
      "step": 33100
    },
    {
      "entropy": 5.702715063095093,
      "epoch": 3.7180322345145167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036778791695435935,
      "loss": 4.9655,
      "mean_token_accuracy": 0.22732293754816055,
      "num_tokens": 73916856.0,
      "step": 33105
    },
    {
      "entropy": 5.527461051940918,
      "epoch": 3.7185938114224744,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036775133971059904,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23560245037078859,
      "num_tokens": 73928268.0,
      "step": 33110
    },
    {
      "entropy": 5.633535575866699,
      "epoch": 3.7191553883304316,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003677147595136001,
      "loss": 5.0193,
      "mean_token_accuracy": 0.22344628423452378,
      "num_tokens": 73940327.0,
      "step": 33115
    },
    {
      "entropy": 5.716259241104126,
      "epoch": 3.7197169652383892,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036767817636452723,
      "loss": 4.9912,
      "mean_token_accuracy": 0.23285637348890303,
      "num_tokens": 73951488.0,
      "step": 33120
    },
    {
      "entropy": 5.6018068313598635,
      "epoch": 3.720278542146347,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036764159026454534,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22457560151815414,
      "num_tokens": 73963479.0,
      "step": 33125
    },
    {
      "entropy": 5.515965175628662,
      "epoch": 3.7208401190543046,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003676050012148193,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22937691062688828,
      "num_tokens": 73974878.0,
      "step": 33130
    },
    {
      "entropy": 5.5814777374267575,
      "epoch": 3.7214016959622622,
      "grad_norm": 1.21875,
      "learning_rate": 0.000367568409216514,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22408433705568315,
      "num_tokens": 73985983.0,
      "step": 33135
    },
    {
      "entropy": 5.630363655090332,
      "epoch": 3.7219632728702194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003675318142707947,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22293757647275925,
      "num_tokens": 73997141.0,
      "step": 33140
    },
    {
      "entropy": 5.670739030838012,
      "epoch": 3.722524849778177,
      "grad_norm": 1.125,
      "learning_rate": 0.0003674952163788266,
      "loss": 4.9359,
      "mean_token_accuracy": 0.22615765482187272,
      "num_tokens": 74008262.0,
      "step": 33145
    },
    {
      "entropy": 5.642913675308227,
      "epoch": 3.7230864266861348,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036745861554177486,
      "loss": 4.9603,
      "mean_token_accuracy": 0.22566355764865875,
      "num_tokens": 74020465.0,
      "step": 33150
    },
    {
      "entropy": 5.663116979598999,
      "epoch": 3.723648003594092,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036742201176080496,
      "loss": 4.9608,
      "mean_token_accuracy": 0.230789452791214,
      "num_tokens": 74031100.0,
      "step": 33155
    },
    {
      "entropy": 5.6703290939331055,
      "epoch": 3.7242095805020496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003673854050370824,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21765979528427123,
      "num_tokens": 74043313.0,
      "step": 33160
    },
    {
      "entropy": 5.595255184173584,
      "epoch": 3.7247711574100073,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036734879537177276,
      "loss": 4.871,
      "mean_token_accuracy": 0.22671747356653213,
      "num_tokens": 74054540.0,
      "step": 33165
    },
    {
      "entropy": 5.704985475540161,
      "epoch": 3.725332734317965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003673121827660415,
      "loss": 5.1221,
      "mean_token_accuracy": 0.2152329668402672,
      "num_tokens": 74065636.0,
      "step": 33170
    },
    {
      "entropy": 5.591818332672119,
      "epoch": 3.7258943112259226,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036727556722105465,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22654181271791457,
      "num_tokens": 74076516.0,
      "step": 33175
    },
    {
      "entropy": 5.657384443283081,
      "epoch": 3.72645588813388,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036723894873797786,
      "loss": 5.0052,
      "mean_token_accuracy": 0.22011086195707322,
      "num_tokens": 74087481.0,
      "step": 33180
    },
    {
      "entropy": 5.721255874633789,
      "epoch": 3.7270174650418375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036720232731797726,
      "loss": 5.0157,
      "mean_token_accuracy": 0.22500188648700714,
      "num_tokens": 74098577.0,
      "step": 33185
    },
    {
      "entropy": 5.714065074920654,
      "epoch": 3.727579041949795,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036716570296221873,
      "loss": 4.9968,
      "mean_token_accuracy": 0.21358095109462738,
      "num_tokens": 74108813.0,
      "step": 33190
    },
    {
      "entropy": 5.601620674133301,
      "epoch": 3.7281406188577524,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036712907567186857,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22124974578619003,
      "num_tokens": 74118937.0,
      "step": 33195
    },
    {
      "entropy": 5.640904760360717,
      "epoch": 3.72870219576571,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003670924454480928,
      "loss": 4.9893,
      "mean_token_accuracy": 0.2252463236451149,
      "num_tokens": 74129879.0,
      "step": 33200
    },
    {
      "entropy": 5.606813430786133,
      "epoch": 3.7292637726736677,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003670558122920579,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22847734689712523,
      "num_tokens": 74140455.0,
      "step": 33205
    },
    {
      "entropy": 5.613111066818237,
      "epoch": 3.729825349581625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036701917620493026,
      "loss": 5.0045,
      "mean_token_accuracy": 0.22376242876052857,
      "num_tokens": 74152508.0,
      "step": 33210
    },
    {
      "entropy": 5.686007308959961,
      "epoch": 3.7303869264895826,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003669825371878764,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2359789565205574,
      "num_tokens": 74163700.0,
      "step": 33215
    },
    {
      "entropy": 5.607150745391846,
      "epoch": 3.7309485033975403,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036694589524206275,
      "loss": 4.9721,
      "mean_token_accuracy": 0.22480219155550002,
      "num_tokens": 74174715.0,
      "step": 33220
    },
    {
      "entropy": 5.689783048629761,
      "epoch": 3.731510080305498,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003669092503686562,
      "loss": 5.003,
      "mean_token_accuracy": 0.22625354528427125,
      "num_tokens": 74186679.0,
      "step": 33225
    },
    {
      "entropy": 5.67692608833313,
      "epoch": 3.7320716572134556,
      "grad_norm": 1.125,
      "learning_rate": 0.00036687260256882347,
      "loss": 4.9507,
      "mean_token_accuracy": 0.22821559756994247,
      "num_tokens": 74196382.0,
      "step": 33230
    },
    {
      "entropy": 5.652218437194824,
      "epoch": 3.732633234121413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036683595184373147,
      "loss": 5.0075,
      "mean_token_accuracy": 0.22923788130283357,
      "num_tokens": 74207587.0,
      "step": 33235
    },
    {
      "entropy": 5.575684404373169,
      "epoch": 3.7331948110293705,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003667992981945471,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22510542571544648,
      "num_tokens": 74218826.0,
      "step": 33240
    },
    {
      "entropy": 5.7281615257263185,
      "epoch": 3.733756387937328,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036676264162243744,
      "loss": 5.0593,
      "mean_token_accuracy": 0.22492226362228393,
      "num_tokens": 74230450.0,
      "step": 33245
    },
    {
      "entropy": 5.692827415466309,
      "epoch": 3.7343179648452853,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003667259821285698,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21784417927265168,
      "num_tokens": 74241438.0,
      "step": 33250
    },
    {
      "entropy": 5.699987077713013,
      "epoch": 3.734879541753243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003666893197141112,
      "loss": 5.0509,
      "mean_token_accuracy": 0.22216208428144454,
      "num_tokens": 74252460.0,
      "step": 33255
    },
    {
      "entropy": 5.65326156616211,
      "epoch": 3.7354411186612007,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003666526543802291,
      "loss": 4.933,
      "mean_token_accuracy": 0.23034279346466063,
      "num_tokens": 74262984.0,
      "step": 33260
    },
    {
      "entropy": 5.651962041854858,
      "epoch": 3.7360026955691583,
      "grad_norm": 1.125,
      "learning_rate": 0.000366615986128091,
      "loss": 5.0766,
      "mean_token_accuracy": 0.22094163596630095,
      "num_tokens": 74274368.0,
      "step": 33265
    },
    {
      "entropy": 5.566403722763061,
      "epoch": 3.736564272477116,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003665793149588644,
      "loss": 4.9121,
      "mean_token_accuracy": 0.2361702188849449,
      "num_tokens": 74286394.0,
      "step": 33270
    },
    {
      "entropy": 5.569729948043824,
      "epoch": 3.737125849385073,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036654264087371686,
      "loss": 4.8928,
      "mean_token_accuracy": 0.23079442083835602,
      "num_tokens": 74296803.0,
      "step": 33275
    },
    {
      "entropy": 5.668832683563233,
      "epoch": 3.737687426293031,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036650596387381615,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2288878783583641,
      "num_tokens": 74306471.0,
      "step": 33280
    },
    {
      "entropy": 5.66676607131958,
      "epoch": 3.7382490032009885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036646928396033,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23278652727603913,
      "num_tokens": 74318875.0,
      "step": 33285
    },
    {
      "entropy": 5.625677871704101,
      "epoch": 3.7388105801089457,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036643260113442634,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22360328882932662,
      "num_tokens": 74328464.0,
      "step": 33290
    },
    {
      "entropy": 5.656302738189697,
      "epoch": 3.7393721570169034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003663959153972732,
      "loss": 4.9607,
      "mean_token_accuracy": 0.23328703343868257,
      "num_tokens": 74339462.0,
      "step": 33295
    },
    {
      "entropy": 5.662198257446289,
      "epoch": 3.739933733924861,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003663592267500388,
      "loss": 5.0077,
      "mean_token_accuracy": 0.2237364411354065,
      "num_tokens": 74350667.0,
      "step": 33300
    },
    {
      "entropy": 5.580706548690796,
      "epoch": 3.7404953108328187,
      "grad_norm": 1.125,
      "learning_rate": 0.0003663225351938911,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22948286235332488,
      "num_tokens": 74361734.0,
      "step": 33305
    },
    {
      "entropy": 5.64864706993103,
      "epoch": 3.741056887740776,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003662858407299984,
      "loss": 5.0234,
      "mean_token_accuracy": 0.2162955582141876,
      "num_tokens": 74373465.0,
      "step": 33310
    },
    {
      "entropy": 5.6675632953643795,
      "epoch": 3.7416184646487336,
      "grad_norm": 1.125,
      "learning_rate": 0.00036624914335952924,
      "loss": 4.951,
      "mean_token_accuracy": 0.22192691564559935,
      "num_tokens": 74384853.0,
      "step": 33315
    },
    {
      "entropy": 5.5797483921051025,
      "epoch": 3.7421800415566913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036621244308365195,
      "loss": 4.8369,
      "mean_token_accuracy": 0.22648457884788514,
      "num_tokens": 74394356.0,
      "step": 33320
    },
    {
      "entropy": 5.650160217285157,
      "epoch": 3.742741618464649,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036617573990353503,
      "loss": 5.023,
      "mean_token_accuracy": 0.2192048415541649,
      "num_tokens": 74406721.0,
      "step": 33325
    },
    {
      "entropy": 5.672672986984253,
      "epoch": 3.743303195372606,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003661390338203472,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22507069557905196,
      "num_tokens": 74418217.0,
      "step": 33330
    },
    {
      "entropy": 5.599617338180542,
      "epoch": 3.743864772280564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003661023248352572,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22752737402915954,
      "num_tokens": 74428644.0,
      "step": 33335
    },
    {
      "entropy": 5.573272085189819,
      "epoch": 3.7444263491885215,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003660656129494339,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2322968900203705,
      "num_tokens": 74439575.0,
      "step": 33340
    },
    {
      "entropy": 5.641100883483887,
      "epoch": 3.7449879260964787,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036602889816404613,
      "loss": 5.0037,
      "mean_token_accuracy": 0.2238263502717018,
      "num_tokens": 74450191.0,
      "step": 33345
    },
    {
      "entropy": 5.647699403762817,
      "epoch": 3.7455495030044363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036599218048026294,
      "loss": 4.99,
      "mean_token_accuracy": 0.2251133680343628,
      "num_tokens": 74460816.0,
      "step": 33350
    },
    {
      "entropy": 5.65433292388916,
      "epoch": 3.746111079912394,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036595545989925345,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23354181945323943,
      "num_tokens": 74470999.0,
      "step": 33355
    },
    {
      "entropy": 5.60689377784729,
      "epoch": 3.7466726568203517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036591873642218676,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23813909590244292,
      "num_tokens": 74480884.0,
      "step": 33360
    },
    {
      "entropy": 5.551606798171997,
      "epoch": 3.7472342337283093,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036588201005023223,
      "loss": 4.9491,
      "mean_token_accuracy": 0.22651530504226686,
      "num_tokens": 74492500.0,
      "step": 33365
    },
    {
      "entropy": 5.620770025253296,
      "epoch": 3.7477958106362665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003658452807845593,
      "loss": 4.932,
      "mean_token_accuracy": 0.2277318298816681,
      "num_tokens": 74504635.0,
      "step": 33370
    },
    {
      "entropy": 5.692186069488526,
      "epoch": 3.748357387544224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036580854862633747,
      "loss": 5.0032,
      "mean_token_accuracy": 0.22101709991693497,
      "num_tokens": 74515743.0,
      "step": 33375
    },
    {
      "entropy": 5.6051593780517575,
      "epoch": 3.748918964452182,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003657718135767362,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22221226692199708,
      "num_tokens": 74528089.0,
      "step": 33380
    },
    {
      "entropy": 5.714623069763183,
      "epoch": 3.749480541360139,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003657350756369251,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22858311831951142,
      "num_tokens": 74538543.0,
      "step": 33385
    },
    {
      "entropy": 5.719079446792603,
      "epoch": 3.7500421182680967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036569833480807403,
      "loss": 5.0423,
      "mean_token_accuracy": 0.22411667555570602,
      "num_tokens": 74549825.0,
      "step": 33390
    },
    {
      "entropy": 5.589680528640747,
      "epoch": 3.7506036951760544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003656615910913528,
      "loss": 4.8883,
      "mean_token_accuracy": 0.23068648725748062,
      "num_tokens": 74560304.0,
      "step": 33395
    },
    {
      "entropy": 5.5315406799316404,
      "epoch": 3.751165272084012,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036562484448793135,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2268046259880066,
      "num_tokens": 74571549.0,
      "step": 33400
    },
    {
      "entropy": 5.68357629776001,
      "epoch": 3.7517268489919693,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003655880949989798,
      "loss": 5.0136,
      "mean_token_accuracy": 0.22150442004203796,
      "num_tokens": 74581865.0,
      "step": 33405
    },
    {
      "entropy": 5.633614635467529,
      "epoch": 3.752288425899927,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036555134262566806,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22586355954408646,
      "num_tokens": 74592079.0,
      "step": 33410
    },
    {
      "entropy": 5.64742259979248,
      "epoch": 3.7528500028078846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003655145873691665,
      "loss": 4.975,
      "mean_token_accuracy": 0.22479832917451859,
      "num_tokens": 74603058.0,
      "step": 33415
    },
    {
      "entropy": 5.620590686798096,
      "epoch": 3.7534115797158423,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036547782923064537,
      "loss": 4.9774,
      "mean_token_accuracy": 0.22484598755836488,
      "num_tokens": 74614008.0,
      "step": 33420
    },
    {
      "entropy": 5.725003623962403,
      "epoch": 3.7539731566237995,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036544106821127505,
      "loss": 5.0628,
      "mean_token_accuracy": 0.21339548081159593,
      "num_tokens": 74625146.0,
      "step": 33425
    },
    {
      "entropy": 5.7360086917877195,
      "epoch": 3.754534733531757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000365404304312226,
      "loss": 5.0551,
      "mean_token_accuracy": 0.22342599481344222,
      "num_tokens": 74636578.0,
      "step": 33430
    },
    {
      "entropy": 5.622378253936768,
      "epoch": 3.755096310439715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036536753753466897,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2246295228600502,
      "num_tokens": 74648551.0,
      "step": 33435
    },
    {
      "entropy": 5.620927953720093,
      "epoch": 3.755657887347672,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003653307678797744,
      "loss": 4.8382,
      "mean_token_accuracy": 0.23341196179389953,
      "num_tokens": 74658712.0,
      "step": 33440
    },
    {
      "entropy": 5.583125972747803,
      "epoch": 3.7562194642556297,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003652939953487131,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2255986139178276,
      "num_tokens": 74670196.0,
      "step": 33445
    },
    {
      "entropy": 5.678762006759643,
      "epoch": 3.7567810411635874,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036525721994265606,
      "loss": 5.0137,
      "mean_token_accuracy": 0.22643790394067764,
      "num_tokens": 74680577.0,
      "step": 33450
    },
    {
      "entropy": 5.655512571334839,
      "epoch": 3.757342618071545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003652204416627741,
      "loss": 4.8866,
      "mean_token_accuracy": 0.23234240859746932,
      "num_tokens": 74692054.0,
      "step": 33455
    },
    {
      "entropy": 5.641273355484008,
      "epoch": 3.7579041949795027,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036518366051023824,
      "loss": 4.9687,
      "mean_token_accuracy": 0.22623401284217834,
      "num_tokens": 74703324.0,
      "step": 33460
    },
    {
      "entropy": 5.557391786575318,
      "epoch": 3.75846577188746,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003651468764862197,
      "loss": 4.9182,
      "mean_token_accuracy": 0.23003669530153276,
      "num_tokens": 74713028.0,
      "step": 33465
    },
    {
      "entropy": 5.566825103759766,
      "epoch": 3.7590273487954176,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036511008959188966,
      "loss": 4.9152,
      "mean_token_accuracy": 0.2235660195350647,
      "num_tokens": 74723964.0,
      "step": 33470
    },
    {
      "entropy": 5.645557594299317,
      "epoch": 3.759588925703375,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036507329982841935,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22863034009933472,
      "num_tokens": 74734928.0,
      "step": 33475
    },
    {
      "entropy": 5.7247633934021,
      "epoch": 3.7601505026113324,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003650365071969803,
      "loss": 5.0783,
      "mean_token_accuracy": 0.22078145891427994,
      "num_tokens": 74747198.0,
      "step": 33480
    },
    {
      "entropy": 5.672784614562988,
      "epoch": 3.76071207951929,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003649997116987439,
      "loss": 4.9752,
      "mean_token_accuracy": 0.2286621242761612,
      "num_tokens": 74758180.0,
      "step": 33485
    },
    {
      "entropy": 5.654520034790039,
      "epoch": 3.7612736564272478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003649629133348818,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2262137070298195,
      "num_tokens": 74769227.0,
      "step": 33490
    },
    {
      "entropy": 5.643897819519043,
      "epoch": 3.7618352333352054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036492611210656574,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22829956859350203,
      "num_tokens": 74780347.0,
      "step": 33495
    },
    {
      "entropy": 5.631436538696289,
      "epoch": 3.7623968102431626,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003648893080149673,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2242441326379776,
      "num_tokens": 74791902.0,
      "step": 33500
    },
    {
      "entropy": 5.608282136917114,
      "epoch": 3.7629583871511203,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036485250106125844,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2224905788898468,
      "num_tokens": 74803530.0,
      "step": 33505
    },
    {
      "entropy": 5.632586574554443,
      "epoch": 3.763519964059078,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000364815691246611,
      "loss": 4.9268,
      "mean_token_accuracy": 0.23350360840559006,
      "num_tokens": 74815128.0,
      "step": 33510
    },
    {
      "entropy": 5.708566617965698,
      "epoch": 3.7640815409670356,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036477887857219726,
      "loss": 5.1218,
      "mean_token_accuracy": 0.21991842091083527,
      "num_tokens": 74827729.0,
      "step": 33515
    },
    {
      "entropy": 5.5769267082214355,
      "epoch": 3.764643117874993,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036474206303918914,
      "loss": 4.9179,
      "mean_token_accuracy": 0.22490644752979277,
      "num_tokens": 74837922.0,
      "step": 33520
    },
    {
      "entropy": 5.61744270324707,
      "epoch": 3.7652046947829505,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000364705244648759,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22958477437496186,
      "num_tokens": 74848239.0,
      "step": 33525
    },
    {
      "entropy": 5.6762457370758055,
      "epoch": 3.765766271690908,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003646684234020789,
      "loss": 4.9802,
      "mean_token_accuracy": 0.2237373411655426,
      "num_tokens": 74860454.0,
      "step": 33530
    },
    {
      "entropy": 5.578694581985474,
      "epoch": 3.7663278485988654,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036463159930032163,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22481112629175187,
      "num_tokens": 74871849.0,
      "step": 33535
    },
    {
      "entropy": 5.652529716491699,
      "epoch": 3.766889425506823,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036459477234465935,
      "loss": 5.0286,
      "mean_token_accuracy": 0.21500314623117447,
      "num_tokens": 74883877.0,
      "step": 33540
    },
    {
      "entropy": 5.812592840194702,
      "epoch": 3.7674510024147807,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036455794253626486,
      "loss": 5.2068,
      "mean_token_accuracy": 0.21721251904964448,
      "num_tokens": 74897218.0,
      "step": 33545
    },
    {
      "entropy": 5.725434589385986,
      "epoch": 3.7680125793227384,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036452110987631066,
      "loss": 4.9703,
      "mean_token_accuracy": 0.2206196293234825,
      "num_tokens": 74908708.0,
      "step": 33550
    },
    {
      "entropy": 5.652174234390259,
      "epoch": 3.768574156230696,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036448427436596966,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2317342832684517,
      "num_tokens": 74919664.0,
      "step": 33555
    },
    {
      "entropy": 5.596145677566528,
      "epoch": 3.7691357331386532,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003644474360064146,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22648061215877532,
      "num_tokens": 74930386.0,
      "step": 33560
    },
    {
      "entropy": 5.604064512252807,
      "epoch": 3.769697310046611,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003644105947988185,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2194860801100731,
      "num_tokens": 74942835.0,
      "step": 33565
    },
    {
      "entropy": 5.66646556854248,
      "epoch": 3.7702588869545686,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036437375074435446,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22008954137563705,
      "num_tokens": 74954332.0,
      "step": 33570
    },
    {
      "entropy": 5.722876834869385,
      "epoch": 3.770820463862526,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003643369038441955,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2223044142127037,
      "num_tokens": 74965530.0,
      "step": 33575
    },
    {
      "entropy": 5.630555725097656,
      "epoch": 3.7713820407704834,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036430005409951476,
      "loss": 4.986,
      "mean_token_accuracy": 0.22010679841041564,
      "num_tokens": 74976319.0,
      "step": 33580
    },
    {
      "entropy": 5.638914585113525,
      "epoch": 3.771943617678441,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036426320151148575,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2209792673587799,
      "num_tokens": 74987948.0,
      "step": 33585
    },
    {
      "entropy": 5.6294371604919435,
      "epoch": 3.7725051945863988,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036422634608128184,
      "loss": 4.9791,
      "mean_token_accuracy": 0.22459350675344467,
      "num_tokens": 74998989.0,
      "step": 33590
    },
    {
      "entropy": 5.673090887069702,
      "epoch": 3.773066771494356,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036418948781007636,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22766402512788772,
      "num_tokens": 75010510.0,
      "step": 33595
    },
    {
      "entropy": 5.645921325683593,
      "epoch": 3.7736283484023136,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036415262669904305,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21641819924116135,
      "num_tokens": 75020708.0,
      "step": 33600
    },
    {
      "entropy": 5.681043577194214,
      "epoch": 3.7741899253102713,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036411576274935546,
      "loss": 4.9832,
      "mean_token_accuracy": 0.22690669000148772,
      "num_tokens": 75031468.0,
      "step": 33605
    },
    {
      "entropy": 5.592912054061889,
      "epoch": 3.774751502218229,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036407889596218757,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23164460808038712,
      "num_tokens": 75042170.0,
      "step": 33610
    },
    {
      "entropy": 5.589148616790771,
      "epoch": 3.775313079126186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000364042026338713,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2277824029326439,
      "num_tokens": 75053916.0,
      "step": 33615
    },
    {
      "entropy": 5.62470326423645,
      "epoch": 3.775874656034144,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036400515388010574,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2349364697933197,
      "num_tokens": 75065026.0,
      "step": 33620
    },
    {
      "entropy": 5.582204961776734,
      "epoch": 3.7764362329421015,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003639682785875399,
      "loss": 4.9666,
      "mean_token_accuracy": 0.22269817739725112,
      "num_tokens": 75075617.0,
      "step": 33625
    },
    {
      "entropy": 5.627172517776489,
      "epoch": 3.7769978098500587,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036393140046218957,
      "loss": 4.9606,
      "mean_token_accuracy": 0.22151016741991042,
      "num_tokens": 75085931.0,
      "step": 33630
    },
    {
      "entropy": 5.6991808891296385,
      "epoch": 3.7775593867580164,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036389451950522895,
      "loss": 4.985,
      "mean_token_accuracy": 0.23327501267194747,
      "num_tokens": 75096404.0,
      "step": 33635
    },
    {
      "entropy": 5.719199275970459,
      "epoch": 3.778120963665974,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036385763571783236,
      "loss": 5.0609,
      "mean_token_accuracy": 0.22069209367036818,
      "num_tokens": 75108300.0,
      "step": 33640
    },
    {
      "entropy": 5.6599242210388185,
      "epoch": 3.7786825405739317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003638207491011742,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22671174705028535,
      "num_tokens": 75119799.0,
      "step": 33645
    },
    {
      "entropy": 5.547819423675537,
      "epoch": 3.7792441174818894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003637838596564289,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22774609327316284,
      "num_tokens": 75130803.0,
      "step": 33650
    },
    {
      "entropy": 5.625913286209107,
      "epoch": 3.7798056943898466,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036374696738477105,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2309648483991623,
      "num_tokens": 75140621.0,
      "step": 33655
    },
    {
      "entropy": 5.630714082717896,
      "epoch": 3.7803672712978043,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003637100722873754,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22234544903039932,
      "num_tokens": 75152446.0,
      "step": 33660
    },
    {
      "entropy": 5.669089365005493,
      "epoch": 3.780928848205762,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003636731743654166,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2223552167415619,
      "num_tokens": 75162936.0,
      "step": 33665
    },
    {
      "entropy": 5.5465660572052,
      "epoch": 3.781490425113719,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036363627362006955,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22877311557531357,
      "num_tokens": 75174184.0,
      "step": 33670
    },
    {
      "entropy": 5.491238784790039,
      "epoch": 3.782052002021677,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003635993700525092,
      "loss": 4.8716,
      "mean_token_accuracy": 0.23014989048242568,
      "num_tokens": 75185609.0,
      "step": 33675
    },
    {
      "entropy": 5.7114355087280275,
      "epoch": 3.7826135789296345,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036356246366391054,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2214427947998047,
      "num_tokens": 75196098.0,
      "step": 33680
    },
    {
      "entropy": 5.748769044876099,
      "epoch": 3.783175155837592,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036352555445544864,
      "loss": 5.0123,
      "mean_token_accuracy": 0.2238513246178627,
      "num_tokens": 75206411.0,
      "step": 33685
    },
    {
      "entropy": 5.591867542266845,
      "epoch": 3.78373673274555,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036348864242829874,
      "loss": 4.8726,
      "mean_token_accuracy": 0.22575775682926177,
      "num_tokens": 75217029.0,
      "step": 33690
    },
    {
      "entropy": 5.529708766937256,
      "epoch": 3.784298309653507,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036345172758363623,
      "loss": 4.8819,
      "mean_token_accuracy": 0.22255726754665375,
      "num_tokens": 75228124.0,
      "step": 33695
    },
    {
      "entropy": 5.535009860992432,
      "epoch": 3.7848598865614647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003634148099226364,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22043834924697875,
      "num_tokens": 75239776.0,
      "step": 33700
    },
    {
      "entropy": 5.57876148223877,
      "epoch": 3.7854214634694223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036337788944647466,
      "loss": 4.8649,
      "mean_token_accuracy": 0.22858378142118455,
      "num_tokens": 75251354.0,
      "step": 33705
    },
    {
      "entropy": 5.680274105072021,
      "epoch": 3.7859830403773795,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003633409661563267,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21707622110843658,
      "num_tokens": 75263167.0,
      "step": 33710
    },
    {
      "entropy": 5.5809091091156,
      "epoch": 3.786544617285337,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036330404005336807,
      "loss": 4.919,
      "mean_token_accuracy": 0.23034048527479173,
      "num_tokens": 75274200.0,
      "step": 33715
    },
    {
      "entropy": 5.641228723526001,
      "epoch": 3.787106194193295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003632671111387746,
      "loss": 5.0087,
      "mean_token_accuracy": 0.22747375518083573,
      "num_tokens": 75287060.0,
      "step": 33720
    },
    {
      "entropy": 5.704783391952515,
      "epoch": 3.787667771101252,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036323017941372203,
      "loss": 5.0124,
      "mean_token_accuracy": 0.22785856276750566,
      "num_tokens": 75297862.0,
      "step": 33725
    },
    {
      "entropy": 5.715614223480225,
      "epoch": 3.7882293480092097,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003631932448793864,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21968714743852616,
      "num_tokens": 75308844.0,
      "step": 33730
    },
    {
      "entropy": 5.7194504737854,
      "epoch": 3.7887909249171674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003631563075369436,
      "loss": 5.0597,
      "mean_token_accuracy": 0.22027010172605516,
      "num_tokens": 75321035.0,
      "step": 33735
    },
    {
      "entropy": 5.566005611419678,
      "epoch": 3.789352501825125,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036311936738756977,
      "loss": 4.8338,
      "mean_token_accuracy": 0.22755513787269593,
      "num_tokens": 75331407.0,
      "step": 33740
    },
    {
      "entropy": 5.591706418991089,
      "epoch": 3.7899140787330827,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003630824244324411,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22862876504659652,
      "num_tokens": 75343082.0,
      "step": 33745
    },
    {
      "entropy": 5.611850786209106,
      "epoch": 3.79047565564104,
      "grad_norm": 1.125,
      "learning_rate": 0.0003630454786727339,
      "loss": 4.905,
      "mean_token_accuracy": 0.2310652405023575,
      "num_tokens": 75353345.0,
      "step": 33750
    },
    {
      "entropy": 5.707428789138794,
      "epoch": 3.7910372325489976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003630085301096245,
      "loss": 5.0644,
      "mean_token_accuracy": 0.22429730296134948,
      "num_tokens": 75364039.0,
      "step": 33755
    },
    {
      "entropy": 5.732564544677734,
      "epoch": 3.7915988094569553,
      "grad_norm": 1.125,
      "learning_rate": 0.00036297157874428933,
      "loss": 4.9846,
      "mean_token_accuracy": 0.22490770369768143,
      "num_tokens": 75374166.0,
      "step": 33760
    },
    {
      "entropy": 5.646312093734741,
      "epoch": 3.7921603863649125,
      "grad_norm": 1.125,
      "learning_rate": 0.00036293462457790505,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22448398619890214,
      "num_tokens": 75384792.0,
      "step": 33765
    },
    {
      "entropy": 5.649959325790405,
      "epoch": 3.79272196327287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036289766761164814,
      "loss": 5.017,
      "mean_token_accuracy": 0.21905437260866165,
      "num_tokens": 75395478.0,
      "step": 33770
    },
    {
      "entropy": 5.65677695274353,
      "epoch": 3.793283540180828,
      "grad_norm": 1.125,
      "learning_rate": 0.00036286070784669543,
      "loss": 5.0059,
      "mean_token_accuracy": 0.2251300647854805,
      "num_tokens": 75406293.0,
      "step": 33775
    },
    {
      "entropy": 5.679422998428345,
      "epoch": 3.7938451170887855,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036282374528422373,
      "loss": 4.9926,
      "mean_token_accuracy": 0.23189188838005065,
      "num_tokens": 75417588.0,
      "step": 33780
    },
    {
      "entropy": 5.571109056472778,
      "epoch": 3.794406693996743,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036278677992540993,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22649744749069214,
      "num_tokens": 75428651.0,
      "step": 33785
    },
    {
      "entropy": 5.69052152633667,
      "epoch": 3.7949682709047003,
      "grad_norm": 1.15625,
      "learning_rate": 0.000362749811771431,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22789925038814546,
      "num_tokens": 75438808.0,
      "step": 33790
    },
    {
      "entropy": 5.615707540512085,
      "epoch": 3.795529847812658,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003627128408234641,
      "loss": 4.8604,
      "mean_token_accuracy": 0.2233271449804306,
      "num_tokens": 75449326.0,
      "step": 33795
    },
    {
      "entropy": 5.613779163360595,
      "epoch": 3.7960914247206157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003626758670826861,
      "loss": 4.9839,
      "mean_token_accuracy": 0.22514677196741104,
      "num_tokens": 75460499.0,
      "step": 33800
    },
    {
      "entropy": 5.605423307418823,
      "epoch": 3.796653001628573,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003626388905502747,
      "loss": 4.9461,
      "mean_token_accuracy": 0.23020306676626207,
      "num_tokens": 75471031.0,
      "step": 33805
    },
    {
      "entropy": 5.657457971572876,
      "epoch": 3.7972145785365305,
      "grad_norm": 1.125,
      "learning_rate": 0.000362601911227407,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22791180163621902,
      "num_tokens": 75483608.0,
      "step": 33810
    },
    {
      "entropy": 5.6922351837158205,
      "epoch": 3.797776155444488,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003625649291152604,
      "loss": 4.9948,
      "mean_token_accuracy": 0.23235345482826233,
      "num_tokens": 75495221.0,
      "step": 33815
    },
    {
      "entropy": 5.617900753021241,
      "epoch": 3.7983377323524454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036252794421501257,
      "loss": 4.9101,
      "mean_token_accuracy": 0.23149592727422713,
      "num_tokens": 75507245.0,
      "step": 33820
    },
    {
      "entropy": 5.713507509231567,
      "epoch": 3.798899309260403,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036249095652784096,
      "loss": 5.0559,
      "mean_token_accuracy": 0.21982128024101258,
      "num_tokens": 75518525.0,
      "step": 33825
    },
    {
      "entropy": 5.6400065422058105,
      "epoch": 3.7994608861683608,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036245396605492344,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21886951923370362,
      "num_tokens": 75529674.0,
      "step": 33830
    },
    {
      "entropy": 5.569281530380249,
      "epoch": 3.8000224630763184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003624169727974376,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23693425357341766,
      "num_tokens": 75539596.0,
      "step": 33835
    },
    {
      "entropy": 5.569328117370605,
      "epoch": 3.800584039984276,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036237997675656167,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22932648211717604,
      "num_tokens": 75551164.0,
      "step": 33840
    },
    {
      "entropy": 5.692564058303833,
      "epoch": 3.8011456168922333,
      "grad_norm": 1.0,
      "learning_rate": 0.0003623429779334732,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21649116575717925,
      "num_tokens": 75563758.0,
      "step": 33845
    },
    {
      "entropy": 5.751452445983887,
      "epoch": 3.801707193800191,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036230597632935053,
      "loss": 5.0785,
      "mean_token_accuracy": 0.22411869168281556,
      "num_tokens": 75574332.0,
      "step": 33850
    },
    {
      "entropy": 5.670445919036865,
      "epoch": 3.8022687707081486,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003622689719453717,
      "loss": 5.0751,
      "mean_token_accuracy": 0.2159237891435623,
      "num_tokens": 75586229.0,
      "step": 33855
    },
    {
      "entropy": 5.695751047134399,
      "epoch": 3.802830347616106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036223196478271495,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22606064528226852,
      "num_tokens": 75597396.0,
      "step": 33860
    },
    {
      "entropy": 5.670016813278198,
      "epoch": 3.8033919245240635,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003621949548425585,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21845293790102005,
      "num_tokens": 75609171.0,
      "step": 33865
    },
    {
      "entropy": 5.709834909439087,
      "epoch": 3.803953501432021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000362157942126081,
      "loss": 5.0195,
      "mean_token_accuracy": 0.22851179242134095,
      "num_tokens": 75619855.0,
      "step": 33870
    },
    {
      "entropy": 5.728710174560547,
      "epoch": 3.804515078339979,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003621209266344608,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2240785226225853,
      "num_tokens": 75631810.0,
      "step": 33875
    },
    {
      "entropy": 5.602294206619263,
      "epoch": 3.8050766552479365,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003620839083688765,
      "loss": 4.9777,
      "mean_token_accuracy": 0.23301631957292557,
      "num_tokens": 75644894.0,
      "step": 33880
    },
    {
      "entropy": 5.629120397567749,
      "epoch": 3.8056382321558937,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003620468873305068,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22876843065023422,
      "num_tokens": 75657357.0,
      "step": 33885
    },
    {
      "entropy": 5.6451764583587645,
      "epoch": 3.8061998090638514,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036200986352053036,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22509985119104386,
      "num_tokens": 75667357.0,
      "step": 33890
    },
    {
      "entropy": 5.620304679870605,
      "epoch": 3.806761385971809,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036197283694012615,
      "loss": 4.879,
      "mean_token_accuracy": 0.2299247056245804,
      "num_tokens": 75678758.0,
      "step": 33895
    },
    {
      "entropy": 5.706650161743164,
      "epoch": 3.8073229628797662,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003619358075904732,
      "loss": 4.9413,
      "mean_token_accuracy": 0.23393814712762834,
      "num_tokens": 75688392.0,
      "step": 33900
    },
    {
      "entropy": 5.634753084182739,
      "epoch": 3.807884539787724,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036189877547275023,
      "loss": 4.9173,
      "mean_token_accuracy": 0.22898991107940675,
      "num_tokens": 75699839.0,
      "step": 33905
    },
    {
      "entropy": 5.697948694229126,
      "epoch": 3.8084461166956816,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036186174058813666,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2252582788467407,
      "num_tokens": 75711877.0,
      "step": 33910
    },
    {
      "entropy": 5.599661445617675,
      "epoch": 3.8090076936036388,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036182470293781154,
      "loss": 4.8805,
      "mean_token_accuracy": 0.23095754235982896,
      "num_tokens": 75721646.0,
      "step": 33915
    },
    {
      "entropy": 5.65549602508545,
      "epoch": 3.8095692705115964,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003617876625229543,
      "loss": 5.0653,
      "mean_token_accuracy": 0.21543714106082917,
      "num_tokens": 75733291.0,
      "step": 33920
    },
    {
      "entropy": 5.642237424850464,
      "epoch": 3.810130847419554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003617506193447442,
      "loss": 4.9627,
      "mean_token_accuracy": 0.23293279707431794,
      "num_tokens": 75744475.0,
      "step": 33925
    },
    {
      "entropy": 5.5831938743591305,
      "epoch": 3.8106924243275118,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003617135734043606,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22468263208866118,
      "num_tokens": 75755098.0,
      "step": 33930
    },
    {
      "entropy": 5.701311111450195,
      "epoch": 3.8112540012354694,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003616765247029833,
      "loss": 4.9549,
      "mean_token_accuracy": 0.23208729773759842,
      "num_tokens": 75767405.0,
      "step": 33935
    },
    {
      "entropy": 5.669096326828003,
      "epoch": 3.8118155781434266,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036163947324179186,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21450380682945253,
      "num_tokens": 75778614.0,
      "step": 33940
    },
    {
      "entropy": 5.622358894348144,
      "epoch": 3.8123771550513843,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036160241902196587,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23189041167497634,
      "num_tokens": 75789087.0,
      "step": 33945
    },
    {
      "entropy": 5.630401611328125,
      "epoch": 3.812938731959342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036156536204468535,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21974524706602097,
      "num_tokens": 75799431.0,
      "step": 33950
    },
    {
      "entropy": 5.618727207183838,
      "epoch": 3.813500308867299,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036152830231113013,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2240130662918091,
      "num_tokens": 75810149.0,
      "step": 33955
    },
    {
      "entropy": 5.609409809112549,
      "epoch": 3.814061885775257,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003614912398224802,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2189757212996483,
      "num_tokens": 75820072.0,
      "step": 33960
    },
    {
      "entropy": 5.657251262664795,
      "epoch": 3.8146234626832145,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003614541745799156,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21717006117105483,
      "num_tokens": 75831017.0,
      "step": 33965
    },
    {
      "entropy": 5.628020620346069,
      "epoch": 3.815185039591172,
      "grad_norm": 1.125,
      "learning_rate": 0.0003614171065846166,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22433853447437285,
      "num_tokens": 75842606.0,
      "step": 33970
    },
    {
      "entropy": 5.6104272365570065,
      "epoch": 3.81574661649913,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003613800358377635,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2315964072942734,
      "num_tokens": 75853820.0,
      "step": 33975
    },
    {
      "entropy": 5.57620210647583,
      "epoch": 3.816308193407087,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003613429623405364,
      "loss": 4.9076,
      "mean_token_accuracy": 0.23225937485694886,
      "num_tokens": 75864959.0,
      "step": 33980
    },
    {
      "entropy": 5.623465251922608,
      "epoch": 3.8168697703150447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036130588609411597,
      "loss": 4.8781,
      "mean_token_accuracy": 0.23458192348480225,
      "num_tokens": 75877247.0,
      "step": 33985
    },
    {
      "entropy": 5.703696537017822,
      "epoch": 3.8174313472230024,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036126880709968264,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2118788868188858,
      "num_tokens": 75888654.0,
      "step": 33990
    },
    {
      "entropy": 5.652743005752564,
      "epoch": 3.8179929241309596,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036123172535841705,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22371436208486556,
      "num_tokens": 75899953.0,
      "step": 33995
    },
    {
      "entropy": 5.656157636642456,
      "epoch": 3.8185545010389172,
      "grad_norm": 1.125,
      "learning_rate": 0.0003611946408714998,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2311303660273552,
      "num_tokens": 75911174.0,
      "step": 34000
    },
    {
      "entropy": 5.586760234832764,
      "epoch": 3.819116077946875,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036115755364011186,
      "loss": 4.9361,
      "mean_token_accuracy": 0.23688559383153915,
      "num_tokens": 75922221.0,
      "step": 34005
    },
    {
      "entropy": 5.535141324996948,
      "epoch": 3.819677654854832,
      "grad_norm": 1.140625,
      "learning_rate": 0.000361120463665434,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22635346204042434,
      "num_tokens": 75932630.0,
      "step": 34010
    },
    {
      "entropy": 5.608441543579102,
      "epoch": 3.82023923176279,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003610833709486471,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2255694881081581,
      "num_tokens": 75944490.0,
      "step": 34015
    },
    {
      "entropy": 5.63274359703064,
      "epoch": 3.8208008086707475,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003610462754909323,
      "loss": 4.9212,
      "mean_token_accuracy": 0.23147680163383483,
      "num_tokens": 75956608.0,
      "step": 34020
    },
    {
      "entropy": 5.670217895507813,
      "epoch": 3.821362385578705,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036100917729347073,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2255997195839882,
      "num_tokens": 75967516.0,
      "step": 34025
    },
    {
      "entropy": 5.581395864486694,
      "epoch": 3.8219239624866628,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036097207635744363,
      "loss": 4.8484,
      "mean_token_accuracy": 0.2294728070497513,
      "num_tokens": 75978329.0,
      "step": 34030
    },
    {
      "entropy": 5.713042259216309,
      "epoch": 3.82248553939462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036093497268403215,
      "loss": 4.9675,
      "mean_token_accuracy": 0.23234012871980667,
      "num_tokens": 75990170.0,
      "step": 34035
    },
    {
      "entropy": 5.607354211807251,
      "epoch": 3.8230471163025777,
      "grad_norm": 1.03125,
      "learning_rate": 0.000360897866274418,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2298387199640274,
      "num_tokens": 76002368.0,
      "step": 34040
    },
    {
      "entropy": 5.607992029190063,
      "epoch": 3.8236086932105353,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036086075712978233,
      "loss": 4.9391,
      "mean_token_accuracy": 0.22751665413379668,
      "num_tokens": 76012867.0,
      "step": 34045
    },
    {
      "entropy": 5.6330023288726805,
      "epoch": 3.8241702701184925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036082364525130684,
      "loss": 5.0049,
      "mean_token_accuracy": 0.2203501656651497,
      "num_tokens": 76023900.0,
      "step": 34050
    },
    {
      "entropy": 5.6169592380523685,
      "epoch": 3.82473184702645,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003607865306401732,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23201765716075898,
      "num_tokens": 76034296.0,
      "step": 34055
    },
    {
      "entropy": 5.678588199615478,
      "epoch": 3.825293423934408,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036074941329756327,
      "loss": 4.9831,
      "mean_token_accuracy": 0.2245662271976471,
      "num_tokens": 76045355.0,
      "step": 34060
    },
    {
      "entropy": 5.61533842086792,
      "epoch": 3.8258550008423655,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003607122932246587,
      "loss": 4.8738,
      "mean_token_accuracy": 0.23585113137960434,
      "num_tokens": 76056158.0,
      "step": 34065
    },
    {
      "entropy": 5.623618125915527,
      "epoch": 3.826416577750323,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036067517042264147,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22408605217933655,
      "num_tokens": 76066930.0,
      "step": 34070
    },
    {
      "entropy": 5.65265679359436,
      "epoch": 3.8269781546582804,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003606380448926936,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22966134548187256,
      "num_tokens": 76078791.0,
      "step": 34075
    },
    {
      "entropy": 5.697808313369751,
      "epoch": 3.827539731566238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003606009166359972,
      "loss": 5.0221,
      "mean_token_accuracy": 0.22355708330869675,
      "num_tokens": 76090356.0,
      "step": 34080
    },
    {
      "entropy": 5.563124132156372,
      "epoch": 3.8281013084741957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036056378565373434,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22913725674152374,
      "num_tokens": 76101137.0,
      "step": 34085
    },
    {
      "entropy": 5.527953577041626,
      "epoch": 3.828662885382153,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003605266519470874,
      "loss": 4.8166,
      "mean_token_accuracy": 0.22855840772390365,
      "num_tokens": 76111661.0,
      "step": 34090
    },
    {
      "entropy": 5.669996452331543,
      "epoch": 3.8292244622901106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003604895155172386,
      "loss": 4.9638,
      "mean_token_accuracy": 0.23123592436313628,
      "num_tokens": 76122670.0,
      "step": 34095
    },
    {
      "entropy": 5.6825943946838375,
      "epoch": 3.8297860391980683,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003604523763653706,
      "loss": 5.0674,
      "mean_token_accuracy": 0.22045151591300965,
      "num_tokens": 76133899.0,
      "step": 34100
    },
    {
      "entropy": 5.686571311950684,
      "epoch": 3.8303476161060255,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036041523449266573,
      "loss": 5.0113,
      "mean_token_accuracy": 0.21912098675966263,
      "num_tokens": 76144865.0,
      "step": 34105
    },
    {
      "entropy": 5.722179412841797,
      "epoch": 3.830909193013983,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036037808990030667,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22690102756023406,
      "num_tokens": 76157392.0,
      "step": 34110
    },
    {
      "entropy": 5.687779808044434,
      "epoch": 3.831470769921941,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003603409425894761,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22162825167179107,
      "num_tokens": 76168339.0,
      "step": 34115
    },
    {
      "entropy": 5.59902515411377,
      "epoch": 3.8320323468298985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003603037925613568,
      "loss": 4.9323,
      "mean_token_accuracy": 0.22993853837251663,
      "num_tokens": 76179377.0,
      "step": 34120
    },
    {
      "entropy": 5.639101934432984,
      "epoch": 3.832593923737856,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003602666398171317,
      "loss": 4.9487,
      "mean_token_accuracy": 0.2214517042040825,
      "num_tokens": 76190030.0,
      "step": 34125
    },
    {
      "entropy": 5.557655668258667,
      "epoch": 3.8331555006458133,
      "grad_norm": 1.125,
      "learning_rate": 0.00036022948435798373,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22884094417095185,
      "num_tokens": 76201339.0,
      "step": 34130
    },
    {
      "entropy": 5.664643573760986,
      "epoch": 3.833717077553771,
      "grad_norm": 1.171875,
      "learning_rate": 0.000360192326185096,
      "loss": 4.9854,
      "mean_token_accuracy": 0.23073702454566955,
      "num_tokens": 76211760.0,
      "step": 34135
    },
    {
      "entropy": 5.685590076446533,
      "epoch": 3.8342786544617287,
      "grad_norm": 1.125,
      "learning_rate": 0.0003601551652996514,
      "loss": 5.0113,
      "mean_token_accuracy": 0.22460865825414658,
      "num_tokens": 76222573.0,
      "step": 34140
    },
    {
      "entropy": 5.6983953475952145,
      "epoch": 3.834840231369686,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003601180017028334,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21607346534729005,
      "num_tokens": 76233547.0,
      "step": 34145
    },
    {
      "entropy": 5.613628196716308,
      "epoch": 3.8354018082776435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003600808353958252,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22708118408918382,
      "num_tokens": 76246067.0,
      "step": 34150
    },
    {
      "entropy": 5.566878938674927,
      "epoch": 3.835963385185601,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003600436663798102,
      "loss": 4.8323,
      "mean_token_accuracy": 0.2373175099492073,
      "num_tokens": 76257698.0,
      "step": 34155
    },
    {
      "entropy": 5.573920249938965,
      "epoch": 3.836524962093559,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036000649465597185,
      "loss": 4.8991,
      "mean_token_accuracy": 0.23347548246383668,
      "num_tokens": 76268533.0,
      "step": 34160
    },
    {
      "entropy": 5.575894069671631,
      "epoch": 3.8370865390015165,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035996932022549377,
      "loss": 4.864,
      "mean_token_accuracy": 0.23420187532901765,
      "num_tokens": 76278529.0,
      "step": 34165
    },
    {
      "entropy": 5.600338363647461,
      "epoch": 3.8376481159094737,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035993214308955955,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21968107372522355,
      "num_tokens": 76289461.0,
      "step": 34170
    },
    {
      "entropy": 5.665293836593628,
      "epoch": 3.8382096928174314,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000359894963249353,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22926202416419983,
      "num_tokens": 76300641.0,
      "step": 34175
    },
    {
      "entropy": 5.648373937606811,
      "epoch": 3.838771269725389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035985778070605785,
      "loss": 5.0551,
      "mean_token_accuracy": 0.21854605674743652,
      "num_tokens": 76310918.0,
      "step": 34180
    },
    {
      "entropy": 5.587065887451172,
      "epoch": 3.8393328466333463,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003598205954608581,
      "loss": 4.8551,
      "mean_token_accuracy": 0.2341305360198021,
      "num_tokens": 76323037.0,
      "step": 34185
    },
    {
      "entropy": 5.643431711196899,
      "epoch": 3.839894423541304,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035978340751493767,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22835810482501984,
      "num_tokens": 76334252.0,
      "step": 34190
    },
    {
      "entropy": 5.67451548576355,
      "epoch": 3.8404560004492616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035974621686948063,
      "loss": 5.0371,
      "mean_token_accuracy": 0.22175079137086867,
      "num_tokens": 76344928.0,
      "step": 34195
    },
    {
      "entropy": 5.692766761779785,
      "epoch": 3.841017577357219,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003597090235256712,
      "loss": 5.0526,
      "mean_token_accuracy": 0.21789965480566026,
      "num_tokens": 76356529.0,
      "step": 34200
    },
    {
      "entropy": 5.6208751678466795,
      "epoch": 3.8415791542651765,
      "grad_norm": 1.25,
      "learning_rate": 0.0003596718274846936,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21967916786670685,
      "num_tokens": 76366842.0,
      "step": 34205
    },
    {
      "entropy": 5.667011690139771,
      "epoch": 3.842140731173134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003596346287477322,
      "loss": 5.0,
      "mean_token_accuracy": 0.22421439439058305,
      "num_tokens": 76377109.0,
      "step": 34210
    },
    {
      "entropy": 5.639051818847657,
      "epoch": 3.842702308081092,
      "grad_norm": 1.125,
      "learning_rate": 0.0003595974273159714,
      "loss": 4.9595,
      "mean_token_accuracy": 0.2222393661737442,
      "num_tokens": 76387621.0,
      "step": 34215
    },
    {
      "entropy": 5.673620080947876,
      "epoch": 3.8432638849890495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003595602231905957,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21821044534444808,
      "num_tokens": 76398537.0,
      "step": 34220
    },
    {
      "entropy": 5.665403175354004,
      "epoch": 3.8438254618970067,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035952301637278973,
      "loss": 4.9536,
      "mean_token_accuracy": 0.22851288616657256,
      "num_tokens": 76410192.0,
      "step": 34225
    },
    {
      "entropy": 5.619189882278443,
      "epoch": 3.8443870388049644,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035948580686373803,
      "loss": 4.937,
      "mean_token_accuracy": 0.23202238082885743,
      "num_tokens": 76420024.0,
      "step": 34230
    },
    {
      "entropy": 5.634699869155884,
      "epoch": 3.844948615712922,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035944859466462554,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2298835277557373,
      "num_tokens": 76430833.0,
      "step": 34235
    },
    {
      "entropy": 5.618652963638306,
      "epoch": 3.8455101926208792,
      "grad_norm": 1.125,
      "learning_rate": 0.000359411379776637,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2232523649930954,
      "num_tokens": 76442800.0,
      "step": 34240
    },
    {
      "entropy": 5.696603727340698,
      "epoch": 3.846071769528837,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035937416220095745,
      "loss": 4.9432,
      "mean_token_accuracy": 0.2294759914278984,
      "num_tokens": 76454988.0,
      "step": 34245
    },
    {
      "entropy": 5.763772869110108,
      "epoch": 3.8466333464367946,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003593369419387718,
      "loss": 5.061,
      "mean_token_accuracy": 0.21542100012302398,
      "num_tokens": 76466523.0,
      "step": 34250
    },
    {
      "entropy": 5.771168851852417,
      "epoch": 3.847194923344752,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003592997189912652,
      "loss": 5.0813,
      "mean_token_accuracy": 0.21084815859794617,
      "num_tokens": 76478691.0,
      "step": 34255
    },
    {
      "entropy": 5.677971458435058,
      "epoch": 3.84775650025271,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035926249335962296,
      "loss": 5.0492,
      "mean_token_accuracy": 0.22096782326698303,
      "num_tokens": 76490544.0,
      "step": 34260
    },
    {
      "entropy": 5.607558298110962,
      "epoch": 3.848318077160667,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035922526504503016,
      "loss": 4.8777,
      "mean_token_accuracy": 0.22817958295345306,
      "num_tokens": 76500269.0,
      "step": 34265
    },
    {
      "entropy": 5.55791916847229,
      "epoch": 3.8488796540686248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003591880340486723,
      "loss": 4.891,
      "mean_token_accuracy": 0.2242135375738144,
      "num_tokens": 76511458.0,
      "step": 34270
    },
    {
      "entropy": 5.594175529479981,
      "epoch": 3.8494412309765824,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003591508003717347,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21962914913892745,
      "num_tokens": 76521722.0,
      "step": 34275
    },
    {
      "entropy": 5.700991153717041,
      "epoch": 3.8500028078845396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035911356401540306,
      "loss": 5.0525,
      "mean_token_accuracy": 0.2273401528596878,
      "num_tokens": 76532935.0,
      "step": 34280
    },
    {
      "entropy": 5.626913833618164,
      "epoch": 3.8505643847924973,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003590763249808629,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22796329110860825,
      "num_tokens": 76543545.0,
      "step": 34285
    },
    {
      "entropy": 5.607088947296143,
      "epoch": 3.851125961700455,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035903908326929995,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23198058903217317,
      "num_tokens": 76553621.0,
      "step": 34290
    },
    {
      "entropy": 5.689903593063354,
      "epoch": 3.851687538608412,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035900183888190005,
      "loss": 4.971,
      "mean_token_accuracy": 0.2324637219309807,
      "num_tokens": 76563154.0,
      "step": 34295
    },
    {
      "entropy": 5.690414810180664,
      "epoch": 3.85224911551637,
      "grad_norm": 1.125,
      "learning_rate": 0.0003589645918198489,
      "loss": 4.9307,
      "mean_token_accuracy": 0.22093158066272736,
      "num_tokens": 76573930.0,
      "step": 34300
    },
    {
      "entropy": 5.622734212875367,
      "epoch": 3.8528106924243275,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035892734208433264,
      "loss": 4.882,
      "mean_token_accuracy": 0.2305169329047203,
      "num_tokens": 76585216.0,
      "step": 34305
    },
    {
      "entropy": 5.63922233581543,
      "epoch": 3.853372269332285,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035889008967653725,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2316674143075943,
      "num_tokens": 76596479.0,
      "step": 34310
    },
    {
      "entropy": 5.699024343490601,
      "epoch": 3.853933846240243,
      "grad_norm": 1.25,
      "learning_rate": 0.0003588528345976489,
      "loss": 5.0529,
      "mean_token_accuracy": 0.22079471200704576,
      "num_tokens": 76607123.0,
      "step": 34315
    },
    {
      "entropy": 5.712420415878296,
      "epoch": 3.8544954231482,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003588155768488537,
      "loss": 5.003,
      "mean_token_accuracy": 0.23065245896577835,
      "num_tokens": 76616898.0,
      "step": 34320
    },
    {
      "entropy": 5.620301294326782,
      "epoch": 3.8550570000561577,
      "grad_norm": 1.40625,
      "learning_rate": 0.00035877831643133803,
      "loss": 4.8141,
      "mean_token_accuracy": 0.23450421839952468,
      "num_tokens": 76627528.0,
      "step": 34325
    },
    {
      "entropy": 5.641709041595459,
      "epoch": 3.8556185769641154,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003587410533462883,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22716473042964935,
      "num_tokens": 76639173.0,
      "step": 34330
    },
    {
      "entropy": 5.64822244644165,
      "epoch": 3.8561801538720726,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035870378759489104,
      "loss": 4.9945,
      "mean_token_accuracy": 0.22346815466880798,
      "num_tokens": 76648970.0,
      "step": 34335
    },
    {
      "entropy": 5.58773455619812,
      "epoch": 3.8567417307800302,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003586665191783326,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2277997687458992,
      "num_tokens": 76659101.0,
      "step": 34340
    },
    {
      "entropy": 5.6623132705688475,
      "epoch": 3.857303307687988,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003586292480977998,
      "loss": 4.961,
      "mean_token_accuracy": 0.22656034529209138,
      "num_tokens": 76670087.0,
      "step": 34345
    },
    {
      "entropy": 5.687554979324341,
      "epoch": 3.8578648845959456,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035859197435447927,
      "loss": 4.991,
      "mean_token_accuracy": 0.22757034450769426,
      "num_tokens": 76681116.0,
      "step": 34350
    },
    {
      "entropy": 5.610872650146485,
      "epoch": 3.8584264615039032,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035855469794955787,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2247127428650856,
      "num_tokens": 76692050.0,
      "step": 34355
    },
    {
      "entropy": 5.766586399078369,
      "epoch": 3.8589880384118604,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035851741888422254,
      "loss": 5.0442,
      "mean_token_accuracy": 0.22109283208847047,
      "num_tokens": 76704338.0,
      "step": 34360
    },
    {
      "entropy": 5.6027388095855715,
      "epoch": 3.859549615319818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003584801371596601,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2321038395166397,
      "num_tokens": 76715300.0,
      "step": 34365
    },
    {
      "entropy": 5.692982196807861,
      "epoch": 3.8601111922277758,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003584428527770578,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22063495218753815,
      "num_tokens": 76726654.0,
      "step": 34370
    },
    {
      "entropy": 5.543818521499634,
      "epoch": 3.860672769135733,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035840556573760266,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23517182618379592,
      "num_tokens": 76737087.0,
      "step": 34375
    },
    {
      "entropy": 5.581181573867798,
      "epoch": 3.8612343460436906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003583682760424819,
      "loss": 4.892,
      "mean_token_accuracy": 0.23366724103689193,
      "num_tokens": 76748006.0,
      "step": 34380
    },
    {
      "entropy": 5.5793126106262205,
      "epoch": 3.8617959229516483,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035833098369288303,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2376499891281128,
      "num_tokens": 76758063.0,
      "step": 34385
    },
    {
      "entropy": 5.625092601776123,
      "epoch": 3.8623574998596055,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035829368868999325,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22871581315994263,
      "num_tokens": 76769080.0,
      "step": 34390
    },
    {
      "entropy": 5.635976886749267,
      "epoch": 3.862919076767563,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035825639103500005,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2257082134485245,
      "num_tokens": 76780743.0,
      "step": 34395
    },
    {
      "entropy": 5.6297248840332035,
      "epoch": 3.863480653675521,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035821909072909113,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22914607375860213,
      "num_tokens": 76792277.0,
      "step": 34400
    },
    {
      "entropy": 5.628566074371338,
      "epoch": 3.8640422305834785,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035818178777345417,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2291790559887886,
      "num_tokens": 76802979.0,
      "step": 34405
    },
    {
      "entropy": 5.582688093185425,
      "epoch": 3.864603807491436,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003581444821692767,
      "loss": 4.9031,
      "mean_token_accuracy": 0.23050492405891418,
      "num_tokens": 76815016.0,
      "step": 34410
    },
    {
      "entropy": 5.560706567764282,
      "epoch": 3.8651653843993934,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003581071739177467,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2359859585762024,
      "num_tokens": 76826360.0,
      "step": 34415
    },
    {
      "entropy": 5.690055704116821,
      "epoch": 3.865726961307351,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035806986302005205,
      "loss": 5.0034,
      "mean_token_accuracy": 0.23076231777668,
      "num_tokens": 76838876.0,
      "step": 34420
    },
    {
      "entropy": 5.585469341278076,
      "epoch": 3.8662885382153087,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003580325494773808,
      "loss": 4.8937,
      "mean_token_accuracy": 0.23289667218923568,
      "num_tokens": 76849156.0,
      "step": 34425
    },
    {
      "entropy": 5.648881483078003,
      "epoch": 3.866850115123266,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035799523329092093,
      "loss": 5.0598,
      "mean_token_accuracy": 0.2210083231329918,
      "num_tokens": 76859683.0,
      "step": 34430
    },
    {
      "entropy": 5.591734457015991,
      "epoch": 3.8674116920312236,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035795791446186073,
      "loss": 4.8662,
      "mean_token_accuracy": 0.2351946249604225,
      "num_tokens": 76870509.0,
      "step": 34435
    },
    {
      "entropy": 5.68055214881897,
      "epoch": 3.8679732689391813,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003579205929913882,
      "loss": 4.9952,
      "mean_token_accuracy": 0.22488543689250945,
      "num_tokens": 76882181.0,
      "step": 34440
    },
    {
      "entropy": 5.657679462432862,
      "epoch": 3.868534845847139,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003578832688806919,
      "loss": 5.0303,
      "mean_token_accuracy": 0.21812645941972733,
      "num_tokens": 76894709.0,
      "step": 34445
    },
    {
      "entropy": 5.6326089859008786,
      "epoch": 3.8690964227550966,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003578459421309602,
      "loss": 4.8968,
      "mean_token_accuracy": 0.23265665173530578,
      "num_tokens": 76905974.0,
      "step": 34450
    },
    {
      "entropy": 5.750566196441651,
      "epoch": 3.869657999663054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003578086127433816,
      "loss": 5.0631,
      "mean_token_accuracy": 0.22397608757019044,
      "num_tokens": 76918009.0,
      "step": 34455
    },
    {
      "entropy": 5.6830517768859865,
      "epoch": 3.8702195765710115,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035777128071914454,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22797795981168748,
      "num_tokens": 76928750.0,
      "step": 34460
    },
    {
      "entropy": 5.662598133087158,
      "epoch": 3.870781153478969,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003577339460594378,
      "loss": 5.0698,
      "mean_token_accuracy": 0.22183688133955,
      "num_tokens": 76939989.0,
      "step": 34465
    },
    {
      "entropy": 5.518632936477661,
      "epoch": 3.8713427303869263,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003576966087654502,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23338814675807953,
      "num_tokens": 76950026.0,
      "step": 34470
    },
    {
      "entropy": 5.720526313781738,
      "epoch": 3.871904307294884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035765926883837054,
      "loss": 5.0656,
      "mean_token_accuracy": 0.2178984120488167,
      "num_tokens": 76961198.0,
      "step": 34475
    },
    {
      "entropy": 5.6352269649505615,
      "epoch": 3.8724658842028417,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003576219262793877,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23418422937393188,
      "num_tokens": 76973012.0,
      "step": 34480
    },
    {
      "entropy": 5.645347213745117,
      "epoch": 3.873027461110799,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035758458108969056,
      "loss": 4.9511,
      "mean_token_accuracy": 0.22731645107269288,
      "num_tokens": 76983613.0,
      "step": 34485
    },
    {
      "entropy": 5.703821992874145,
      "epoch": 3.8735890380187565,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003575472332704684,
      "loss": 5.084,
      "mean_token_accuracy": 0.2136029526591301,
      "num_tokens": 76995471.0,
      "step": 34490
    },
    {
      "entropy": 5.599470758438111,
      "epoch": 3.874150614926714,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035750988282291037,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23358374536037446,
      "num_tokens": 77005777.0,
      "step": 34495
    },
    {
      "entropy": 5.657947635650634,
      "epoch": 3.874712191834672,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035747252974820565,
      "loss": 5.0408,
      "mean_token_accuracy": 0.22017768323421477,
      "num_tokens": 77017099.0,
      "step": 34500
    },
    {
      "entropy": 5.601150751113892,
      "epoch": 3.8752737687426295,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035743517404754354,
      "loss": 5.0238,
      "mean_token_accuracy": 0.22185489386320115,
      "num_tokens": 77028350.0,
      "step": 34505
    },
    {
      "entropy": 5.626949167251587,
      "epoch": 3.8758353456505867,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003573978157221136,
      "loss": 4.9407,
      "mean_token_accuracy": 0.2350357547402382,
      "num_tokens": 77039355.0,
      "step": 34510
    },
    {
      "entropy": 5.735683345794678,
      "epoch": 3.8763969225585444,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035736045477310517,
      "loss": 5.0684,
      "mean_token_accuracy": 0.22087581753730773,
      "num_tokens": 77049966.0,
      "step": 34515
    },
    {
      "entropy": 5.678094911575317,
      "epoch": 3.876958499466502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035732309120170786,
      "loss": 4.938,
      "mean_token_accuracy": 0.2284926414489746,
      "num_tokens": 77061102.0,
      "step": 34520
    },
    {
      "entropy": 5.625797939300537,
      "epoch": 3.8775200763744593,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035728572500911155,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2214153677225113,
      "num_tokens": 77071516.0,
      "step": 34525
    },
    {
      "entropy": 5.634793472290039,
      "epoch": 3.878081653282417,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035724835619650573,
      "loss": 5.0488,
      "mean_token_accuracy": 0.2227686643600464,
      "num_tokens": 77081924.0,
      "step": 34530
    },
    {
      "entropy": 5.667774200439453,
      "epoch": 3.8786432301903746,
      "grad_norm": 1.125,
      "learning_rate": 0.0003572109847650804,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22516768872737886,
      "num_tokens": 77093064.0,
      "step": 34535
    },
    {
      "entropy": 5.654484653472901,
      "epoch": 3.8792048070983323,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035717361071602546,
      "loss": 4.9902,
      "mean_token_accuracy": 0.22562641948461531,
      "num_tokens": 77104428.0,
      "step": 34540
    },
    {
      "entropy": 5.633862400054932,
      "epoch": 3.87976638400629,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035713623405053083,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23819794058799743,
      "num_tokens": 77114949.0,
      "step": 34545
    },
    {
      "entropy": 5.67378454208374,
      "epoch": 3.880327960914247,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035709885476978666,
      "loss": 5.0343,
      "mean_token_accuracy": 0.22437143474817275,
      "num_tokens": 77125178.0,
      "step": 34550
    },
    {
      "entropy": 5.56029372215271,
      "epoch": 3.880889537822205,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035706147287498303,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23234472423791885,
      "num_tokens": 77135967.0,
      "step": 34555
    },
    {
      "entropy": 5.701435136795044,
      "epoch": 3.8814511147301625,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035702408836731043,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2295663133263588,
      "num_tokens": 77148052.0,
      "step": 34560
    },
    {
      "entropy": 5.699023199081421,
      "epoch": 3.8820126916381197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000356986701247959,
      "loss": 4.9706,
      "mean_token_accuracy": 0.22973718792200087,
      "num_tokens": 77159008.0,
      "step": 34565
    },
    {
      "entropy": 5.618608522415161,
      "epoch": 3.8825742685460773,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003569493115181192,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2223699226975441,
      "num_tokens": 77169846.0,
      "step": 34570
    },
    {
      "entropy": 5.608992004394532,
      "epoch": 3.883135845454035,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003569119191789815,
      "loss": 4.9811,
      "mean_token_accuracy": 0.22661227285861968,
      "num_tokens": 77182881.0,
      "step": 34575
    },
    {
      "entropy": 5.64581413269043,
      "epoch": 3.8836974223619922,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035687452423173655,
      "loss": 4.9432,
      "mean_token_accuracy": 0.22373273074626923,
      "num_tokens": 77193852.0,
      "step": 34580
    },
    {
      "entropy": 5.696767854690552,
      "epoch": 3.88425899926995,
      "grad_norm": 1.109375,
      "learning_rate": 0.000356837126677575,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2233378529548645,
      "num_tokens": 77205141.0,
      "step": 34585
    },
    {
      "entropy": 5.7162402153015135,
      "epoch": 3.8848205761779075,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003567997265176876,
      "loss": 5.028,
      "mean_token_accuracy": 0.21322995275259018,
      "num_tokens": 77217521.0,
      "step": 34590
    },
    {
      "entropy": 5.7629499435424805,
      "epoch": 3.885382153085865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035676232375326513,
      "loss": 5.0426,
      "mean_token_accuracy": 0.2200607419013977,
      "num_tokens": 77227468.0,
      "step": 34595
    },
    {
      "entropy": 5.583069801330566,
      "epoch": 3.885943729993823,
      "grad_norm": 1.125,
      "learning_rate": 0.00035672491838549867,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22796124517917632,
      "num_tokens": 77237908.0,
      "step": 34600
    },
    {
      "entropy": 5.548809623718261,
      "epoch": 3.88650530690178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000356687510415579,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23141830563545226,
      "num_tokens": 77248621.0,
      "step": 34605
    },
    {
      "entropy": 5.6721936702728275,
      "epoch": 3.8870668838097377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003566500998446973,
      "loss": 5.0153,
      "mean_token_accuracy": 0.2186934009194374,
      "num_tokens": 77260261.0,
      "step": 34610
    },
    {
      "entropy": 5.647407054901123,
      "epoch": 3.8876284607176954,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035661268667404485,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22239390909671783,
      "num_tokens": 77270985.0,
      "step": 34615
    },
    {
      "entropy": 5.672834968566894,
      "epoch": 3.8881900376256526,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003565752709048127,
      "loss": 4.995,
      "mean_token_accuracy": 0.2309156596660614,
      "num_tokens": 77281876.0,
      "step": 34620
    },
    {
      "entropy": 5.5849920272827145,
      "epoch": 3.8887516145336103,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003565378525381924,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22982754111289977,
      "num_tokens": 77293141.0,
      "step": 34625
    },
    {
      "entropy": 5.640004444122314,
      "epoch": 3.889313191441568,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003565004315753752,
      "loss": 4.9613,
      "mean_token_accuracy": 0.22957645058631898,
      "num_tokens": 77305723.0,
      "step": 34630
    },
    {
      "entropy": 5.584830236434937,
      "epoch": 3.8898747683495256,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035646300801755266,
      "loss": 4.898,
      "mean_token_accuracy": 0.2329012170433998,
      "num_tokens": 77317736.0,
      "step": 34635
    },
    {
      "entropy": 5.596729040145874,
      "epoch": 3.8904363452574833,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035642558186591635,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2244092419743538,
      "num_tokens": 77328293.0,
      "step": 34640
    },
    {
      "entropy": 5.659111022949219,
      "epoch": 3.8909979221654405,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003563881531216579,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2241094097495079,
      "num_tokens": 77339371.0,
      "step": 34645
    },
    {
      "entropy": 5.675293684005737,
      "epoch": 3.891559499073398,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003563507217859691,
      "loss": 5.0177,
      "mean_token_accuracy": 0.2169698104262352,
      "num_tokens": 77350234.0,
      "step": 34650
    },
    {
      "entropy": 5.6839048862457275,
      "epoch": 3.892121075981356,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003563132878600419,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22735877931118012,
      "num_tokens": 77361144.0,
      "step": 34655
    },
    {
      "entropy": 5.659879922866821,
      "epoch": 3.892682652889313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003562758513450678,
      "loss": 4.987,
      "mean_token_accuracy": 0.23237016201019287,
      "num_tokens": 77372090.0,
      "step": 34660
    },
    {
      "entropy": 5.689804553985596,
      "epoch": 3.8932442297972707,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003562384122422393,
      "loss": 5.0595,
      "mean_token_accuracy": 0.22016306817531586,
      "num_tokens": 77382974.0,
      "step": 34665
    },
    {
      "entropy": 5.683719348907471,
      "epoch": 3.8938058067052284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003562009705527481,
      "loss": 5.0051,
      "mean_token_accuracy": 0.22234435081481935,
      "num_tokens": 77395971.0,
      "step": 34670
    },
    {
      "entropy": 5.694003391265869,
      "epoch": 3.8943673836131856,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035616352627778665,
      "loss": 4.9685,
      "mean_token_accuracy": 0.21935991644859315,
      "num_tokens": 77405776.0,
      "step": 34675
    },
    {
      "entropy": 5.6897780418396,
      "epoch": 3.8949289605211432,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035612607941854695,
      "loss": 4.9984,
      "mean_token_accuracy": 0.22246501892805098,
      "num_tokens": 77417668.0,
      "step": 34680
    },
    {
      "entropy": 5.678988981246948,
      "epoch": 3.895490537429101,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003560886299762214,
      "loss": 4.9416,
      "mean_token_accuracy": 0.22773105800151824,
      "num_tokens": 77428581.0,
      "step": 34685
    },
    {
      "entropy": 5.697328710556031,
      "epoch": 3.8960521143370586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003560511779520025,
      "loss": 5.0389,
      "mean_token_accuracy": 0.22396613508462906,
      "num_tokens": 77441587.0,
      "step": 34690
    },
    {
      "entropy": 5.656877851486206,
      "epoch": 3.896613691245016,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003560137233470826,
      "loss": 4.9289,
      "mean_token_accuracy": 0.23142073750495912,
      "num_tokens": 77452579.0,
      "step": 34695
    },
    {
      "entropy": 5.596844291687011,
      "epoch": 3.8971752681529734,
      "grad_norm": 1.125,
      "learning_rate": 0.00035597626616265434,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22630835175514222,
      "num_tokens": 77464465.0,
      "step": 34700
    },
    {
      "entropy": 5.593823385238648,
      "epoch": 3.897736845060931,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035593880639991037,
      "loss": 4.9227,
      "mean_token_accuracy": 0.2214621514081955,
      "num_tokens": 77474944.0,
      "step": 34705
    },
    {
      "entropy": 5.586379480361939,
      "epoch": 3.8982984219688888,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035590134406004336,
      "loss": 4.8453,
      "mean_token_accuracy": 0.23688877671957015,
      "num_tokens": 77485389.0,
      "step": 34710
    },
    {
      "entropy": 5.651964521408081,
      "epoch": 3.898859998876846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003558638791442463,
      "loss": 4.9025,
      "mean_token_accuracy": 0.2276395708322525,
      "num_tokens": 77496834.0,
      "step": 34715
    },
    {
      "entropy": 5.579259586334229,
      "epoch": 3.8994215757848036,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003558264116537119,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22775835692882537,
      "num_tokens": 77508123.0,
      "step": 34720
    },
    {
      "entropy": 5.6063940048217775,
      "epoch": 3.8999831526927613,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003557889415896332,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22734274864196777,
      "num_tokens": 77519599.0,
      "step": 34725
    },
    {
      "entropy": 5.597504806518555,
      "epoch": 3.900544729600719,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035575146895320327,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22355655580759048,
      "num_tokens": 77532241.0,
      "step": 34730
    },
    {
      "entropy": 5.710771226882935,
      "epoch": 3.9011063065086766,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003557139937456152,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2243884727358818,
      "num_tokens": 77542702.0,
      "step": 34735
    },
    {
      "entropy": 5.625005197525025,
      "epoch": 3.901667883416634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035567651596806236,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2220536395907402,
      "num_tokens": 77554235.0,
      "step": 34740
    },
    {
      "entropy": 5.630262184143066,
      "epoch": 3.9022294603245915,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035563903562173786,
      "loss": 4.9579,
      "mean_token_accuracy": 0.23084431886672974,
      "num_tokens": 77565704.0,
      "step": 34745
    },
    {
      "entropy": 5.622745609283447,
      "epoch": 3.902791037232549,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035560155270783527,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22374411821365356,
      "num_tokens": 77576524.0,
      "step": 34750
    },
    {
      "entropy": 5.665405225753784,
      "epoch": 3.9033526141405064,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003555640672275479,
      "loss": 4.9284,
      "mean_token_accuracy": 0.23046302050352097,
      "num_tokens": 77587292.0,
      "step": 34755
    },
    {
      "entropy": 5.649006128311157,
      "epoch": 3.903914191048464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003555265791820695,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21662099063396453,
      "num_tokens": 77597958.0,
      "step": 34760
    },
    {
      "entropy": 5.699815559387207,
      "epoch": 3.9044757679564217,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003554890885725935,
      "loss": 5.052,
      "mean_token_accuracy": 0.22027798742055893,
      "num_tokens": 77609264.0,
      "step": 34765
    },
    {
      "entropy": 5.732881641387939,
      "epoch": 3.905037344864379,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003554515954003137,
      "loss": 5.0309,
      "mean_token_accuracy": 0.22916645109653472,
      "num_tokens": 77621224.0,
      "step": 34770
    },
    {
      "entropy": 5.66120867729187,
      "epoch": 3.9055989217723366,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003554140996664239,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2210492104291916,
      "num_tokens": 77632009.0,
      "step": 34775
    },
    {
      "entropy": 5.622309398651123,
      "epoch": 3.9061604986802942,
      "grad_norm": 1.15625,
      "learning_rate": 0.000355376601372118,
      "loss": 4.946,
      "mean_token_accuracy": 0.22363899946212767,
      "num_tokens": 77643094.0,
      "step": 34780
    },
    {
      "entropy": 5.660679721832276,
      "epoch": 3.906722075588252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003553391005185899,
      "loss": 4.992,
      "mean_token_accuracy": 0.23107970505952835,
      "num_tokens": 77654049.0,
      "step": 34785
    },
    {
      "entropy": 5.628866958618164,
      "epoch": 3.9072836524962096,
      "grad_norm": 1.125,
      "learning_rate": 0.0003553015971070338,
      "loss": 4.9087,
      "mean_token_accuracy": 0.2288317233324051,
      "num_tokens": 77664820.0,
      "step": 34790
    },
    {
      "entropy": 5.756520080566406,
      "epoch": 3.907845229404167,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003552640911386436,
      "loss": 5.0688,
      "mean_token_accuracy": 0.21874265372753143,
      "num_tokens": 77675107.0,
      "step": 34795
    },
    {
      "entropy": 5.656947374343872,
      "epoch": 3.9084068063121244,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035522658261461366,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22912580966949464,
      "num_tokens": 77685487.0,
      "step": 34800
    },
    {
      "entropy": 5.646736478805542,
      "epoch": 3.908968383220082,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003551890715361382,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22966018617153167,
      "num_tokens": 77697391.0,
      "step": 34805
    },
    {
      "entropy": 5.63563027381897,
      "epoch": 3.9095299601280393,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003551515579044116,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22597309648990632,
      "num_tokens": 77707270.0,
      "step": 34810
    },
    {
      "entropy": 5.524936532974243,
      "epoch": 3.910091537035997,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035511404172062827,
      "loss": 4.8605,
      "mean_token_accuracy": 0.2283810555934906,
      "num_tokens": 77717164.0,
      "step": 34815
    },
    {
      "entropy": 5.710162878036499,
      "epoch": 3.9106531139439547,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003550765229859828,
      "loss": 5.0222,
      "mean_token_accuracy": 0.22574121803045272,
      "num_tokens": 77728142.0,
      "step": 34820
    },
    {
      "entropy": 5.692923927307129,
      "epoch": 3.9112146908519123,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003550390017016698,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2338757872581482,
      "num_tokens": 77740162.0,
      "step": 34825
    },
    {
      "entropy": 5.645911931991577,
      "epoch": 3.91177626775987,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035500147786888387,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23565286099910737,
      "num_tokens": 77751574.0,
      "step": 34830
    },
    {
      "entropy": 5.710404014587402,
      "epoch": 3.912337844667827,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035496395148881994,
      "loss": 5.0178,
      "mean_token_accuracy": 0.22431949377059937,
      "num_tokens": 77762919.0,
      "step": 34835
    },
    {
      "entropy": 5.627382564544678,
      "epoch": 3.912899421575785,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035492642256267277,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22512372434139252,
      "num_tokens": 77774578.0,
      "step": 34840
    },
    {
      "entropy": 5.668759202957153,
      "epoch": 3.9134609984837425,
      "grad_norm": 1.125,
      "learning_rate": 0.00035488889109163725,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22292887419462204,
      "num_tokens": 77784872.0,
      "step": 34845
    },
    {
      "entropy": 5.611009836196899,
      "epoch": 3.9140225753916997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003548513570769085,
      "loss": 4.9361,
      "mean_token_accuracy": 0.22652028053998946,
      "num_tokens": 77796159.0,
      "step": 34850
    },
    {
      "entropy": 5.589690113067627,
      "epoch": 3.9145841522996574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003548138205196816,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2319112613797188,
      "num_tokens": 77805779.0,
      "step": 34855
    },
    {
      "entropy": 5.668914079666138,
      "epoch": 3.915145729207615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003547762814211517,
      "loss": 5.0111,
      "mean_token_accuracy": 0.2302881211042404,
      "num_tokens": 77818102.0,
      "step": 34860
    },
    {
      "entropy": 5.661305904388428,
      "epoch": 3.9157073061155723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000354738739782514,
      "loss": 5.0171,
      "mean_token_accuracy": 0.2189672350883484,
      "num_tokens": 77830107.0,
      "step": 34865
    },
    {
      "entropy": 5.685315275192261,
      "epoch": 3.91626888302353,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000354701195604964,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22537013292312622,
      "num_tokens": 77841388.0,
      "step": 34870
    },
    {
      "entropy": 5.655983400344849,
      "epoch": 3.9168304599314876,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003546636488896969,
      "loss": 5.0142,
      "mean_token_accuracy": 0.2217336431145668,
      "num_tokens": 77853418.0,
      "step": 34875
    },
    {
      "entropy": 5.712381076812744,
      "epoch": 3.9173920368394453,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003546260996379083,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22389772534370422,
      "num_tokens": 77865019.0,
      "step": 34880
    },
    {
      "entropy": 5.675012588500977,
      "epoch": 3.917953613747403,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000354588547850794,
      "loss": 4.9209,
      "mean_token_accuracy": 0.23291132748126983,
      "num_tokens": 77876546.0,
      "step": 34885
    },
    {
      "entropy": 5.707903337478638,
      "epoch": 3.91851519065536,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003545509935295493,
      "loss": 5.025,
      "mean_token_accuracy": 0.22603488564491273,
      "num_tokens": 77888081.0,
      "step": 34890
    },
    {
      "entropy": 5.579881095886231,
      "epoch": 3.919076767563318,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003545134366753701,
      "loss": 4.8653,
      "mean_token_accuracy": 0.23291546553373338,
      "num_tokens": 77900106.0,
      "step": 34895
    },
    {
      "entropy": 5.619477796554565,
      "epoch": 3.9196383444712755,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003544758772894523,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2333684042096138,
      "num_tokens": 77910452.0,
      "step": 34900
    },
    {
      "entropy": 5.706148481369018,
      "epoch": 3.9201999213792327,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035443831537299177,
      "loss": 5.0113,
      "mean_token_accuracy": 0.22581467032432556,
      "num_tokens": 77921826.0,
      "step": 34905
    },
    {
      "entropy": 5.767303562164306,
      "epoch": 3.9207614982871903,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003544007509271844,
      "loss": 5.0695,
      "mean_token_accuracy": 0.22148324847221373,
      "num_tokens": 77933909.0,
      "step": 34910
    },
    {
      "entropy": 5.774835205078125,
      "epoch": 3.921323075195148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035436318395322636,
      "loss": 5.0408,
      "mean_token_accuracy": 0.22365380823612213,
      "num_tokens": 77945232.0,
      "step": 34915
    },
    {
      "entropy": 5.661491107940674,
      "epoch": 3.9218846521031057,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003543256144523137,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22614498436450958,
      "num_tokens": 77957305.0,
      "step": 34920
    },
    {
      "entropy": 5.605467653274536,
      "epoch": 3.9224462290110633,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003542880424256428,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23277090787887572,
      "num_tokens": 77968036.0,
      "step": 34925
    },
    {
      "entropy": 5.6246631145477295,
      "epoch": 3.9230078059190205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003542504678744099,
      "loss": 4.9157,
      "mean_token_accuracy": 0.229895880818367,
      "num_tokens": 77978826.0,
      "step": 34930
    },
    {
      "entropy": 5.634543418884277,
      "epoch": 3.923569382826978,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003542128907998113,
      "loss": 4.9388,
      "mean_token_accuracy": 0.22603039890527726,
      "num_tokens": 77989820.0,
      "step": 34935
    },
    {
      "entropy": 5.659372329711914,
      "epoch": 3.924130959734936,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003541753112030435,
      "loss": 4.9784,
      "mean_token_accuracy": 0.22368978559970856,
      "num_tokens": 78000114.0,
      "step": 34940
    },
    {
      "entropy": 5.599851655960083,
      "epoch": 3.924692536642893,
      "grad_norm": 1.28125,
      "learning_rate": 0.00035413772908530317,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23050281405448914,
      "num_tokens": 78010680.0,
      "step": 34945
    },
    {
      "entropy": 5.645400524139404,
      "epoch": 3.9252541135508507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003541001444477869,
      "loss": 4.9234,
      "mean_token_accuracy": 0.2230563923716545,
      "num_tokens": 78022172.0,
      "step": 34950
    },
    {
      "entropy": 5.657908391952515,
      "epoch": 3.9258156904588084,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035406255729169124,
      "loss": 4.9665,
      "mean_token_accuracy": 0.21412609666585922,
      "num_tokens": 78033535.0,
      "step": 34955
    },
    {
      "entropy": 5.698249197006225,
      "epoch": 3.9263772673667656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003540249676182131,
      "loss": 4.9655,
      "mean_token_accuracy": 0.21980341970920564,
      "num_tokens": 78044555.0,
      "step": 34960
    },
    {
      "entropy": 5.649522352218628,
      "epoch": 3.9269388442747233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035398737542854936,
      "loss": 4.9108,
      "mean_token_accuracy": 0.2345434159040451,
      "num_tokens": 78056078.0,
      "step": 34965
    },
    {
      "entropy": 5.661337566375733,
      "epoch": 3.927500421182681,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003539497807238969,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22717979103326796,
      "num_tokens": 78068324.0,
      "step": 34970
    },
    {
      "entropy": 5.610949897766114,
      "epoch": 3.9280619980906386,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035391218350545287,
      "loss": 4.9035,
      "mean_token_accuracy": 0.2256702482700348,
      "num_tokens": 78080013.0,
      "step": 34975
    },
    {
      "entropy": 5.646019077301025,
      "epoch": 3.9286235749985963,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035387458377441434,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2306706964969635,
      "num_tokens": 78091494.0,
      "step": 34980
    },
    {
      "entropy": 5.671958351135254,
      "epoch": 3.9291851519065535,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003538369815319784,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22717946469783784,
      "num_tokens": 78102037.0,
      "step": 34985
    },
    {
      "entropy": 5.658925580978393,
      "epoch": 3.929746728814511,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035379937677934245,
      "loss": 5.0257,
      "mean_token_accuracy": 0.22027140259742736,
      "num_tokens": 78112651.0,
      "step": 34990
    },
    {
      "entropy": 5.685484600067139,
      "epoch": 3.930308305722469,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003537617695177037,
      "loss": 5.0136,
      "mean_token_accuracy": 0.22483717203140258,
      "num_tokens": 78123273.0,
      "step": 34995
    },
    {
      "entropy": 5.669126367568969,
      "epoch": 3.930869882630426,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035372415974825974,
      "loss": 5.0197,
      "mean_token_accuracy": 0.2212323412299156,
      "num_tokens": 78133868.0,
      "step": 35000
    },
    {
      "entropy": 5.747037124633789,
      "epoch": 3.9314314595383837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035368654747220795,
      "loss": 5.034,
      "mean_token_accuracy": 0.2156943514943123,
      "num_tokens": 78145001.0,
      "step": 35005
    },
    {
      "entropy": 5.752141857147217,
      "epoch": 3.9319930364463413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035364893269074606,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2192138358950615,
      "num_tokens": 78155699.0,
      "step": 35010
    },
    {
      "entropy": 5.623638248443603,
      "epoch": 3.932554613354299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035361131540507157,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23188912570476533,
      "num_tokens": 78165941.0,
      "step": 35015
    },
    {
      "entropy": 5.616686010360718,
      "epoch": 3.9331161902622567,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003535736956163823,
      "loss": 4.9816,
      "mean_token_accuracy": 0.23329856842756272,
      "num_tokens": 78177049.0,
      "step": 35020
    },
    {
      "entropy": 5.678004312515259,
      "epoch": 3.933677767170214,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035353607332587613,
      "loss": 5.064,
      "mean_token_accuracy": 0.22186149805784225,
      "num_tokens": 78188834.0,
      "step": 35025
    },
    {
      "entropy": 5.736487245559692,
      "epoch": 3.9342393440781716,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003534984485347509,
      "loss": 5.0011,
      "mean_token_accuracy": 0.2290984332561493,
      "num_tokens": 78199847.0,
      "step": 35030
    },
    {
      "entropy": 5.53919095993042,
      "epoch": 3.934800920986129,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003534608212442046,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2304687425494194,
      "num_tokens": 78210747.0,
      "step": 35035
    },
    {
      "entropy": 5.623755311965942,
      "epoch": 3.9353624978940864,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035342319145543533,
      "loss": 5.0007,
      "mean_token_accuracy": 0.2289229080080986,
      "num_tokens": 78221711.0,
      "step": 35040
    },
    {
      "entropy": 5.7520441055297855,
      "epoch": 3.935924074802044,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035338555916964126,
      "loss": 5.1073,
      "mean_token_accuracy": 0.21687812805175782,
      "num_tokens": 78233957.0,
      "step": 35045
    },
    {
      "entropy": 5.621125745773315,
      "epoch": 3.9364856517100018,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035334792438802056,
      "loss": 4.995,
      "mean_token_accuracy": 0.2249815508723259,
      "num_tokens": 78245403.0,
      "step": 35050
    },
    {
      "entropy": 5.700217342376709,
      "epoch": 3.937047228617959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003533102871117716,
      "loss": 4.949,
      "mean_token_accuracy": 0.22328965067863465,
      "num_tokens": 78256062.0,
      "step": 35055
    },
    {
      "entropy": 5.682314014434814,
      "epoch": 3.9376088055259166,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003532726473420927,
      "loss": 4.9846,
      "mean_token_accuracy": 0.226555897295475,
      "num_tokens": 78265895.0,
      "step": 35060
    },
    {
      "entropy": 5.697548913955688,
      "epoch": 3.9381703824338743,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003532350050801823,
      "loss": 4.9502,
      "mean_token_accuracy": 0.2241071194410324,
      "num_tokens": 78276752.0,
      "step": 35065
    },
    {
      "entropy": 5.686281347274781,
      "epoch": 3.938731959341832,
      "grad_norm": 1.234375,
      "learning_rate": 0.00035319736032723905,
      "loss": 4.9805,
      "mean_token_accuracy": 0.23192053735256196,
      "num_tokens": 78288357.0,
      "step": 35070
    },
    {
      "entropy": 5.664483594894409,
      "epoch": 3.9392935362497896,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003531597130844616,
      "loss": 4.9364,
      "mean_token_accuracy": 0.23247686326503753,
      "num_tokens": 78299558.0,
      "step": 35075
    },
    {
      "entropy": 5.754700374603272,
      "epoch": 3.939855113157747,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035312206335304843,
      "loss": 4.9691,
      "mean_token_accuracy": 0.22530128806829453,
      "num_tokens": 78311443.0,
      "step": 35080
    },
    {
      "entropy": 5.718343782424927,
      "epoch": 3.9404166900657045,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035308441113419855,
      "loss": 5.053,
      "mean_token_accuracy": 0.22238852381706237,
      "num_tokens": 78322575.0,
      "step": 35085
    },
    {
      "entropy": 5.642999982833862,
      "epoch": 3.940978266973662,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035304675642911074,
      "loss": 4.9515,
      "mean_token_accuracy": 0.2249840721487999,
      "num_tokens": 78333583.0,
      "step": 35090
    },
    {
      "entropy": 5.6465901851654055,
      "epoch": 3.9415398438816194,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003530090992389838,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22829961776733398,
      "num_tokens": 78343339.0,
      "step": 35095
    },
    {
      "entropy": 5.704099369049072,
      "epoch": 3.942101420789577,
      "grad_norm": 1.203125,
      "learning_rate": 0.000352971439565017,
      "loss": 4.926,
      "mean_token_accuracy": 0.22666538059711455,
      "num_tokens": 78354613.0,
      "step": 35100
    },
    {
      "entropy": 5.626445770263672,
      "epoch": 3.9426629976975347,
      "grad_norm": 1.265625,
      "learning_rate": 0.00035293377740840934,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2290654733777046,
      "num_tokens": 78365105.0,
      "step": 35105
    },
    {
      "entropy": 5.579329538345337,
      "epoch": 3.9432245746054924,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035289611277036,
      "loss": 4.8457,
      "mean_token_accuracy": 0.23825504928827285,
      "num_tokens": 78375311.0,
      "step": 35110
    },
    {
      "entropy": 5.598333406448364,
      "epoch": 3.94378615151345,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003528584456520681,
      "loss": 4.8981,
      "mean_token_accuracy": 0.23469826728105544,
      "num_tokens": 78386653.0,
      "step": 35115
    },
    {
      "entropy": 5.539033079147339,
      "epoch": 3.9443477284214072,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035282077605473317,
      "loss": 4.905,
      "mean_token_accuracy": 0.22940988838672638,
      "num_tokens": 78397296.0,
      "step": 35120
    },
    {
      "entropy": 5.655440378189087,
      "epoch": 3.944909305329365,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003527831039795545,
      "loss": 4.9042,
      "mean_token_accuracy": 0.23389618396759032,
      "num_tokens": 78408825.0,
      "step": 35125
    },
    {
      "entropy": 5.588826322555542,
      "epoch": 3.9454708822373226,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003527454294277316,
      "loss": 4.8909,
      "mean_token_accuracy": 0.23430146872997284,
      "num_tokens": 78419895.0,
      "step": 35130
    },
    {
      "entropy": 5.583400821685791,
      "epoch": 3.94603245914528,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003527077524004642,
      "loss": 4.8258,
      "mean_token_accuracy": 0.23154741525650024,
      "num_tokens": 78430111.0,
      "step": 35135
    },
    {
      "entropy": 5.636906719207763,
      "epoch": 3.9465940360532374,
      "grad_norm": 1.125,
      "learning_rate": 0.00035267007289895176,
      "loss": 4.9962,
      "mean_token_accuracy": 0.2219121426343918,
      "num_tokens": 78441778.0,
      "step": 35140
    },
    {
      "entropy": 5.642813110351563,
      "epoch": 3.947155612961195,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035263239092439417,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2332245409488678,
      "num_tokens": 78451930.0,
      "step": 35145
    },
    {
      "entropy": 5.5847025394439695,
      "epoch": 3.9477171898691523,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035259470647799104,
      "loss": 4.9064,
      "mean_token_accuracy": 0.2276376113295555,
      "num_tokens": 78463002.0,
      "step": 35150
    },
    {
      "entropy": 5.589416790008545,
      "epoch": 3.94827876677711,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035255701956094236,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22389927506446838,
      "num_tokens": 78473573.0,
      "step": 35155
    },
    {
      "entropy": 5.645166873931885,
      "epoch": 3.9488403436850676,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035251933017444814,
      "loss": 4.9845,
      "mean_token_accuracy": 0.22570974975824357,
      "num_tokens": 78485336.0,
      "step": 35160
    },
    {
      "entropy": 5.68372483253479,
      "epoch": 3.9494019205930253,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035248163831970837,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22945573925971985,
      "num_tokens": 78497801.0,
      "step": 35165
    },
    {
      "entropy": 5.611038446426392,
      "epoch": 3.949963497500983,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035244394399792304,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22712777256965638,
      "num_tokens": 78509288.0,
      "step": 35170
    },
    {
      "entropy": 5.643189716339111,
      "epoch": 3.95052507440894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003524062472102927,
      "loss": 4.8735,
      "mean_token_accuracy": 0.23039824664592742,
      "num_tokens": 78519871.0,
      "step": 35175
    },
    {
      "entropy": 5.641613340377807,
      "epoch": 3.951086651316898,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003523685479580174,
      "loss": 4.9795,
      "mean_token_accuracy": 0.22204203456640242,
      "num_tokens": 78530287.0,
      "step": 35180
    },
    {
      "entropy": 5.63554573059082,
      "epoch": 3.9516482282248555,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003523308462422974,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22474739104509353,
      "num_tokens": 78542341.0,
      "step": 35185
    },
    {
      "entropy": 5.573857069015503,
      "epoch": 3.9522098051328127,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003522931420643334,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23602579087018966,
      "num_tokens": 78552533.0,
      "step": 35190
    },
    {
      "entropy": 5.655317115783691,
      "epoch": 3.9527713820407704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003522554354253257,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22755298763513565,
      "num_tokens": 78564007.0,
      "step": 35195
    },
    {
      "entropy": 5.632915353775024,
      "epoch": 3.953332958948728,
      "grad_norm": 1.25,
      "learning_rate": 0.00035221772632647503,
      "loss": 4.9877,
      "mean_token_accuracy": 0.2217197000980377,
      "num_tokens": 78574015.0,
      "step": 35200
    },
    {
      "entropy": 5.699917984008789,
      "epoch": 3.9538945358566857,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035218001476898194,
      "loss": 5.0257,
      "mean_token_accuracy": 0.22562156021595,
      "num_tokens": 78586230.0,
      "step": 35205
    },
    {
      "entropy": 5.6437540531158445,
      "epoch": 3.9544561127646434,
      "grad_norm": 1.125,
      "learning_rate": 0.0003521423007540472,
      "loss": 4.9766,
      "mean_token_accuracy": 0.22955967038869857,
      "num_tokens": 78596819.0,
      "step": 35210
    },
    {
      "entropy": 5.629056072235107,
      "epoch": 3.9550176896726006,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003521045842828718,
      "loss": 4.9334,
      "mean_token_accuracy": 0.23116403371095656,
      "num_tokens": 78608788.0,
      "step": 35215
    },
    {
      "entropy": 5.659788703918457,
      "epoch": 3.9555792665805583,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003520668653566565,
      "loss": 4.9517,
      "mean_token_accuracy": 0.22564833909273146,
      "num_tokens": 78619124.0,
      "step": 35220
    },
    {
      "entropy": 5.750182104110718,
      "epoch": 3.956140843488516,
      "grad_norm": 1.125,
      "learning_rate": 0.0003520291439766024,
      "loss": 5.0342,
      "mean_token_accuracy": 0.22838675528764724,
      "num_tokens": 78630799.0,
      "step": 35225
    },
    {
      "entropy": 5.589577007293701,
      "epoch": 3.956702420396473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035199142014391034,
      "loss": 4.8871,
      "mean_token_accuracy": 0.23018446564674377,
      "num_tokens": 78641664.0,
      "step": 35230
    },
    {
      "entropy": 5.667150926589966,
      "epoch": 3.957263997304431,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035195369385978165,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22753285765647888,
      "num_tokens": 78652025.0,
      "step": 35235
    },
    {
      "entropy": 5.655558061599732,
      "epoch": 3.9578255742123885,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035191596512541747,
      "loss": 4.9379,
      "mean_token_accuracy": 0.237081079185009,
      "num_tokens": 78663539.0,
      "step": 35240
    },
    {
      "entropy": 5.623299407958984,
      "epoch": 3.9583871511203457,
      "grad_norm": 1.125,
      "learning_rate": 0.0003518782339420192,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2185833215713501,
      "num_tokens": 78674398.0,
      "step": 35245
    },
    {
      "entropy": 5.658600759506226,
      "epoch": 3.9589487280283033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035184050031078815,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2154463544487953,
      "num_tokens": 78685945.0,
      "step": 35250
    },
    {
      "entropy": 5.625453329086303,
      "epoch": 3.959510304936261,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035180276423292566,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21997968703508378,
      "num_tokens": 78696664.0,
      "step": 35255
    },
    {
      "entropy": 5.643620491027832,
      "epoch": 3.9600718818442187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003517650257096334,
      "loss": 4.8375,
      "mean_token_accuracy": 0.23632104694843292,
      "num_tokens": 78706812.0,
      "step": 35260
    },
    {
      "entropy": 5.596327400207519,
      "epoch": 3.9606334587521763,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035172728474211306,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22970400601625443,
      "num_tokens": 78716657.0,
      "step": 35265
    },
    {
      "entropy": 5.652878665924073,
      "epoch": 3.9611950356601335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035168954133156613,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2228786438703537,
      "num_tokens": 78728313.0,
      "step": 35270
    },
    {
      "entropy": 5.664108800888061,
      "epoch": 3.961756612568091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003516517954791944,
      "loss": 4.9662,
      "mean_token_accuracy": 0.2308400183916092,
      "num_tokens": 78739888.0,
      "step": 35275
    },
    {
      "entropy": 5.666188335418701,
      "epoch": 3.962318189476049,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035161404718619986,
      "loss": 4.943,
      "mean_token_accuracy": 0.22812678813934326,
      "num_tokens": 78750624.0,
      "step": 35280
    },
    {
      "entropy": 5.635391664505005,
      "epoch": 3.962879766384006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003515762964537844,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22959134876728057,
      "num_tokens": 78762506.0,
      "step": 35285
    },
    {
      "entropy": 5.670600318908692,
      "epoch": 3.9634413432919637,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003515385432831499,
      "loss": 4.9582,
      "mean_token_accuracy": 0.22470832467079163,
      "num_tokens": 78772565.0,
      "step": 35290
    },
    {
      "entropy": 5.618021535873413,
      "epoch": 3.9640029201999214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003515007876754984,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22946800291538239,
      "num_tokens": 78784107.0,
      "step": 35295
    },
    {
      "entropy": 5.692856025695801,
      "epoch": 3.964564497107879,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003514630296320323,
      "loss": 5.0087,
      "mean_token_accuracy": 0.22372142225503922,
      "num_tokens": 78796182.0,
      "step": 35300
    },
    {
      "entropy": 5.71171817779541,
      "epoch": 3.9651260740158367,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003514252691539536,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22997506707906723,
      "num_tokens": 78806620.0,
      "step": 35305
    },
    {
      "entropy": 5.636563491821289,
      "epoch": 3.965687650923794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035138750624246466,
      "loss": 4.8874,
      "mean_token_accuracy": 0.2371609166264534,
      "num_tokens": 78817034.0,
      "step": 35310
    },
    {
      "entropy": 5.684043836593628,
      "epoch": 3.9662492278317516,
      "grad_norm": 2.171875,
      "learning_rate": 0.000351349740898768,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2299385353922844,
      "num_tokens": 78828974.0,
      "step": 35315
    },
    {
      "entropy": 5.680231714248658,
      "epoch": 3.9668108047397093,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003513119731240659,
      "loss": 4.9647,
      "mean_token_accuracy": 0.22779695689678192,
      "num_tokens": 78840514.0,
      "step": 35320
    },
    {
      "entropy": 5.631758785247802,
      "epoch": 3.9673723816476665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035127420291956104,
      "loss": 4.9905,
      "mean_token_accuracy": 0.2283450946211815,
      "num_tokens": 78850839.0,
      "step": 35325
    },
    {
      "entropy": 5.653752040863037,
      "epoch": 3.967933958555624,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003512364302864559,
      "loss": 4.986,
      "mean_token_accuracy": 0.2279742032289505,
      "num_tokens": 78862931.0,
      "step": 35330
    },
    {
      "entropy": 5.663985919952393,
      "epoch": 3.968495535463582,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003511986552259533,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2270403578877449,
      "num_tokens": 78872324.0,
      "step": 35335
    },
    {
      "entropy": 5.58043360710144,
      "epoch": 3.969057112371539,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035116087773925594,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22696414589881897,
      "num_tokens": 78884075.0,
      "step": 35340
    },
    {
      "entropy": 5.54282956123352,
      "epoch": 3.9696186892794967,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003511230978275667,
      "loss": 4.8668,
      "mean_token_accuracy": 0.23055271953344345,
      "num_tokens": 78894804.0,
      "step": 35345
    },
    {
      "entropy": 5.708126592636108,
      "epoch": 3.9701802661874543,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035108531549208846,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21529026329517365,
      "num_tokens": 78906069.0,
      "step": 35350
    },
    {
      "entropy": 5.804078102111816,
      "epoch": 3.970741843095412,
      "grad_norm": 1.125,
      "learning_rate": 0.00035104753073402433,
      "loss": 5.0861,
      "mean_token_accuracy": 0.21969297975301744,
      "num_tokens": 78918490.0,
      "step": 35355
    },
    {
      "entropy": 5.696898937225342,
      "epoch": 3.9713034200033697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003510097435545774,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2212388351559639,
      "num_tokens": 78929790.0,
      "step": 35360
    },
    {
      "entropy": 5.667677402496338,
      "epoch": 3.971864996911327,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035097195395495057,
      "loss": 4.9467,
      "mean_token_accuracy": 0.23576954752206802,
      "num_tokens": 78941600.0,
      "step": 35365
    },
    {
      "entropy": 5.7254596710205075,
      "epoch": 3.9724265738192845,
      "grad_norm": 1.125,
      "learning_rate": 0.0003509341619363474,
      "loss": 5.0849,
      "mean_token_accuracy": 0.22435242533683777,
      "num_tokens": 78953764.0,
      "step": 35370
    },
    {
      "entropy": 5.700806140899658,
      "epoch": 3.972988150727242,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035089636749997095,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2280383139848709,
      "num_tokens": 78964781.0,
      "step": 35375
    },
    {
      "entropy": 5.743687009811401,
      "epoch": 3.9735497276351994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035085857064702487,
      "loss": 5.017,
      "mean_token_accuracy": 0.2213798001408577,
      "num_tokens": 78976325.0,
      "step": 35380
    },
    {
      "entropy": 5.645967435836792,
      "epoch": 3.974111304543157,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003508207713787123,
      "loss": 4.8458,
      "mean_token_accuracy": 0.2329242408275604,
      "num_tokens": 78987174.0,
      "step": 35385
    },
    {
      "entropy": 5.667509603500366,
      "epoch": 3.9746728814511147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035078296969623716,
      "loss": 4.9478,
      "mean_token_accuracy": 0.23049305528402328,
      "num_tokens": 78999195.0,
      "step": 35390
    },
    {
      "entropy": 5.531387138366699,
      "epoch": 3.9752344583590724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035074516560080273,
      "loss": 4.8044,
      "mean_token_accuracy": 0.23689245730638503,
      "num_tokens": 79010981.0,
      "step": 35395
    },
    {
      "entropy": 5.61497802734375,
      "epoch": 3.97579603526703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003507073590936129,
      "loss": 4.978,
      "mean_token_accuracy": 0.23153850585222244,
      "num_tokens": 79022059.0,
      "step": 35400
    },
    {
      "entropy": 5.779169940948487,
      "epoch": 3.9763576121749873,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035066955017587135,
      "loss": 5.1011,
      "mean_token_accuracy": 0.2260509267449379,
      "num_tokens": 79033261.0,
      "step": 35405
    },
    {
      "entropy": 5.63051962852478,
      "epoch": 3.976919189082945,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000350631738848782,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22133505940437317,
      "num_tokens": 79044518.0,
      "step": 35410
    },
    {
      "entropy": 5.630279397964477,
      "epoch": 3.9774807659909026,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003505939251135488,
      "loss": 4.9529,
      "mean_token_accuracy": 0.22076242566108703,
      "num_tokens": 79055202.0,
      "step": 35415
    },
    {
      "entropy": 5.672166109085083,
      "epoch": 3.97804234289886,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003505561089713756,
      "loss": 4.9783,
      "mean_token_accuracy": 0.2234543278813362,
      "num_tokens": 79066448.0,
      "step": 35420
    },
    {
      "entropy": 5.667074871063233,
      "epoch": 3.9786039198068175,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003505182904234667,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22453461438417435,
      "num_tokens": 79077745.0,
      "step": 35425
    },
    {
      "entropy": 5.758225631713867,
      "epoch": 3.979165496714775,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003504804694710261,
      "loss": 5.0348,
      "mean_token_accuracy": 0.21680830121040345,
      "num_tokens": 79088011.0,
      "step": 35430
    },
    {
      "entropy": 5.776441669464111,
      "epoch": 3.9797270736227324,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003504426461152581,
      "loss": 5.0612,
      "mean_token_accuracy": 0.2117028057575226,
      "num_tokens": 79098718.0,
      "step": 35435
    },
    {
      "entropy": 5.639050245285034,
      "epoch": 3.98028865053069,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003504048203573669,
      "loss": 4.9459,
      "mean_token_accuracy": 0.222758150100708,
      "num_tokens": 79108216.0,
      "step": 35440
    },
    {
      "entropy": 5.616494226455688,
      "epoch": 3.9808502274386477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003503669921985571,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22322695404291154,
      "num_tokens": 79119709.0,
      "step": 35445
    },
    {
      "entropy": 5.727732372283936,
      "epoch": 3.9814118043466054,
      "grad_norm": 1.125,
      "learning_rate": 0.000350329161640033,
      "loss": 5.0383,
      "mean_token_accuracy": 0.22627424448728561,
      "num_tokens": 79131824.0,
      "step": 35450
    },
    {
      "entropy": 5.671210479736328,
      "epoch": 3.981973381254563,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035029132868299914,
      "loss": 4.9544,
      "mean_token_accuracy": 0.22034438848495483,
      "num_tokens": 79142855.0,
      "step": 35455
    },
    {
      "entropy": 5.7263490676879885,
      "epoch": 3.9825349581625202,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035025349332866017,
      "loss": 5.0866,
      "mean_token_accuracy": 0.2160090133547783,
      "num_tokens": 79153759.0,
      "step": 35460
    },
    {
      "entropy": 5.678020858764649,
      "epoch": 3.983096535070478,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035021565557822085,
      "loss": 4.9542,
      "mean_token_accuracy": 0.22884318977594376,
      "num_tokens": 79163586.0,
      "step": 35465
    },
    {
      "entropy": 5.697628498077393,
      "epoch": 3.9836581119784356,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003501778154328858,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2163451299071312,
      "num_tokens": 79175067.0,
      "step": 35470
    },
    {
      "entropy": 5.562964963912964,
      "epoch": 3.9842196888863928,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003501399728938599,
      "loss": 4.8531,
      "mean_token_accuracy": 0.2337016850709915,
      "num_tokens": 79186039.0,
      "step": 35475
    },
    {
      "entropy": 5.693966054916382,
      "epoch": 3.9847812657943504,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035010212796234827,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2192348435521126,
      "num_tokens": 79198571.0,
      "step": 35480
    },
    {
      "entropy": 5.575818777084351,
      "epoch": 3.985342842702308,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003500642806395557,
      "loss": 4.8762,
      "mean_token_accuracy": 0.23160358965396882,
      "num_tokens": 79210253.0,
      "step": 35485
    },
    {
      "entropy": 5.598423099517822,
      "epoch": 3.9859044196102658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003500264309266873,
      "loss": 4.937,
      "mean_token_accuracy": 0.22951993644237517,
      "num_tokens": 79221611.0,
      "step": 35490
    },
    {
      "entropy": 5.661969947814941,
      "epoch": 3.9864659965182234,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034998857882494823,
      "loss": 5.0575,
      "mean_token_accuracy": 0.2170454040169716,
      "num_tokens": 79233270.0,
      "step": 35495
    },
    {
      "entropy": 5.675623226165771,
      "epoch": 3.9870275734261806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034995072433554377,
      "loss": 5.0081,
      "mean_token_accuracy": 0.21848268955945968,
      "num_tokens": 79245338.0,
      "step": 35500
    },
    {
      "entropy": 5.710799264907837,
      "epoch": 3.9875891503341383,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003499128674596792,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21942146122455597,
      "num_tokens": 79255558.0,
      "step": 35505
    },
    {
      "entropy": 5.739508628845215,
      "epoch": 3.988150727242096,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034987500819855976,
      "loss": 5.0317,
      "mean_token_accuracy": 0.23050396889448166,
      "num_tokens": 79266075.0,
      "step": 35510
    },
    {
      "entropy": 5.7291343212127686,
      "epoch": 3.988712304150053,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034983714655339115,
      "loss": 5.0293,
      "mean_token_accuracy": 0.22381548285484315,
      "num_tokens": 79277667.0,
      "step": 35515
    },
    {
      "entropy": 5.5923912525177,
      "epoch": 3.989273881058011,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003497992825253787,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2289977103471756,
      "num_tokens": 79288755.0,
      "step": 35520
    },
    {
      "entropy": 5.651436805725098,
      "epoch": 3.9898354579659685,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034976141611572817,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22377213835716248,
      "num_tokens": 79299048.0,
      "step": 35525
    },
    {
      "entropy": 5.621462726593018,
      "epoch": 3.9903970348739257,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034972354732564506,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22848345041275026,
      "num_tokens": 79310847.0,
      "step": 35530
    },
    {
      "entropy": 5.698372840881348,
      "epoch": 3.9909586117818834,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034968567615633516,
      "loss": 5.005,
      "mean_token_accuracy": 0.21875516176223755,
      "num_tokens": 79322496.0,
      "step": 35535
    },
    {
      "entropy": 5.6719554424285885,
      "epoch": 3.991520188689841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034964780260900445,
      "loss": 4.9913,
      "mean_token_accuracy": 0.22509938329458237,
      "num_tokens": 79334216.0,
      "step": 35540
    },
    {
      "entropy": 5.593036794662476,
      "epoch": 3.9920817655977987,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034960992668485875,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23670327812433242,
      "num_tokens": 79344320.0,
      "step": 35545
    },
    {
      "entropy": 5.680321884155274,
      "epoch": 3.9926433425057564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034957204838510406,
      "loss": 5.1235,
      "mean_token_accuracy": 0.21707790344953537,
      "num_tokens": 79355775.0,
      "step": 35550
    },
    {
      "entropy": 5.607705402374267,
      "epoch": 3.9932049194137136,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034953416771094645,
      "loss": 4.8695,
      "mean_token_accuracy": 0.22734489142894745,
      "num_tokens": 79366866.0,
      "step": 35555
    },
    {
      "entropy": 5.648572731018066,
      "epoch": 3.9937664963216712,
      "grad_norm": 1.140625,
      "learning_rate": 0.000349496284663592,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22108977138996125,
      "num_tokens": 79378599.0,
      "step": 35560
    },
    {
      "entropy": 5.599613094329834,
      "epoch": 3.994328073229629,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034945839924424696,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23044802248477936,
      "num_tokens": 79389283.0,
      "step": 35565
    },
    {
      "entropy": 5.656701850891113,
      "epoch": 3.994889650137586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003494205114541176,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2210273414850235,
      "num_tokens": 79400858.0,
      "step": 35570
    },
    {
      "entropy": 5.686509513854981,
      "epoch": 3.995451227045544,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034938262129441026,
      "loss": 5.0146,
      "mean_token_accuracy": 0.22219268232584,
      "num_tokens": 79411122.0,
      "step": 35575
    },
    {
      "entropy": 5.59342589378357,
      "epoch": 3.9960128039535014,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003493447287663315,
      "loss": 4.8396,
      "mean_token_accuracy": 0.23369989395141602,
      "num_tokens": 79420153.0,
      "step": 35580
    },
    {
      "entropy": 5.708611631393433,
      "epoch": 3.996574380861459,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003493068338710877,
      "loss": 5.0995,
      "mean_token_accuracy": 0.22218070924282074,
      "num_tokens": 79432964.0,
      "step": 35585
    },
    {
      "entropy": 5.651735258102417,
      "epoch": 3.9971359577694168,
      "grad_norm": 1.0,
      "learning_rate": 0.0003492689366098855,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22767164260149003,
      "num_tokens": 79445191.0,
      "step": 35590
    },
    {
      "entropy": 5.684379959106446,
      "epoch": 3.997697534677374,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034923103698393147,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22832554280757905,
      "num_tokens": 79455553.0,
      "step": 35595
    },
    {
      "entropy": 5.690669345855713,
      "epoch": 3.9982591115853316,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003491931349944325,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21954517513513566,
      "num_tokens": 79466781.0,
      "step": 35600
    },
    {
      "entropy": 5.606044292449951,
      "epoch": 3.9988206884932893,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034915523064259533,
      "loss": 4.9482,
      "mean_token_accuracy": 0.23025101870298387,
      "num_tokens": 79478656.0,
      "step": 35605
    },
    {
      "entropy": 5.639211893081665,
      "epoch": 3.9993822654012465,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003491173239296268,
      "loss": 4.9395,
      "mean_token_accuracy": 0.23091977089643478,
      "num_tokens": 79489763.0,
      "step": 35610
    },
    {
      "entropy": 5.673930215835571,
      "epoch": 3.999943842309204,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000349079414856734,
      "loss": 5.0528,
      "mean_token_accuracy": 0.2186564415693283,
      "num_tokens": 79501828.0,
      "step": 35615
    },
    {
      "entropy": 5.677290863460964,
      "epoch": 4.000449261526366,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003490415034251239,
      "loss": 4.9545,
      "mean_token_accuracy": 0.22254778941472372,
      "num_tokens": 79510316.0,
      "step": 35620
    },
    {
      "entropy": 5.598945903778076,
      "epoch": 4.001010838434324,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003490035896360036,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23194278478622438,
      "num_tokens": 79521555.0,
      "step": 35625
    },
    {
      "entropy": 5.741635274887085,
      "epoch": 4.001572415342281,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034896567349058027,
      "loss": 5.0144,
      "mean_token_accuracy": 0.2247587725520134,
      "num_tokens": 79533147.0,
      "step": 35630
    },
    {
      "entropy": 5.739795780181884,
      "epoch": 4.002133992250239,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034892775499006126,
      "loss": 5.0307,
      "mean_token_accuracy": 0.2243322253227234,
      "num_tokens": 79544695.0,
      "step": 35635
    },
    {
      "entropy": 5.580975723266602,
      "epoch": 4.002695569158196,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003488898341356538,
      "loss": 4.7995,
      "mean_token_accuracy": 0.24076153188943863,
      "num_tokens": 79555740.0,
      "step": 35640
    },
    {
      "entropy": 5.654485654830933,
      "epoch": 4.0032571460661535,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034885191092856545,
      "loss": 4.8888,
      "mean_token_accuracy": 0.23098110407590866,
      "num_tokens": 79565648.0,
      "step": 35645
    },
    {
      "entropy": 5.698060512542725,
      "epoch": 4.003818722974112,
      "grad_norm": 1.125,
      "learning_rate": 0.0003488139853700036,
      "loss": 4.9704,
      "mean_token_accuracy": 0.22977199405431747,
      "num_tokens": 79577107.0,
      "step": 35650
    },
    {
      "entropy": 5.634275484085083,
      "epoch": 4.004380299882069,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034877605746117576,
      "loss": 4.9272,
      "mean_token_accuracy": 0.2299410194158554,
      "num_tokens": 79586689.0,
      "step": 35655
    },
    {
      "entropy": 5.729523324966431,
      "epoch": 4.004941876790026,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003487381272032897,
      "loss": 5.0049,
      "mean_token_accuracy": 0.22449325621128083,
      "num_tokens": 79597517.0,
      "step": 35660
    },
    {
      "entropy": 5.688465547561646,
      "epoch": 4.005503453697984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000348700194597553,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22599893510341645,
      "num_tokens": 79609516.0,
      "step": 35665
    },
    {
      "entropy": 5.6143872261047365,
      "epoch": 4.006065030605941,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034866225964517354,
      "loss": 4.9262,
      "mean_token_accuracy": 0.23773615807294846,
      "num_tokens": 79620706.0,
      "step": 35670
    },
    {
      "entropy": 5.638884782791138,
      "epoch": 4.006626607513899,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003486243223473592,
      "loss": 4.8347,
      "mean_token_accuracy": 0.22834660857915878,
      "num_tokens": 79630897.0,
      "step": 35675
    },
    {
      "entropy": 5.6139191627502445,
      "epoch": 4.007188184421857,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034858638270531786,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22647084891796113,
      "num_tokens": 79642426.0,
      "step": 35680
    },
    {
      "entropy": 5.653180313110352,
      "epoch": 4.007749761329814,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034854844072025755,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2288523405790329,
      "num_tokens": 79653491.0,
      "step": 35685
    },
    {
      "entropy": 5.580340957641601,
      "epoch": 4.008311338237772,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034851049639338637,
      "loss": 4.846,
      "mean_token_accuracy": 0.23365572690963746,
      "num_tokens": 79665329.0,
      "step": 35690
    },
    {
      "entropy": 5.634906482696533,
      "epoch": 4.008872915145729,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003484725497259125,
      "loss": 4.8489,
      "mean_token_accuracy": 0.22975773960351945,
      "num_tokens": 79676552.0,
      "step": 35695
    },
    {
      "entropy": 5.601104211807251,
      "epoch": 4.009434492053686,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034843460071904407,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2314855307340622,
      "num_tokens": 79687359.0,
      "step": 35700
    },
    {
      "entropy": 5.635143709182739,
      "epoch": 4.0099960689616445,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034839664937398964,
      "loss": 4.8252,
      "mean_token_accuracy": 0.23611259162425996,
      "num_tokens": 79698728.0,
      "step": 35705
    },
    {
      "entropy": 5.711255216598511,
      "epoch": 4.010557645869602,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003483586956919572,
      "loss": 5.0325,
      "mean_token_accuracy": 0.22311062216758729,
      "num_tokens": 79709018.0,
      "step": 35710
    },
    {
      "entropy": 5.6875081062316895,
      "epoch": 4.011119222777559,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003483207396741556,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22571788281202315,
      "num_tokens": 79720166.0,
      "step": 35715
    },
    {
      "entropy": 5.6463884830474855,
      "epoch": 4.011680799685517,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003482827813217932,
      "loss": 4.7843,
      "mean_token_accuracy": 0.23945293128490447,
      "num_tokens": 79730467.0,
      "step": 35720
    },
    {
      "entropy": 5.669395446777344,
      "epoch": 4.012242376593474,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034824482063607856,
      "loss": 4.9612,
      "mean_token_accuracy": 0.23064793050289153,
      "num_tokens": 79740570.0,
      "step": 35725
    },
    {
      "entropy": 5.62656717300415,
      "epoch": 4.012803953501432,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003482068576182204,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22593430429697037,
      "num_tokens": 79751856.0,
      "step": 35730
    },
    {
      "entropy": 5.639906311035157,
      "epoch": 4.01336553040939,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034816889226942755,
      "loss": 4.9193,
      "mean_token_accuracy": 0.233419506251812,
      "num_tokens": 79763267.0,
      "step": 35735
    },
    {
      "entropy": 5.71825909614563,
      "epoch": 4.013927107317347,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003481309245909088,
      "loss": 5.0106,
      "mean_token_accuracy": 0.230345518887043,
      "num_tokens": 79775202.0,
      "step": 35740
    },
    {
      "entropy": 5.640640449523926,
      "epoch": 4.014488684225305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003480929545838731,
      "loss": 4.9614,
      "mean_token_accuracy": 0.22168252915143966,
      "num_tokens": 79787087.0,
      "step": 35745
    },
    {
      "entropy": 5.723457956314087,
      "epoch": 4.015050261133262,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003480549822495293,
      "loss": 5.013,
      "mean_token_accuracy": 0.22853058576583862,
      "num_tokens": 79799014.0,
      "step": 35750
    },
    {
      "entropy": 5.646758317947388,
      "epoch": 4.015611838041219,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034801700758908653,
      "loss": 4.8624,
      "mean_token_accuracy": 0.24000533819198608,
      "num_tokens": 79810455.0,
      "step": 35755
    },
    {
      "entropy": 5.7107775688171385,
      "epoch": 4.0161734149491775,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034797903060375396,
      "loss": 4.9283,
      "mean_token_accuracy": 0.23000338673591614,
      "num_tokens": 79821614.0,
      "step": 35760
    },
    {
      "entropy": 5.680928373336792,
      "epoch": 4.016734991857135,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003479410512947407,
      "loss": 5.0556,
      "mean_token_accuracy": 0.22389704883098602,
      "num_tokens": 79832418.0,
      "step": 35765
    },
    {
      "entropy": 5.6912530899047855,
      "epoch": 4.017296568765093,
      "grad_norm": 1.28125,
      "learning_rate": 0.00034790306966325606,
      "loss": 5.0255,
      "mean_token_accuracy": 0.22615324407815934,
      "num_tokens": 79844234.0,
      "step": 35770
    },
    {
      "entropy": 5.592253398895264,
      "epoch": 4.01785814567305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003478650857105095,
      "loss": 4.8737,
      "mean_token_accuracy": 0.2333102524280548,
      "num_tokens": 79855284.0,
      "step": 35775
    },
    {
      "entropy": 5.60810399055481,
      "epoch": 4.018419722581007,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003478270994377102,
      "loss": 4.8544,
      "mean_token_accuracy": 0.2276845932006836,
      "num_tokens": 79866169.0,
      "step": 35780
    },
    {
      "entropy": 5.722963047027588,
      "epoch": 4.018981299488965,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034778911084606793,
      "loss": 5.0265,
      "mean_token_accuracy": 0.22111037671566008,
      "num_tokens": 79878044.0,
      "step": 35785
    },
    {
      "entropy": 5.644456672668457,
      "epoch": 4.0195428763969225,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003477511199367922,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23736108094453812,
      "num_tokens": 79888859.0,
      "step": 35790
    },
    {
      "entropy": 5.721826601028442,
      "epoch": 4.02010445330488,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034771312671109256,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2344800978899002,
      "num_tokens": 79899886.0,
      "step": 35795
    },
    {
      "entropy": 5.701833820343017,
      "epoch": 4.020666030212838,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003476751311701787,
      "loss": 4.9679,
      "mean_token_accuracy": 0.23445323705673218,
      "num_tokens": 79911752.0,
      "step": 35800
    },
    {
      "entropy": 5.63209638595581,
      "epoch": 4.021227607120795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003476371333152605,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22705877125263213,
      "num_tokens": 79923076.0,
      "step": 35805
    },
    {
      "entropy": 5.688178205490113,
      "epoch": 4.021789184028753,
      "grad_norm": 1.25,
      "learning_rate": 0.00034759913314754776,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22591314017772673,
      "num_tokens": 79933866.0,
      "step": 35810
    },
    {
      "entropy": 5.678562593460083,
      "epoch": 4.02235076093671,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034756113066825045,
      "loss": 4.9204,
      "mean_token_accuracy": 0.23421999663114548,
      "num_tokens": 79946046.0,
      "step": 35815
    },
    {
      "entropy": 5.6273411273956295,
      "epoch": 4.022912337844668,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003475231258785787,
      "loss": 4.8973,
      "mean_token_accuracy": 0.23118135929107667,
      "num_tokens": 79957887.0,
      "step": 35820
    },
    {
      "entropy": 5.614711570739746,
      "epoch": 4.023473914752626,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003474851187797424,
      "loss": 4.8119,
      "mean_token_accuracy": 0.2348991885781288,
      "num_tokens": 79968101.0,
      "step": 35825
    },
    {
      "entropy": 5.672719192504883,
      "epoch": 4.024035491660583,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034744710937295185,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22466988116502762,
      "num_tokens": 79978176.0,
      "step": 35830
    },
    {
      "entropy": 5.660128450393676,
      "epoch": 4.02459706856854,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00034740909765941725,
      "loss": 4.9082,
      "mean_token_accuracy": 0.2270583212375641,
      "num_tokens": 79990324.0,
      "step": 35835
    },
    {
      "entropy": 5.688468885421753,
      "epoch": 4.025158645476498,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003473710836403489,
      "loss": 4.939,
      "mean_token_accuracy": 0.22525408416986464,
      "num_tokens": 80001898.0,
      "step": 35840
    },
    {
      "entropy": 5.624190902709961,
      "epoch": 4.0257202223844555,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003473330673169572,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22196489125490187,
      "num_tokens": 80012072.0,
      "step": 35845
    },
    {
      "entropy": 5.596902465820312,
      "epoch": 4.026281799292413,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003472950486904525,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23015639930963516,
      "num_tokens": 80022718.0,
      "step": 35850
    },
    {
      "entropy": 5.73530855178833,
      "epoch": 4.026843376200371,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003472570277620455,
      "loss": 4.9758,
      "mean_token_accuracy": 0.2279218092560768,
      "num_tokens": 80033899.0,
      "step": 35855
    },
    {
      "entropy": 5.769401359558105,
      "epoch": 4.027404953108328,
      "grad_norm": 1.125,
      "learning_rate": 0.0003472190045329466,
      "loss": 4.9713,
      "mean_token_accuracy": 0.22567844539880752,
      "num_tokens": 80045398.0,
      "step": 35860
    },
    {
      "entropy": 5.722082424163818,
      "epoch": 4.027966530016286,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003471809790043667,
      "loss": 4.9903,
      "mean_token_accuracy": 0.23100190311670304,
      "num_tokens": 80055186.0,
      "step": 35865
    },
    {
      "entropy": 5.63487982749939,
      "epoch": 4.028528106924243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034714295117751635,
      "loss": 4.9556,
      "mean_token_accuracy": 0.23884495943784714,
      "num_tokens": 80065545.0,
      "step": 35870
    },
    {
      "entropy": 5.6601167678833,
      "epoch": 4.029089683832201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034710492105360653,
      "loss": 4.9963,
      "mean_token_accuracy": 0.2273005798459053,
      "num_tokens": 80077451.0,
      "step": 35875
    },
    {
      "entropy": 5.617845344543457,
      "epoch": 4.029651260740159,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003470668886338481,
      "loss": 4.8694,
      "mean_token_accuracy": 0.23218512386083603,
      "num_tokens": 80088215.0,
      "step": 35880
    },
    {
      "entropy": 5.563166332244873,
      "epoch": 4.030212837648116,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034702885391945184,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23206229954957963,
      "num_tokens": 80098393.0,
      "step": 35885
    },
    {
      "entropy": 5.639608335494995,
      "epoch": 4.030774414556073,
      "grad_norm": 1.125,
      "learning_rate": 0.000346990816911629,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22473621964454651,
      "num_tokens": 80110183.0,
      "step": 35890
    },
    {
      "entropy": 5.600053977966309,
      "epoch": 4.031335991464031,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003469527776115906,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22997664511203766,
      "num_tokens": 80120954.0,
      "step": 35895
    },
    {
      "entropy": 5.620555019378662,
      "epoch": 4.031897568371988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034691473602054793,
      "loss": 4.8245,
      "mean_token_accuracy": 0.2339925065636635,
      "num_tokens": 80131607.0,
      "step": 35900
    },
    {
      "entropy": 5.568237543106079,
      "epoch": 4.032459145279946,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003468766921397121,
      "loss": 4.869,
      "mean_token_accuracy": 0.2319898173213005,
      "num_tokens": 80141700.0,
      "step": 35905
    },
    {
      "entropy": 5.630965948104858,
      "epoch": 4.033020722187904,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003468386459702946,
      "loss": 4.9036,
      "mean_token_accuracy": 0.2287917196750641,
      "num_tokens": 80151776.0,
      "step": 35910
    },
    {
      "entropy": 5.673248767852783,
      "epoch": 4.033582299095861,
      "grad_norm": 1.125,
      "learning_rate": 0.00034680059751350664,
      "loss": 4.9554,
      "mean_token_accuracy": 0.22415705919265747,
      "num_tokens": 80162838.0,
      "step": 35915
    },
    {
      "entropy": 5.6633766174316404,
      "epoch": 4.034143876003819,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034676254677055984,
      "loss": 4.9525,
      "mean_token_accuracy": 0.22750553637742996,
      "num_tokens": 80174104.0,
      "step": 35920
    },
    {
      "entropy": 5.647879505157471,
      "epoch": 4.034705452911776,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034672449374266575,
      "loss": 5.0004,
      "mean_token_accuracy": 0.22384635508060455,
      "num_tokens": 80185015.0,
      "step": 35925
    },
    {
      "entropy": 5.687137222290039,
      "epoch": 4.0352670298197335,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003466864384310359,
      "loss": 4.8752,
      "mean_token_accuracy": 0.23709957003593446,
      "num_tokens": 80196893.0,
      "step": 35930
    },
    {
      "entropy": 5.752750492095947,
      "epoch": 4.035828606727692,
      "grad_norm": 1.125,
      "learning_rate": 0.00034664838083688223,
      "loss": 4.9665,
      "mean_token_accuracy": 0.23025454133749007,
      "num_tokens": 80209910.0,
      "step": 35935
    },
    {
      "entropy": 5.628735065460205,
      "epoch": 4.036390183635649,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034661032096141615,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22587931603193284,
      "num_tokens": 80219780.0,
      "step": 35940
    },
    {
      "entropy": 5.688802242279053,
      "epoch": 4.036951760543606,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003465722588058497,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22450220435857773,
      "num_tokens": 80231316.0,
      "step": 35945
    },
    {
      "entropy": 5.70683856010437,
      "epoch": 4.037513337451564,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034653419437139475,
      "loss": 4.9773,
      "mean_token_accuracy": 0.2262989416718483,
      "num_tokens": 80242622.0,
      "step": 35950
    },
    {
      "entropy": 5.735695123672485,
      "epoch": 4.038074914359521,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003464961276592634,
      "loss": 4.9546,
      "mean_token_accuracy": 0.23459222614765168,
      "num_tokens": 80255357.0,
      "step": 35955
    },
    {
      "entropy": 5.715416145324707,
      "epoch": 4.0386364912674795,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034645805867066756,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22969394624233247,
      "num_tokens": 80267430.0,
      "step": 35960
    },
    {
      "entropy": 5.648274087905884,
      "epoch": 4.039198068175437,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034641998740681933,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2315468192100525,
      "num_tokens": 80277820.0,
      "step": 35965
    },
    {
      "entropy": 5.732408809661865,
      "epoch": 4.039759645083394,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034638191386893115,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21622256338596343,
      "num_tokens": 80289559.0,
      "step": 35970
    },
    {
      "entropy": 5.668887662887573,
      "epoch": 4.040321221991352,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034634383805821504,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22191197276115418,
      "num_tokens": 80300452.0,
      "step": 35975
    },
    {
      "entropy": 5.672245693206787,
      "epoch": 4.040882798899309,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034630575997588353,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22569430470466614,
      "num_tokens": 80311403.0,
      "step": 35980
    },
    {
      "entropy": 5.652348184585572,
      "epoch": 4.0414443758072665,
      "grad_norm": 1.203125,
      "learning_rate": 0.000346267679623149,
      "loss": 4.8667,
      "mean_token_accuracy": 0.23041878193616866,
      "num_tokens": 80323245.0,
      "step": 35985
    },
    {
      "entropy": 5.636755084991455,
      "epoch": 4.042005952715225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034622959700122383,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22211361229419707,
      "num_tokens": 80333666.0,
      "step": 35990
    },
    {
      "entropy": 5.69563856124878,
      "epoch": 4.042567529623182,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003461915121113206,
      "loss": 4.8795,
      "mean_token_accuracy": 0.23878561407327653,
      "num_tokens": 80345291.0,
      "step": 35995
    },
    {
      "entropy": 5.595835447311401,
      "epoch": 4.04312910653114,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003461534249546521,
      "loss": 4.8588,
      "mean_token_accuracy": 0.23141447007656096,
      "num_tokens": 80355612.0,
      "step": 36000
    },
    {
      "epoch": 4.04312910653114,
      "eval_entropy": 5.374812192328972,
      "eval_loss": 5.064660549163818,
      "eval_mean_token_accuracy": 0.22996986190686836,
      "eval_num_tokens": 80355612.0,
      "eval_runtime": 18.4994,
      "eval_samples_per_second": 1676.434,
      "eval_steps_per_second": 209.574,
      "step": 36000
    },
    {
      "entropy": 5.581135559082031,
      "epoch": 4.043690683439097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000346115335532431,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23287715762853622,
      "num_tokens": 80365952.0,
      "step": 36005
    },
    {
      "entropy": 5.689744520187378,
      "epoch": 4.044252260347054,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034607724384587006,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2269822061061859,
      "num_tokens": 80376962.0,
      "step": 36010
    },
    {
      "entropy": 5.734729480743408,
      "epoch": 4.044813837255012,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034603914989618197,
      "loss": 5.0119,
      "mean_token_accuracy": 0.22458259612321854,
      "num_tokens": 80388671.0,
      "step": 36015
    },
    {
      "entropy": 5.709034061431884,
      "epoch": 4.04537541416297,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003460010536845799,
      "loss": 4.9562,
      "mean_token_accuracy": 0.23354559689760207,
      "num_tokens": 80399215.0,
      "step": 36020
    },
    {
      "entropy": 5.668269681930542,
      "epoch": 4.045936991070927,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003459629552122767,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22271010279655457,
      "num_tokens": 80410590.0,
      "step": 36025
    },
    {
      "entropy": 5.6326741695404055,
      "epoch": 4.046498567978885,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003459248544804855,
      "loss": 4.864,
      "mean_token_accuracy": 0.231650909781456,
      "num_tokens": 80422085.0,
      "step": 36030
    },
    {
      "entropy": 5.614691591262817,
      "epoch": 4.047060144886842,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003458867514904194,
      "loss": 4.8339,
      "mean_token_accuracy": 0.2370166674256325,
      "num_tokens": 80434389.0,
      "step": 36035
    },
    {
      "entropy": 5.6468305587768555,
      "epoch": 4.047621721794799,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003458486462432916,
      "loss": 4.9364,
      "mean_token_accuracy": 0.23328206241130828,
      "num_tokens": 80445073.0,
      "step": 36040
    },
    {
      "entropy": 5.636676836013794,
      "epoch": 4.0481832987027575,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034581053874031554,
      "loss": 4.876,
      "mean_token_accuracy": 0.23268700987100602,
      "num_tokens": 80455206.0,
      "step": 36045
    },
    {
      "entropy": 5.634824419021607,
      "epoch": 4.048744875610715,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003457724289827044,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2336013972759247,
      "num_tokens": 80467707.0,
      "step": 36050
    },
    {
      "entropy": 5.642119455337524,
      "epoch": 4.049306452518673,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003457343169716716,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23074193000793458,
      "num_tokens": 80478493.0,
      "step": 36055
    },
    {
      "entropy": 5.613879442214966,
      "epoch": 4.04986802942663,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034569620270843087,
      "loss": 4.9284,
      "mean_token_accuracy": 0.22807333171367644,
      "num_tokens": 80490057.0,
      "step": 36060
    },
    {
      "entropy": 5.666419744491577,
      "epoch": 4.050429606334587,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034565808619419554,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22616769075393678,
      "num_tokens": 80500590.0,
      "step": 36065
    },
    {
      "entropy": 5.665709161758423,
      "epoch": 4.050991183242545,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003456199674301793,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22987350076436996,
      "num_tokens": 80510836.0,
      "step": 36070
    },
    {
      "entropy": 5.639569711685181,
      "epoch": 4.051552760150503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000345581846417596,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22796598821878433,
      "num_tokens": 80521770.0,
      "step": 36075
    },
    {
      "entropy": 5.713111019134521,
      "epoch": 4.05211433705846,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034554372315765927,
      "loss": 4.9418,
      "mean_token_accuracy": 0.23447264283895491,
      "num_tokens": 80532331.0,
      "step": 36080
    },
    {
      "entropy": 5.730151081085205,
      "epoch": 4.052675913966418,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003455055976515831,
      "loss": 4.91,
      "mean_token_accuracy": 0.23115717023611068,
      "num_tokens": 80544858.0,
      "step": 36085
    },
    {
      "entropy": 5.683489513397217,
      "epoch": 4.053237490874375,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034546746990058125,
      "loss": 5.0042,
      "mean_token_accuracy": 0.22221611589193344,
      "num_tokens": 80556493.0,
      "step": 36090
    },
    {
      "entropy": 5.660515880584716,
      "epoch": 4.053799067782332,
      "grad_norm": 1.125,
      "learning_rate": 0.00034542933990586795,
      "loss": 4.9789,
      "mean_token_accuracy": 0.2339089721441269,
      "num_tokens": 80567416.0,
      "step": 36095
    },
    {
      "entropy": 5.6839629173278805,
      "epoch": 4.0543606446902904,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034539120766865713,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22833152115345,
      "num_tokens": 80578175.0,
      "step": 36100
    },
    {
      "entropy": 5.617527961730957,
      "epoch": 4.054922221598248,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003453530731901628,
      "loss": 4.8495,
      "mean_token_accuracy": 0.23694443702697754,
      "num_tokens": 80588755.0,
      "step": 36105
    },
    {
      "entropy": 5.633136367797851,
      "epoch": 4.055483798506206,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003453149364715995,
      "loss": 4.9528,
      "mean_token_accuracy": 0.23046422749757767,
      "num_tokens": 80599734.0,
      "step": 36110
    },
    {
      "entropy": 5.585005855560302,
      "epoch": 4.056045375414163,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034527679751418126,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23945382833480836,
      "num_tokens": 80610765.0,
      "step": 36115
    },
    {
      "entropy": 5.6905440330505375,
      "epoch": 4.05660695232212,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003452386563191226,
      "loss": 4.9785,
      "mean_token_accuracy": 0.23186922520399095,
      "num_tokens": 80622680.0,
      "step": 36120
    },
    {
      "entropy": 5.703726530075073,
      "epoch": 4.057168529230078,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034520051288763784,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22792916297912597,
      "num_tokens": 80633115.0,
      "step": 36125
    },
    {
      "entropy": 5.6802552223205565,
      "epoch": 4.0577301061380355,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003451623672209415,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23025192320346832,
      "num_tokens": 80643750.0,
      "step": 36130
    },
    {
      "entropy": 5.681240463256836,
      "epoch": 4.058291683045993,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034512421932024825,
      "loss": 4.9767,
      "mean_token_accuracy": 0.22425515800714493,
      "num_tokens": 80654392.0,
      "step": 36135
    },
    {
      "entropy": 5.597859573364258,
      "epoch": 4.058853259953951,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003450860691867726,
      "loss": 4.9186,
      "mean_token_accuracy": 0.22677420675754548,
      "num_tokens": 80665509.0,
      "step": 36140
    },
    {
      "entropy": 5.620163202285767,
      "epoch": 4.059414836861908,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034504791682172934,
      "loss": 4.9101,
      "mean_token_accuracy": 0.23080091625452043,
      "num_tokens": 80676895.0,
      "step": 36145
    },
    {
      "entropy": 5.664107990264893,
      "epoch": 4.059976413769866,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003450097622263333,
      "loss": 4.9213,
      "mean_token_accuracy": 0.23079494535923004,
      "num_tokens": 80688321.0,
      "step": 36150
    },
    {
      "entropy": 5.634449863433838,
      "epoch": 4.060537990677823,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00034497160540179924,
      "loss": 4.8076,
      "mean_token_accuracy": 0.24683752357959748,
      "num_tokens": 80699534.0,
      "step": 36155
    },
    {
      "entropy": 5.5535417079925535,
      "epoch": 4.061099567585781,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034493344634934205,
      "loss": 4.8236,
      "mean_token_accuracy": 0.24430159330368043,
      "num_tokens": 80710999.0,
      "step": 36160
    },
    {
      "entropy": 5.604998826980591,
      "epoch": 4.061661144493739,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003448952850701769,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23132922053337096,
      "num_tokens": 80722943.0,
      "step": 36165
    },
    {
      "entropy": 5.6082042217254635,
      "epoch": 4.062222721401696,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034485712156551873,
      "loss": 4.9584,
      "mean_token_accuracy": 0.22614701837301254,
      "num_tokens": 80733142.0,
      "step": 36170
    },
    {
      "entropy": 5.607211542129517,
      "epoch": 4.062784298309653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034481895583658267,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22709420323371887,
      "num_tokens": 80745228.0,
      "step": 36175
    },
    {
      "entropy": 5.666900300979615,
      "epoch": 4.063345875217611,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034478078788458406,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2312960758805275,
      "num_tokens": 80756866.0,
      "step": 36180
    },
    {
      "entropy": 5.614627933502197,
      "epoch": 4.0639074521255685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000344742617710738,
      "loss": 4.8337,
      "mean_token_accuracy": 0.24670223742723466,
      "num_tokens": 80767541.0,
      "step": 36185
    },
    {
      "entropy": 5.631455755233764,
      "epoch": 4.064469029033527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034470444531626003,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2270177960395813,
      "num_tokens": 80779577.0,
      "step": 36190
    },
    {
      "entropy": 5.683849763870239,
      "epoch": 4.065030605941484,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034466627070236545,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2320490449666977,
      "num_tokens": 80790569.0,
      "step": 36195
    },
    {
      "entropy": 5.743772268295288,
      "epoch": 4.065592182849441,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003446280938702698,
      "loss": 5.0359,
      "mean_token_accuracy": 0.22201416045427322,
      "num_tokens": 80800887.0,
      "step": 36200
    },
    {
      "entropy": 5.687555265426636,
      "epoch": 4.066153759757399,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034458991482118865,
      "loss": 4.8955,
      "mean_token_accuracy": 0.228476682305336,
      "num_tokens": 80811967.0,
      "step": 36205
    },
    {
      "entropy": 5.617449569702148,
      "epoch": 4.066715336665356,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003445517335563376,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23408206999301912,
      "num_tokens": 80824770.0,
      "step": 36210
    },
    {
      "entropy": 5.595003366470337,
      "epoch": 4.067276913573314,
      "grad_norm": 1.140625,
      "learning_rate": 0.00034451355007693237,
      "loss": 4.9255,
      "mean_token_accuracy": 0.23212743401527405,
      "num_tokens": 80834913.0,
      "step": 36215
    },
    {
      "entropy": 5.661984777450561,
      "epoch": 4.067838490481272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003444753643841888,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22657689452171326,
      "num_tokens": 80846555.0,
      "step": 36220
    },
    {
      "entropy": 5.734800386428833,
      "epoch": 4.068400067389229,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034443717647932265,
      "loss": 5.0382,
      "mean_token_accuracy": 0.22119188755750657,
      "num_tokens": 80857774.0,
      "step": 36225
    },
    {
      "entropy": 5.752104377746582,
      "epoch": 4.068961644297186,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003443989863635498,
      "loss": 5.007,
      "mean_token_accuracy": 0.22575817555189132,
      "num_tokens": 80868618.0,
      "step": 36230
    },
    {
      "entropy": 5.67226824760437,
      "epoch": 4.069523221205144,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034436079403808647,
      "loss": 5.0101,
      "mean_token_accuracy": 0.22244301438331604,
      "num_tokens": 80880319.0,
      "step": 36235
    },
    {
      "entropy": 5.6568385601043705,
      "epoch": 4.070084798113101,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00034432259950414846,
      "loss": 4.966,
      "mean_token_accuracy": 0.23123422563076018,
      "num_tokens": 80891194.0,
      "step": 36240
    },
    {
      "entropy": 5.591697120666504,
      "epoch": 4.0706463750210595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000344284402762952,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23256979882717133,
      "num_tokens": 80902322.0,
      "step": 36245
    },
    {
      "entropy": 5.717619276046753,
      "epoch": 4.071207951929017,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003442462038157132,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22308236807584764,
      "num_tokens": 80913324.0,
      "step": 36250
    },
    {
      "entropy": 5.6126429557800295,
      "epoch": 4.071769528836974,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003442080026636485,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23457279950380325,
      "num_tokens": 80923360.0,
      "step": 36255
    },
    {
      "entropy": 5.675219535827637,
      "epoch": 4.072331105744932,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003441697993079742,
      "loss": 5.0093,
      "mean_token_accuracy": 0.22020789980888367,
      "num_tokens": 80935359.0,
      "step": 36260
    },
    {
      "entropy": 5.600913858413696,
      "epoch": 4.072892682652889,
      "grad_norm": 1.125,
      "learning_rate": 0.0003441315937499066,
      "loss": 4.8609,
      "mean_token_accuracy": 0.22962099611759185,
      "num_tokens": 80945885.0,
      "step": 36265
    },
    {
      "entropy": 5.699811601638794,
      "epoch": 4.0734542595608465,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034409338599066224,
      "loss": 4.9517,
      "mean_token_accuracy": 0.23188893347978592,
      "num_tokens": 80955700.0,
      "step": 36270
    },
    {
      "entropy": 5.5941516876220705,
      "epoch": 4.074015836468805,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003440551760314577,
      "loss": 4.9046,
      "mean_token_accuracy": 0.23363907486200333,
      "num_tokens": 80967130.0,
      "step": 36275
    },
    {
      "entropy": 5.6470863819122314,
      "epoch": 4.074577413376762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034401696387350954,
      "loss": 4.9439,
      "mean_token_accuracy": 0.2294040948152542,
      "num_tokens": 80979677.0,
      "step": 36280
    },
    {
      "entropy": 5.6660933017730715,
      "epoch": 4.075138990284719,
      "grad_norm": 1.25,
      "learning_rate": 0.0003439787495180345,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22952058166265488,
      "num_tokens": 80990563.0,
      "step": 36285
    },
    {
      "entropy": 5.638608694076538,
      "epoch": 4.075700567192677,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034394053296624936,
      "loss": 4.9124,
      "mean_token_accuracy": 0.23810099810361862,
      "num_tokens": 81001731.0,
      "step": 36290
    },
    {
      "entropy": 5.701417875289917,
      "epoch": 4.076262144100634,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003439023142193709,
      "loss": 4.9962,
      "mean_token_accuracy": 0.22332969903945923,
      "num_tokens": 81013072.0,
      "step": 36295
    },
    {
      "entropy": 5.646234846115112,
      "epoch": 4.0768237210085925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000343864093278616,
      "loss": 4.9363,
      "mean_token_accuracy": 0.233484710752964,
      "num_tokens": 81023628.0,
      "step": 36300
    },
    {
      "entropy": 5.566300344467163,
      "epoch": 4.07738529791655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003438258701452017,
      "loss": 4.8203,
      "mean_token_accuracy": 0.24018087685108186,
      "num_tokens": 81034856.0,
      "step": 36305
    },
    {
      "entropy": 5.677752876281739,
      "epoch": 4.077946874824507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003437876448203449,
      "loss": 4.9653,
      "mean_token_accuracy": 0.23183444142341614,
      "num_tokens": 81046998.0,
      "step": 36310
    },
    {
      "entropy": 5.6527362823486325,
      "epoch": 4.078508451732465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003437494173052629,
      "loss": 4.9232,
      "mean_token_accuracy": 0.23111740052700042,
      "num_tokens": 81058292.0,
      "step": 36315
    },
    {
      "entropy": 5.675209856033325,
      "epoch": 4.079070028640422,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003437111876011728,
      "loss": 4.9019,
      "mean_token_accuracy": 0.23386302441358567,
      "num_tokens": 81069907.0,
      "step": 36320
    },
    {
      "entropy": 5.658973169326782,
      "epoch": 4.0796316055483794,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034367295570929176,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22210567742586135,
      "num_tokens": 81080700.0,
      "step": 36325
    },
    {
      "entropy": 5.615248346328736,
      "epoch": 4.0801931824563376,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003436347216308373,
      "loss": 4.9066,
      "mean_token_accuracy": 0.23083320409059524,
      "num_tokens": 81092484.0,
      "step": 36330
    },
    {
      "entropy": 5.660621118545532,
      "epoch": 4.080754759364295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003435964853670265,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23320140242576598,
      "num_tokens": 81103992.0,
      "step": 36335
    },
    {
      "entropy": 5.681908082962036,
      "epoch": 4.081316336272253,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034355824691907714,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22366584986448287,
      "num_tokens": 81114995.0,
      "step": 36340
    },
    {
      "entropy": 5.537897348403931,
      "epoch": 4.08187791318021,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00034352000628820655,
      "loss": 4.7978,
      "mean_token_accuracy": 0.23680518418550492,
      "num_tokens": 81125627.0,
      "step": 36345
    },
    {
      "entropy": 5.669006776809693,
      "epoch": 4.082439490088167,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003434817634756324,
      "loss": 5.0338,
      "mean_token_accuracy": 0.22275872975587846,
      "num_tokens": 81136957.0,
      "step": 36350
    },
    {
      "entropy": 5.65050745010376,
      "epoch": 4.083001066996125,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003434435184825724,
      "loss": 4.9161,
      "mean_token_accuracy": 0.23256539851427077,
      "num_tokens": 81148840.0,
      "step": 36355
    },
    {
      "entropy": 5.690972375869751,
      "epoch": 4.083562643904083,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034340527131024415,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22674258500337602,
      "num_tokens": 81160385.0,
      "step": 36360
    },
    {
      "entropy": 5.594469976425171,
      "epoch": 4.08412422081204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003433670219598656,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23291660994291305,
      "num_tokens": 81171334.0,
      "step": 36365
    },
    {
      "entropy": 5.639058589935303,
      "epoch": 4.084685797719998,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003433287704326545,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23189856857061386,
      "num_tokens": 81182547.0,
      "step": 36370
    },
    {
      "entropy": 5.655236291885376,
      "epoch": 4.085247374627955,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003432905167298288,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22507529109716415,
      "num_tokens": 81193427.0,
      "step": 36375
    },
    {
      "entropy": 5.684911060333252,
      "epoch": 4.085808951535913,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034325226085260666,
      "loss": 4.87,
      "mean_token_accuracy": 0.23727635145187378,
      "num_tokens": 81204522.0,
      "step": 36380
    },
    {
      "entropy": 5.686816930770874,
      "epoch": 4.0863705284438705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000343214002802206,
      "loss": 5.0191,
      "mean_token_accuracy": 0.22607712149620057,
      "num_tokens": 81215209.0,
      "step": 36385
    },
    {
      "entropy": 5.649035215377808,
      "epoch": 4.086932105351828,
      "grad_norm": 1.125,
      "learning_rate": 0.00034317574257984514,
      "loss": 4.9895,
      "mean_token_accuracy": 0.22569866180419923,
      "num_tokens": 81226083.0,
      "step": 36390
    },
    {
      "entropy": 5.602921485900879,
      "epoch": 4.087493682259786,
      "grad_norm": 1.203125,
      "learning_rate": 0.000343137480186742,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23250085562467576,
      "num_tokens": 81236531.0,
      "step": 36395
    },
    {
      "entropy": 5.6385533809661865,
      "epoch": 4.088055259167743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034309921562411527,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2225663036108017,
      "num_tokens": 81246856.0,
      "step": 36400
    },
    {
      "entropy": 5.68622407913208,
      "epoch": 4.0886168360757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034306094889318295,
      "loss": 4.9876,
      "mean_token_accuracy": 0.2213197484612465,
      "num_tokens": 81259352.0,
      "step": 36405
    },
    {
      "entropy": 5.727842140197754,
      "epoch": 4.089178412983658,
      "grad_norm": 1.25,
      "learning_rate": 0.0003430226799951637,
      "loss": 5.0222,
      "mean_token_accuracy": 0.22818990647792817,
      "num_tokens": 81270020.0,
      "step": 36410
    },
    {
      "entropy": 5.6504371643066404,
      "epoch": 4.089739989891616,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034298440893127584,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22786771655082702,
      "num_tokens": 81281625.0,
      "step": 36415
    },
    {
      "entropy": 5.612018251419068,
      "epoch": 4.090301566799573,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034294613570273817,
      "loss": 4.7945,
      "mean_token_accuracy": 0.23792178332805633,
      "num_tokens": 81292340.0,
      "step": 36420
    },
    {
      "entropy": 5.634851694107056,
      "epoch": 4.090863143707531,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034290786031076917,
      "loss": 4.8809,
      "mean_token_accuracy": 0.23116752803325652,
      "num_tokens": 81304494.0,
      "step": 36425
    },
    {
      "entropy": 5.639945983886719,
      "epoch": 4.091424720615488,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003428695827565875,
      "loss": 4.9278,
      "mean_token_accuracy": 0.23477163910865784,
      "num_tokens": 81315365.0,
      "step": 36430
    },
    {
      "entropy": 5.68047080039978,
      "epoch": 4.091986297523446,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034283130304141204,
      "loss": 4.9349,
      "mean_token_accuracy": 0.2282964214682579,
      "num_tokens": 81326970.0,
      "step": 36435
    },
    {
      "entropy": 5.635422849655152,
      "epoch": 4.092547874431403,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003427930211664615,
      "loss": 4.9075,
      "mean_token_accuracy": 0.23570260852575303,
      "num_tokens": 81337117.0,
      "step": 36440
    },
    {
      "entropy": 5.676207876205444,
      "epoch": 4.093109451339361,
      "grad_norm": 1.15625,
      "learning_rate": 0.000342754737132955,
      "loss": 4.9812,
      "mean_token_accuracy": 0.22117028683423995,
      "num_tokens": 81347984.0,
      "step": 36445
    },
    {
      "entropy": 5.687882137298584,
      "epoch": 4.093671028247319,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034271645094211125,
      "loss": 4.9449,
      "mean_token_accuracy": 0.23468544483184814,
      "num_tokens": 81358685.0,
      "step": 36450
    },
    {
      "entropy": 5.668383550643921,
      "epoch": 4.094232605155276,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003426781625951495,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22925609201192856,
      "num_tokens": 81368724.0,
      "step": 36455
    },
    {
      "entropy": 5.719790267944336,
      "epoch": 4.094794182063233,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034263987209328886,
      "loss": 5.0328,
      "mean_token_accuracy": 0.22495269924402236,
      "num_tokens": 81380570.0,
      "step": 36460
    },
    {
      "entropy": 5.63255352973938,
      "epoch": 4.095355758971191,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003426015794377483,
      "loss": 4.952,
      "mean_token_accuracy": 0.2330651193857193,
      "num_tokens": 81391606.0,
      "step": 36465
    },
    {
      "entropy": 5.732181119918823,
      "epoch": 4.0959173358791485,
      "grad_norm": 1.125,
      "learning_rate": 0.0003425632846297474,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22756431102752686,
      "num_tokens": 81402439.0,
      "step": 36470
    },
    {
      "entropy": 5.66036286354065,
      "epoch": 4.096478912787107,
      "grad_norm": 1.125,
      "learning_rate": 0.00034252498767050525,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22895006686449051,
      "num_tokens": 81413500.0,
      "step": 36475
    },
    {
      "entropy": 5.5654082775115965,
      "epoch": 4.097040489695064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003424866885612413,
      "loss": 4.8439,
      "mean_token_accuracy": 0.22768006175756456,
      "num_tokens": 81424656.0,
      "step": 36480
    },
    {
      "entropy": 5.77014307975769,
      "epoch": 4.097602066603021,
      "grad_norm": 1.203125,
      "learning_rate": 0.000342448387303175,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2325441226363182,
      "num_tokens": 81436597.0,
      "step": 36485
    },
    {
      "entropy": 5.733474588394165,
      "epoch": 4.098163643510979,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034241008389752594,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22032178789377213,
      "num_tokens": 81448955.0,
      "step": 36490
    },
    {
      "entropy": 5.659018564224243,
      "epoch": 4.098725220418936,
      "grad_norm": 1.9375,
      "learning_rate": 0.0003423717783455136,
      "loss": 4.922,
      "mean_token_accuracy": 0.23695116639137268,
      "num_tokens": 81460445.0,
      "step": 36495
    },
    {
      "entropy": 5.666285705566406,
      "epoch": 4.099286797326894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034233347064835775,
      "loss": 4.916,
      "mean_token_accuracy": 0.22875260412693024,
      "num_tokens": 81472103.0,
      "step": 36500
    },
    {
      "entropy": 5.609416866302491,
      "epoch": 4.099848374234852,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003422951608072781,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2292238652706146,
      "num_tokens": 81483051.0,
      "step": 36505
    },
    {
      "entropy": 5.577567005157471,
      "epoch": 4.100409951142809,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003422568488234944,
      "loss": 4.8618,
      "mean_token_accuracy": 0.23099908083677292,
      "num_tokens": 81493755.0,
      "step": 36510
    },
    {
      "entropy": 5.570014905929566,
      "epoch": 4.100971528050766,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034221853469822654,
      "loss": 4.8405,
      "mean_token_accuracy": 0.2392558768391609,
      "num_tokens": 81505738.0,
      "step": 36515
    },
    {
      "entropy": 5.708972644805908,
      "epoch": 4.101533104958724,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00034218021843269444,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2251397341489792,
      "num_tokens": 81516410.0,
      "step": 36520
    },
    {
      "entropy": 5.659409713745117,
      "epoch": 4.1020946818666815,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003421419000281182,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2243536338210106,
      "num_tokens": 81527516.0,
      "step": 36525
    },
    {
      "entropy": 5.691025590896606,
      "epoch": 4.10265625877464,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003421035794857178,
      "loss": 4.9701,
      "mean_token_accuracy": 0.22579001039266586,
      "num_tokens": 81538226.0,
      "step": 36530
    },
    {
      "entropy": 5.725802373886109,
      "epoch": 4.103217835682597,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003420652568067134,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22455905079841615,
      "num_tokens": 81549699.0,
      "step": 36535
    },
    {
      "entropy": 5.729009389877319,
      "epoch": 4.103779412590554,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034202693199232516,
      "loss": 5.047,
      "mean_token_accuracy": 0.2157588168978691,
      "num_tokens": 81560404.0,
      "step": 36540
    },
    {
      "entropy": 5.724722862243652,
      "epoch": 4.104340989498512,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034198860504377343,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22847013026475907,
      "num_tokens": 81571758.0,
      "step": 36545
    },
    {
      "entropy": 5.642856550216675,
      "epoch": 4.104902566406469,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034195027596227857,
      "loss": 4.832,
      "mean_token_accuracy": 0.23348561078310012,
      "num_tokens": 81583555.0,
      "step": 36550
    },
    {
      "entropy": 5.591591453552246,
      "epoch": 4.1054641433144266,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003419119447490609,
      "loss": 4.8905,
      "mean_token_accuracy": 0.23363024294376372,
      "num_tokens": 81594901.0,
      "step": 36555
    },
    {
      "entropy": 5.6618030071258545,
      "epoch": 4.106025720222385,
      "grad_norm": 1.125,
      "learning_rate": 0.000341873611405341,
      "loss": 4.8903,
      "mean_token_accuracy": 0.23750953674316405,
      "num_tokens": 81606229.0,
      "step": 36560
    },
    {
      "entropy": 5.795843267440796,
      "epoch": 4.106587297130342,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003418352759323394,
      "loss": 4.9918,
      "mean_token_accuracy": 0.23272041976451874,
      "num_tokens": 81616543.0,
      "step": 36565
    },
    {
      "entropy": 5.666664409637451,
      "epoch": 4.1071488740383,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034179693833127667,
      "loss": 4.9677,
      "mean_token_accuracy": 0.23321881741285325,
      "num_tokens": 81627517.0,
      "step": 36570
    },
    {
      "entropy": 5.622861099243164,
      "epoch": 4.107710450946257,
      "grad_norm": 1.25,
      "learning_rate": 0.00034175859860337347,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2296190306544304,
      "num_tokens": 81637235.0,
      "step": 36575
    },
    {
      "entropy": 5.657913875579834,
      "epoch": 4.108272027854214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034172025674985076,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2285772055387497,
      "num_tokens": 81648539.0,
      "step": 36580
    },
    {
      "entropy": 5.644637537002564,
      "epoch": 4.1088336047621725,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034168191277192904,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2342006340622902,
      "num_tokens": 81659981.0,
      "step": 36585
    },
    {
      "entropy": 5.694956827163696,
      "epoch": 4.10939518167013,
      "grad_norm": 1.1875,
      "learning_rate": 0.00034164356667082933,
      "loss": 4.8838,
      "mean_token_accuracy": 0.23028636276721953,
      "num_tokens": 81669602.0,
      "step": 36590
    },
    {
      "entropy": 5.6164435863494875,
      "epoch": 4.109956758578087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003416052184477727,
      "loss": 4.9611,
      "mean_token_accuracy": 0.22724911719560623,
      "num_tokens": 81679995.0,
      "step": 36595
    },
    {
      "entropy": 5.597440147399903,
      "epoch": 4.110518335486045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003415668681039801,
      "loss": 4.8935,
      "mean_token_accuracy": 0.23286951780319215,
      "num_tokens": 81691295.0,
      "step": 36600
    },
    {
      "entropy": 5.647963905334473,
      "epoch": 4.111079912394002,
      "grad_norm": 1.125,
      "learning_rate": 0.00034152851564067254,
      "loss": 4.9199,
      "mean_token_accuracy": 0.2293574631214142,
      "num_tokens": 81700966.0,
      "step": 36605
    },
    {
      "entropy": 5.726864051818848,
      "epoch": 4.1116414893019595,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003414901610590712,
      "loss": 5.0162,
      "mean_token_accuracy": 0.22146336883306503,
      "num_tokens": 81713469.0,
      "step": 36610
    },
    {
      "entropy": 5.661908054351807,
      "epoch": 4.112203066209918,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003414518043603974,
      "loss": 4.9247,
      "mean_token_accuracy": 0.228926482796669,
      "num_tokens": 81724748.0,
      "step": 36615
    },
    {
      "entropy": 5.684364366531372,
      "epoch": 4.112764643117875,
      "grad_norm": 1.203125,
      "learning_rate": 0.00034141344554587243,
      "loss": 5.0179,
      "mean_token_accuracy": 0.22436599284410477,
      "num_tokens": 81735694.0,
      "step": 36620
    },
    {
      "entropy": 5.770839071273803,
      "epoch": 4.113326220025833,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003413750846167175,
      "loss": 4.95,
      "mean_token_accuracy": 0.22752096951007844,
      "num_tokens": 81748251.0,
      "step": 36625
    },
    {
      "entropy": 5.662378358840942,
      "epoch": 4.11388779693379,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003413367215741542,
      "loss": 4.8751,
      "mean_token_accuracy": 0.23384592682123184,
      "num_tokens": 81760750.0,
      "step": 36630
    },
    {
      "entropy": 5.697823524475098,
      "epoch": 4.114449373841747,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034129835641940394,
      "loss": 5.0197,
      "mean_token_accuracy": 0.22425539046525955,
      "num_tokens": 81772100.0,
      "step": 36635
    },
    {
      "entropy": 5.687202072143554,
      "epoch": 4.1150109507497055,
      "grad_norm": 1.125,
      "learning_rate": 0.00034125998915368827,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22686814069747924,
      "num_tokens": 81783795.0,
      "step": 36640
    },
    {
      "entropy": 5.6623693943023685,
      "epoch": 4.115572527657663,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034122161977822885,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22961791157722472,
      "num_tokens": 81794713.0,
      "step": 36645
    },
    {
      "entropy": 5.684852266311646,
      "epoch": 4.11613410456562,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003411832482942474,
      "loss": 4.9393,
      "mean_token_accuracy": 0.22881957590579988,
      "num_tokens": 81806493.0,
      "step": 36650
    },
    {
      "entropy": 5.6299714088439945,
      "epoch": 4.116695681473578,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034114487470296575,
      "loss": 4.8186,
      "mean_token_accuracy": 0.23771592676639558,
      "num_tokens": 81818232.0,
      "step": 36655
    },
    {
      "entropy": 5.587494039535523,
      "epoch": 4.117257258381535,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034110649900560554,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22935966402292252,
      "num_tokens": 81829439.0,
      "step": 36660
    },
    {
      "entropy": 5.664763641357422,
      "epoch": 4.117818835289493,
      "grad_norm": 1.125,
      "learning_rate": 0.0003410681212033887,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2294595554471016,
      "num_tokens": 81839585.0,
      "step": 36665
    },
    {
      "entropy": 5.669733190536499,
      "epoch": 4.1183804121974505,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034102974129753726,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2248108357191086,
      "num_tokens": 81850648.0,
      "step": 36670
    },
    {
      "entropy": 5.648495149612427,
      "epoch": 4.118941989105408,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003409913592892733,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22758804857730866,
      "num_tokens": 81862597.0,
      "step": 36675
    },
    {
      "entropy": 5.726099920272827,
      "epoch": 4.119503566013366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003409529751798189,
      "loss": 4.9567,
      "mean_token_accuracy": 0.22899453341960907,
      "num_tokens": 81874540.0,
      "step": 36680
    },
    {
      "entropy": 5.638830757141113,
      "epoch": 4.120065142921323,
      "grad_norm": 1.125,
      "learning_rate": 0.0003409145889703961,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2383740246295929,
      "num_tokens": 81884205.0,
      "step": 36685
    },
    {
      "entropy": 5.6077690601348875,
      "epoch": 4.12062671982928,
      "grad_norm": 1.265625,
      "learning_rate": 0.00034087620066222733,
      "loss": 4.838,
      "mean_token_accuracy": 0.23156075477600097,
      "num_tokens": 81895552.0,
      "step": 36690
    },
    {
      "entropy": 5.7366533279418945,
      "epoch": 4.121188296737238,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003408378102565347,
      "loss": 5.0372,
      "mean_token_accuracy": 0.21879565715789795,
      "num_tokens": 81908169.0,
      "step": 36695
    },
    {
      "entropy": 5.648630428314209,
      "epoch": 4.121749873645196,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003407994177545407,
      "loss": 4.9303,
      "mean_token_accuracy": 0.23350273072719574,
      "num_tokens": 81919465.0,
      "step": 36700
    },
    {
      "entropy": 5.656400680541992,
      "epoch": 4.122311450553153,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003407610231574677,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22509755045175553,
      "num_tokens": 81930558.0,
      "step": 36705
    },
    {
      "entropy": 5.69405083656311,
      "epoch": 4.122873027461111,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003407226264665382,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22842090725898742,
      "num_tokens": 81941325.0,
      "step": 36710
    },
    {
      "entropy": 5.649309206008911,
      "epoch": 4.123434604369068,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034068422768297486,
      "loss": 4.853,
      "mean_token_accuracy": 0.23009466081857682,
      "num_tokens": 81953225.0,
      "step": 36715
    },
    {
      "entropy": 5.694357013702392,
      "epoch": 4.123996181277026,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003406458268080001,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22668455243110658,
      "num_tokens": 81964365.0,
      "step": 36720
    },
    {
      "entropy": 5.624047374725341,
      "epoch": 4.1245577581849835,
      "grad_norm": 1.21875,
      "learning_rate": 0.00034060742384283695,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2268099531531334,
      "num_tokens": 81974648.0,
      "step": 36725
    },
    {
      "entropy": 5.627902173995972,
      "epoch": 4.125119335092941,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003405690187887078,
      "loss": 4.871,
      "mean_token_accuracy": 0.2371126666665077,
      "num_tokens": 81985502.0,
      "step": 36730
    },
    {
      "entropy": 5.733272695541382,
      "epoch": 4.125680912000899,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00034053061164683574,
      "loss": 4.9851,
      "mean_token_accuracy": 0.22780778259038925,
      "num_tokens": 81998279.0,
      "step": 36735
    },
    {
      "entropy": 5.6431725978851315,
      "epoch": 4.126242488908856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034049220241844357,
      "loss": 4.9153,
      "mean_token_accuracy": 0.23194042593240738,
      "num_tokens": 82009911.0,
      "step": 36740
    },
    {
      "entropy": 5.589267253875732,
      "epoch": 4.126804065816813,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003404537911047543,
      "loss": 4.8449,
      "mean_token_accuracy": 0.238273024559021,
      "num_tokens": 82020219.0,
      "step": 36745
    },
    {
      "entropy": 5.58437671661377,
      "epoch": 4.127365642724771,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034041537770699095,
      "loss": 4.7911,
      "mean_token_accuracy": 0.23759319335222245,
      "num_tokens": 82030997.0,
      "step": 36750
    },
    {
      "entropy": 5.682664060592652,
      "epoch": 4.127927219632729,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003403769622263765,
      "loss": 4.9471,
      "mean_token_accuracy": 0.2328087270259857,
      "num_tokens": 82043384.0,
      "step": 36755
    },
    {
      "entropy": 5.62950553894043,
      "epoch": 4.128488796540687,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00034033854466413425,
      "loss": 4.8539,
      "mean_token_accuracy": 0.23211251348257064,
      "num_tokens": 82053346.0,
      "step": 36760
    },
    {
      "entropy": 5.626910161972046,
      "epoch": 4.129050373448644,
      "grad_norm": 1.125,
      "learning_rate": 0.0003403001250214875,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23033001720905305,
      "num_tokens": 82063628.0,
      "step": 36765
    },
    {
      "entropy": 5.644896507263184,
      "epoch": 4.129611950356601,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003402617032996593,
      "loss": 4.9112,
      "mean_token_accuracy": 0.23688815236091615,
      "num_tokens": 82074097.0,
      "step": 36770
    },
    {
      "entropy": 5.647297668457031,
      "epoch": 4.130173527264559,
      "grad_norm": 1.125,
      "learning_rate": 0.0003402232794998733,
      "loss": 4.9924,
      "mean_token_accuracy": 0.22751137018203735,
      "num_tokens": 82085159.0,
      "step": 36775
    },
    {
      "entropy": 5.666732931137085,
      "epoch": 4.130735104172516,
      "grad_norm": 1.234375,
      "learning_rate": 0.00034018485362335266,
      "loss": 4.9591,
      "mean_token_accuracy": 0.228511543571949,
      "num_tokens": 82096299.0,
      "step": 36780
    },
    {
      "entropy": 5.628242111206054,
      "epoch": 4.131296681080474,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003401464256713211,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2298291251063347,
      "num_tokens": 82107111.0,
      "step": 36785
    },
    {
      "entropy": 5.657581853866577,
      "epoch": 4.131858257988432,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00034010799564500197,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2298978239297867,
      "num_tokens": 82117889.0,
      "step": 36790
    },
    {
      "entropy": 5.604825258255005,
      "epoch": 4.132419834896389,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003400695635456191,
      "loss": 4.8866,
      "mean_token_accuracy": 0.23236483335494995,
      "num_tokens": 82128380.0,
      "step": 36795
    },
    {
      "entropy": 5.602800559997559,
      "epoch": 4.132981411804346,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003400311293743961,
      "loss": 4.8154,
      "mean_token_accuracy": 0.23896071761846543,
      "num_tokens": 82138368.0,
      "step": 36800
    },
    {
      "entropy": 5.694503688812256,
      "epoch": 4.133542988712304,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003399926931325567,
      "loss": 5.0027,
      "mean_token_accuracy": 0.22790478765964509,
      "num_tokens": 82148449.0,
      "step": 36805
    },
    {
      "entropy": 5.559890222549439,
      "epoch": 4.1341045656202615,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033995425482132464,
      "loss": 4.8288,
      "mean_token_accuracy": 0.23462088853120805,
      "num_tokens": 82160937.0,
      "step": 36810
    },
    {
      "entropy": 5.6204689025878904,
      "epoch": 4.13466614252822,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033991581444192405,
      "loss": 4.8467,
      "mean_token_accuracy": 0.23708777725696564,
      "num_tokens": 82171546.0,
      "step": 36815
    },
    {
      "entropy": 5.652288293838501,
      "epoch": 4.135227719436177,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003398773719955787,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23397375494241715,
      "num_tokens": 82182613.0,
      "step": 36820
    },
    {
      "entropy": 5.556546211242676,
      "epoch": 4.135789296344134,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003398389274835126,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23620132803916932,
      "num_tokens": 82193757.0,
      "step": 36825
    },
    {
      "entropy": 5.631818628311157,
      "epoch": 4.136350873252092,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033980048090694997,
      "loss": 4.952,
      "mean_token_accuracy": 0.22495618611574172,
      "num_tokens": 82205601.0,
      "step": 36830
    },
    {
      "entropy": 5.688161993026734,
      "epoch": 4.136912450160049,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003397620322671149,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2297626867890358,
      "num_tokens": 82217022.0,
      "step": 36835
    },
    {
      "entropy": 5.623322200775147,
      "epoch": 4.137474027068007,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033972358156523163,
      "loss": 4.864,
      "mean_token_accuracy": 0.23102327287197114,
      "num_tokens": 82227682.0,
      "step": 36840
    },
    {
      "entropy": 5.598960733413696,
      "epoch": 4.138035603975965,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003396851288025243,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23037390559911727,
      "num_tokens": 82238754.0,
      "step": 36845
    },
    {
      "entropy": 5.7539843082427975,
      "epoch": 4.138597180883922,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003396466739802175,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2275544211268425,
      "num_tokens": 82249536.0,
      "step": 36850
    },
    {
      "entropy": 5.667871141433716,
      "epoch": 4.13915875779188,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003396082170995355,
      "loss": 4.9044,
      "mean_token_accuracy": 0.23054079264402388,
      "num_tokens": 82260374.0,
      "step": 36855
    },
    {
      "entropy": 5.644634628295899,
      "epoch": 4.139720334699837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003395697581617028,
      "loss": 4.8968,
      "mean_token_accuracy": 0.23027797788381577,
      "num_tokens": 82271801.0,
      "step": 36860
    },
    {
      "entropy": 5.668447542190552,
      "epoch": 4.1402819116077945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033953129716794396,
      "loss": 5.0456,
      "mean_token_accuracy": 0.22257663309574127,
      "num_tokens": 82283149.0,
      "step": 36865
    },
    {
      "entropy": 5.686263656616211,
      "epoch": 4.140843488515753,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003394928341194836,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22687898874282836,
      "num_tokens": 82293003.0,
      "step": 36870
    },
    {
      "entropy": 5.707385301589966,
      "epoch": 4.14140506542371,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003394543690175464,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23022557348012923,
      "num_tokens": 82303734.0,
      "step": 36875
    },
    {
      "entropy": 5.592858695983887,
      "epoch": 4.141966642331667,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033941590186335703,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22857795804738998,
      "num_tokens": 82314569.0,
      "step": 36880
    },
    {
      "entropy": 5.634070873260498,
      "epoch": 4.142528219239625,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003393774326581404,
      "loss": 4.9249,
      "mean_token_accuracy": 0.23042971938848494,
      "num_tokens": 82325174.0,
      "step": 36885
    },
    {
      "entropy": 5.6895294189453125,
      "epoch": 4.143089796147582,
      "grad_norm": 1.125,
      "learning_rate": 0.00033933896140312135,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2302846521139145,
      "num_tokens": 82337448.0,
      "step": 36890
    },
    {
      "entropy": 5.720573091506958,
      "epoch": 4.1436513730555395,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033930048809952484,
      "loss": 4.9847,
      "mean_token_accuracy": 0.22517139762639998,
      "num_tokens": 82348705.0,
      "step": 36895
    },
    {
      "entropy": 5.737167406082153,
      "epoch": 4.144212949963498,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033926201274857575,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22975176721811294,
      "num_tokens": 82360439.0,
      "step": 36900
    },
    {
      "entropy": 5.645040464401245,
      "epoch": 4.144774526871455,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033922353535149934,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23251235485076904,
      "num_tokens": 82371346.0,
      "step": 36905
    },
    {
      "entropy": 5.604332685470581,
      "epoch": 4.145336103779413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003391850559095207,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2378978908061981,
      "num_tokens": 82382221.0,
      "step": 36910
    },
    {
      "entropy": 5.669711494445801,
      "epoch": 4.14589768068737,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003391465744238649,
      "loss": 4.9654,
      "mean_token_accuracy": 0.2215780109167099,
      "num_tokens": 82392290.0,
      "step": 36915
    },
    {
      "entropy": 5.73162956237793,
      "epoch": 4.146459257595327,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003391080908957573,
      "loss": 4.9896,
      "mean_token_accuracy": 0.22164733409881593,
      "num_tokens": 82403513.0,
      "step": 36920
    },
    {
      "entropy": 5.622276544570923,
      "epoch": 4.1470208345032855,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033906960532642327,
      "loss": 4.8495,
      "mean_token_accuracy": 0.24176120460033418,
      "num_tokens": 82414537.0,
      "step": 36925
    },
    {
      "entropy": 5.60680456161499,
      "epoch": 4.147582411411243,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003390311177170881,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2366253539919853,
      "num_tokens": 82425330.0,
      "step": 36930
    },
    {
      "entropy": 5.526845264434814,
      "epoch": 4.1481439883192,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003389926280689773,
      "loss": 4.857,
      "mean_token_accuracy": 0.23218757808208465,
      "num_tokens": 82435703.0,
      "step": 36935
    },
    {
      "entropy": 5.66630187034607,
      "epoch": 4.148705565227158,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033895413638331646,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2323676601052284,
      "num_tokens": 82447302.0,
      "step": 36940
    },
    {
      "entropy": 5.650882530212402,
      "epoch": 4.149267142135115,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003389156426613311,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22370182871818542,
      "num_tokens": 82458379.0,
      "step": 36945
    },
    {
      "entropy": 5.69480619430542,
      "epoch": 4.149828719043073,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033887714690424687,
      "loss": 4.9044,
      "mean_token_accuracy": 0.23151775300502778,
      "num_tokens": 82469728.0,
      "step": 36950
    },
    {
      "entropy": 5.698250722885132,
      "epoch": 4.150390295951031,
      "grad_norm": 1.125,
      "learning_rate": 0.00033883864911328954,
      "loss": 5.0186,
      "mean_token_accuracy": 0.22475594878196717,
      "num_tokens": 82482216.0,
      "step": 36955
    },
    {
      "entropy": 5.710199689865112,
      "epoch": 4.150951872858988,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033880014928968483,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22978626936674118,
      "num_tokens": 82494511.0,
      "step": 36960
    },
    {
      "entropy": 5.661095428466797,
      "epoch": 4.151513449766946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033876164743465865,
      "loss": 4.9606,
      "mean_token_accuracy": 0.23045448064804078,
      "num_tokens": 82505982.0,
      "step": 36965
    },
    {
      "entropy": 5.630374574661255,
      "epoch": 4.152075026674903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003387231435494368,
      "loss": 4.8716,
      "mean_token_accuracy": 0.23504676967859267,
      "num_tokens": 82516682.0,
      "step": 36970
    },
    {
      "entropy": 5.653599739074707,
      "epoch": 4.15263660358286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033868463763524545,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22914841324090957,
      "num_tokens": 82527742.0,
      "step": 36975
    },
    {
      "entropy": 5.645258808135987,
      "epoch": 4.1531981804908185,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033864612969331046,
      "loss": 4.8841,
      "mean_token_accuracy": 0.23174204528331757,
      "num_tokens": 82538130.0,
      "step": 36980
    },
    {
      "entropy": 5.625742197036743,
      "epoch": 4.153759757398776,
      "grad_norm": 1.125,
      "learning_rate": 0.0003386076197248581,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22879607677459718,
      "num_tokens": 82549695.0,
      "step": 36985
    },
    {
      "entropy": 5.688390016555786,
      "epoch": 4.154321334306733,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003385691077311144,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22398089319467546,
      "num_tokens": 82560778.0,
      "step": 36990
    },
    {
      "entropy": 5.622143888473511,
      "epoch": 4.154882911214691,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033853059371330565,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22837408781051635,
      "num_tokens": 82572021.0,
      "step": 36995
    },
    {
      "entropy": 5.740124130249024,
      "epoch": 4.155444488122648,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003384920776726582,
      "loss": 4.9542,
      "mean_token_accuracy": 0.23189810812473297,
      "num_tokens": 82583173.0,
      "step": 37000
    },
    {
      "entropy": 5.64091534614563,
      "epoch": 4.156006065030606,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033845355961039823,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23411096334457399,
      "num_tokens": 82594279.0,
      "step": 37005
    },
    {
      "entropy": 5.693149042129517,
      "epoch": 4.1565676419385635,
      "grad_norm": 1.234375,
      "learning_rate": 0.00033841503952775255,
      "loss": 5.0246,
      "mean_token_accuracy": 0.22846726030111314,
      "num_tokens": 82605038.0,
      "step": 37010
    },
    {
      "entropy": 5.738646078109741,
      "epoch": 4.157129218846521,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033837651742594734,
      "loss": 5.0376,
      "mean_token_accuracy": 0.22764464169740678,
      "num_tokens": 82616627.0,
      "step": 37015
    },
    {
      "entropy": 5.644404888153076,
      "epoch": 4.157690795754479,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033833799330620917,
      "loss": 4.9287,
      "mean_token_accuracy": 0.22764657735824584,
      "num_tokens": 82628057.0,
      "step": 37020
    },
    {
      "entropy": 5.655875253677368,
      "epoch": 4.158252372662436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003382994671697647,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22727904021739959,
      "num_tokens": 82639890.0,
      "step": 37025
    },
    {
      "entropy": 5.613913202285767,
      "epoch": 4.158813949570393,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033826093901784065,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2259336993098259,
      "num_tokens": 82651770.0,
      "step": 37030
    },
    {
      "entropy": 5.644197082519531,
      "epoch": 4.159375526478351,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033822240885166384,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23037231713533401,
      "num_tokens": 82662159.0,
      "step": 37035
    },
    {
      "entropy": 5.67485466003418,
      "epoch": 4.159937103386309,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033818387667246095,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2326274558901787,
      "num_tokens": 82674348.0,
      "step": 37040
    },
    {
      "entropy": 5.664977025985718,
      "epoch": 4.160498680294267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000338145342481459,
      "loss": 4.9169,
      "mean_token_accuracy": 0.23487394154071808,
      "num_tokens": 82685674.0,
      "step": 37045
    },
    {
      "entropy": 5.744552135467529,
      "epoch": 4.161060257202224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003381068062798848,
      "loss": 5.0342,
      "mean_token_accuracy": 0.21588269919157027,
      "num_tokens": 82696992.0,
      "step": 37050
    },
    {
      "entropy": 5.753073787689209,
      "epoch": 4.161621834110181,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003380682680689654,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22501230537891387,
      "num_tokens": 82708741.0,
      "step": 37055
    },
    {
      "entropy": 5.662639617919922,
      "epoch": 4.162183411018139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033802972784992784,
      "loss": 4.9425,
      "mean_token_accuracy": 0.22651711404323577,
      "num_tokens": 82720687.0,
      "step": 37060
    },
    {
      "entropy": 5.612665224075317,
      "epoch": 4.1627449879260965,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003379911856239994,
      "loss": 4.9378,
      "mean_token_accuracy": 0.22756485491991044,
      "num_tokens": 82731863.0,
      "step": 37065
    },
    {
      "entropy": 5.624758625030518,
      "epoch": 4.163306564834054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003379526413924071,
      "loss": 4.8939,
      "mean_token_accuracy": 0.23332503139972688,
      "num_tokens": 82742931.0,
      "step": 37070
    },
    {
      "entropy": 5.688798666000366,
      "epoch": 4.163868141742012,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003379140951563784,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22772988975048064,
      "num_tokens": 82754244.0,
      "step": 37075
    },
    {
      "entropy": 5.6182033061981205,
      "epoch": 4.164429718649969,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033787554691714033,
      "loss": 4.8981,
      "mean_token_accuracy": 0.22454496324062348,
      "num_tokens": 82764929.0,
      "step": 37080
    },
    {
      "entropy": 5.622734642028808,
      "epoch": 4.164991295557926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00033783699667592066,
      "loss": 4.8631,
      "mean_token_accuracy": 0.2365850642323494,
      "num_tokens": 82775767.0,
      "step": 37085
    },
    {
      "entropy": 5.625581169128418,
      "epoch": 4.165552872465884,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033779844443394656,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22672875672578813,
      "num_tokens": 82786902.0,
      "step": 37090
    },
    {
      "entropy": 5.749755048751831,
      "epoch": 4.166114449373842,
      "grad_norm": 1.28125,
      "learning_rate": 0.00033775989019244556,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22929200083017348,
      "num_tokens": 82797677.0,
      "step": 37095
    },
    {
      "entropy": 5.733327770233155,
      "epoch": 4.1666760262818,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033772133395264535,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22410861551761627,
      "num_tokens": 82807246.0,
      "step": 37100
    },
    {
      "entropy": 5.687112474441529,
      "epoch": 4.167237603189757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033768277571577364,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22956956475973128,
      "num_tokens": 82818983.0,
      "step": 37105
    },
    {
      "entropy": 5.630052137374878,
      "epoch": 4.167799180097714,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033764421548305795,
      "loss": 4.8922,
      "mean_token_accuracy": 0.23497757613658904,
      "num_tokens": 82829340.0,
      "step": 37110
    },
    {
      "entropy": 5.6473533630371096,
      "epoch": 4.168360757005672,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033760565325572614,
      "loss": 4.9537,
      "mean_token_accuracy": 0.22150865346193313,
      "num_tokens": 82840741.0,
      "step": 37115
    },
    {
      "entropy": 5.691431093215942,
      "epoch": 4.168922333913629,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003375670890350061,
      "loss": 4.938,
      "mean_token_accuracy": 0.2270449236035347,
      "num_tokens": 82851284.0,
      "step": 37120
    },
    {
      "entropy": 5.534418964385987,
      "epoch": 4.169483910821587,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033752852282212557,
      "loss": 4.7464,
      "mean_token_accuracy": 0.2360808953642845,
      "num_tokens": 82861856.0,
      "step": 37125
    },
    {
      "entropy": 5.689512586593628,
      "epoch": 4.170045487729545,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003374899546183128,
      "loss": 5.0064,
      "mean_token_accuracy": 0.22077351808547974,
      "num_tokens": 82873354.0,
      "step": 37130
    },
    {
      "entropy": 5.659934425354004,
      "epoch": 4.170607064637502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033745138442479544,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22701056748628617,
      "num_tokens": 82883891.0,
      "step": 37135
    },
    {
      "entropy": 5.675482845306396,
      "epoch": 4.17116864154546,
      "grad_norm": 1.25,
      "learning_rate": 0.0003374128122428019,
      "loss": 4.8439,
      "mean_token_accuracy": 0.2348674476146698,
      "num_tokens": 82893685.0,
      "step": 37140
    },
    {
      "entropy": 5.620890378952026,
      "epoch": 4.171730218453417,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003373742380735602,
      "loss": 4.9258,
      "mean_token_accuracy": 0.23908283263444902,
      "num_tokens": 82904947.0,
      "step": 37145
    },
    {
      "entropy": 5.634110164642334,
      "epoch": 4.1722917953613745,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003373356619182985,
      "loss": 5.0184,
      "mean_token_accuracy": 0.22283550798892976,
      "num_tokens": 82915368.0,
      "step": 37150
    },
    {
      "entropy": 5.655040979385376,
      "epoch": 4.172853372269333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003372970837782453,
      "loss": 4.897,
      "mean_token_accuracy": 0.22708261460065843,
      "num_tokens": 82926594.0,
      "step": 37155
    },
    {
      "entropy": 5.643754196166992,
      "epoch": 4.17341494917729,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003372585036546287,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23409923315048217,
      "num_tokens": 82937699.0,
      "step": 37160
    },
    {
      "entropy": 5.673220872879028,
      "epoch": 4.173976526085247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003372199215486772,
      "loss": 4.9865,
      "mean_token_accuracy": 0.22732727825641633,
      "num_tokens": 82948822.0,
      "step": 37165
    },
    {
      "entropy": 5.672927570343018,
      "epoch": 4.174538102993205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003371813374616192,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2249355599284172,
      "num_tokens": 82959597.0,
      "step": 37170
    },
    {
      "entropy": 5.699595308303833,
      "epoch": 4.175099679901162,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033714275139468336,
      "loss": 4.9498,
      "mean_token_accuracy": 0.2257800057530403,
      "num_tokens": 82970740.0,
      "step": 37175
    },
    {
      "entropy": 5.745236921310425,
      "epoch": 4.17566125680912,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003371041633490982,
      "loss": 5.0023,
      "mean_token_accuracy": 0.22012673914432526,
      "num_tokens": 82982923.0,
      "step": 37180
    },
    {
      "entropy": 5.584204816818238,
      "epoch": 4.176222833717078,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003370655733260924,
      "loss": 4.8749,
      "mean_token_accuracy": 0.2265549883246422,
      "num_tokens": 82993271.0,
      "step": 37185
    },
    {
      "entropy": 5.682012367248535,
      "epoch": 4.176784410625035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033702698132689467,
      "loss": 4.9061,
      "mean_token_accuracy": 0.23330171704292296,
      "num_tokens": 83004688.0,
      "step": 37190
    },
    {
      "entropy": 5.667706918716431,
      "epoch": 4.177345987532993,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003369883873527338,
      "loss": 4.993,
      "mean_token_accuracy": 0.22238563597202302,
      "num_tokens": 83016310.0,
      "step": 37195
    },
    {
      "entropy": 5.739289569854736,
      "epoch": 4.17790756444095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033694979140483865,
      "loss": 5.0005,
      "mean_token_accuracy": 0.22475082725286483,
      "num_tokens": 83027327.0,
      "step": 37200
    },
    {
      "entropy": 5.679482793807983,
      "epoch": 4.1784691413489075,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003369111934844381,
      "loss": 4.956,
      "mean_token_accuracy": 0.22567104697227477,
      "num_tokens": 83038493.0,
      "step": 37205
    },
    {
      "entropy": 5.639640998840332,
      "epoch": 4.179030718256866,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033687259359276115,
      "loss": 4.8717,
      "mean_token_accuracy": 0.231198987364769,
      "num_tokens": 83050082.0,
      "step": 37210
    },
    {
      "entropy": 5.62581877708435,
      "epoch": 4.179592295164823,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003368339917310368,
      "loss": 4.868,
      "mean_token_accuracy": 0.232979454100132,
      "num_tokens": 83061234.0,
      "step": 37215
    },
    {
      "entropy": 5.6532752990722654,
      "epoch": 4.18015387207278,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003367953879004942,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22799071967601775,
      "num_tokens": 83071454.0,
      "step": 37220
    },
    {
      "entropy": 5.683274793624878,
      "epoch": 4.180715448980738,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003367567821023625,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22592619955539703,
      "num_tokens": 83083204.0,
      "step": 37225
    },
    {
      "entropy": 5.64222469329834,
      "epoch": 4.181277025888695,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003367181743378708,
      "loss": 4.884,
      "mean_token_accuracy": 0.23264414221048355,
      "num_tokens": 83094654.0,
      "step": 37230
    },
    {
      "entropy": 5.72699990272522,
      "epoch": 4.181838602796653,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033667956460824866,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2227422460913658,
      "num_tokens": 83105539.0,
      "step": 37235
    },
    {
      "entropy": 5.674021434783936,
      "epoch": 4.182400179704611,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003366409529147251,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2302113026380539,
      "num_tokens": 83115741.0,
      "step": 37240
    },
    {
      "entropy": 5.674088382720948,
      "epoch": 4.182961756612568,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033660233925852977,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23317456543445586,
      "num_tokens": 83126739.0,
      "step": 37245
    },
    {
      "entropy": 5.6030162334442135,
      "epoch": 4.183523333520526,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033656372364089206,
      "loss": 4.8257,
      "mean_token_accuracy": 0.24182742983102798,
      "num_tokens": 83138821.0,
      "step": 37250
    },
    {
      "entropy": 5.681314992904663,
      "epoch": 4.184084910428483,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033652510606304155,
      "loss": 4.9946,
      "mean_token_accuracy": 0.22268572598695754,
      "num_tokens": 83149757.0,
      "step": 37255
    },
    {
      "entropy": 5.662266683578491,
      "epoch": 4.18464648733644,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033648648652620766,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22266867160797119,
      "num_tokens": 83161627.0,
      "step": 37260
    },
    {
      "entropy": 5.6449565410614015,
      "epoch": 4.1852080642443985,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033644786503162036,
      "loss": 4.9016,
      "mean_token_accuracy": 0.23284645527601242,
      "num_tokens": 83172903.0,
      "step": 37265
    },
    {
      "entropy": 5.568699741363526,
      "epoch": 4.185769641152356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003364092415805091,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23712618350982667,
      "num_tokens": 83184142.0,
      "step": 37270
    },
    {
      "entropy": 5.6847024917602536,
      "epoch": 4.186331218060313,
      "grad_norm": 1.125,
      "learning_rate": 0.00033637061617410377,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22544221580028534,
      "num_tokens": 83195206.0,
      "step": 37275
    },
    {
      "entropy": 5.66923246383667,
      "epoch": 4.186892794968271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003363319888136343,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23602731376886368,
      "num_tokens": 83207579.0,
      "step": 37280
    },
    {
      "entropy": 5.706981611251831,
      "epoch": 4.187454371876228,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003362933595003305,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22509384751319886,
      "num_tokens": 83219964.0,
      "step": 37285
    },
    {
      "entropy": 5.618392276763916,
      "epoch": 4.188015948784186,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003362547282354223,
      "loss": 4.8859,
      "mean_token_accuracy": 0.23147504031658173,
      "num_tokens": 83231241.0,
      "step": 37290
    },
    {
      "entropy": 5.626342725753784,
      "epoch": 4.188577525692144,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003362160950201398,
      "loss": 4.861,
      "mean_token_accuracy": 0.2384617358446121,
      "num_tokens": 83241784.0,
      "step": 37295
    },
    {
      "entropy": 5.61759934425354,
      "epoch": 4.189139102600101,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003361774598557132,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23246927708387374,
      "num_tokens": 83252815.0,
      "step": 37300
    },
    {
      "entropy": 5.625223922729492,
      "epoch": 4.189700679508059,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033613882274337244,
      "loss": 4.9387,
      "mean_token_accuracy": 0.2326512321829796,
      "num_tokens": 83263938.0,
      "step": 37305
    },
    {
      "entropy": 5.721153402328492,
      "epoch": 4.190262256416016,
      "grad_norm": 1.25,
      "learning_rate": 0.0003361001836843479,
      "loss": 4.9997,
      "mean_token_accuracy": 0.22058703005313873,
      "num_tokens": 83274746.0,
      "step": 37310
    },
    {
      "entropy": 5.735708284378052,
      "epoch": 4.190823833323973,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003360615426798697,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2169579565525055,
      "num_tokens": 83285707.0,
      "step": 37315
    },
    {
      "entropy": 5.751377296447754,
      "epoch": 4.1913854102319315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003360228997311685,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22991060167551042,
      "num_tokens": 83297175.0,
      "step": 37320
    },
    {
      "entropy": 5.66524977684021,
      "epoch": 4.191946987139889,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003359842548394744,
      "loss": 4.8326,
      "mean_token_accuracy": 0.23846022337675093,
      "num_tokens": 83307731.0,
      "step": 37325
    },
    {
      "entropy": 5.679303455352783,
      "epoch": 4.192508564047847,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003359456080060179,
      "loss": 4.9255,
      "mean_token_accuracy": 0.2311746969819069,
      "num_tokens": 83318698.0,
      "step": 37330
    },
    {
      "entropy": 5.568059968948364,
      "epoch": 4.193070140955804,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003359069592320297,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22726290971040725,
      "num_tokens": 83330085.0,
      "step": 37335
    },
    {
      "entropy": 5.7077028274536135,
      "epoch": 4.193631717863761,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003358683085187403,
      "loss": 4.9755,
      "mean_token_accuracy": 0.2276623472571373,
      "num_tokens": 83340606.0,
      "step": 37340
    },
    {
      "entropy": 5.639530324935913,
      "epoch": 4.194193294771719,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033582965586738037,
      "loss": 4.907,
      "mean_token_accuracy": 0.22841438353061677,
      "num_tokens": 83352177.0,
      "step": 37345
    },
    {
      "entropy": 5.715217447280883,
      "epoch": 4.1947548716796765,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003357910012791804,
      "loss": 5.0236,
      "mean_token_accuracy": 0.22119660675525665,
      "num_tokens": 83362474.0,
      "step": 37350
    },
    {
      "entropy": 5.681159734725952,
      "epoch": 4.195316448587634,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003357523447553716,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22901953607797623,
      "num_tokens": 83373187.0,
      "step": 37355
    },
    {
      "entropy": 5.622400760650635,
      "epoch": 4.195878025495592,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033571368629718457,
      "loss": 4.8631,
      "mean_token_accuracy": 0.23143758177757262,
      "num_tokens": 83385110.0,
      "step": 37360
    },
    {
      "entropy": 5.686730527877808,
      "epoch": 4.196439602403549,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033567502590585007,
      "loss": 4.9706,
      "mean_token_accuracy": 0.23058436810970306,
      "num_tokens": 83396692.0,
      "step": 37365
    },
    {
      "entropy": 5.691313648223877,
      "epoch": 4.197001179311506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003356363635825994,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23407007604837418,
      "num_tokens": 83408433.0,
      "step": 37370
    },
    {
      "entropy": 5.728839969635009,
      "epoch": 4.197562756219464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003355976993286633,
      "loss": 4.9706,
      "mean_token_accuracy": 0.23295272141695023,
      "num_tokens": 83419169.0,
      "step": 37375
    },
    {
      "entropy": 5.703592824935913,
      "epoch": 4.198124333127422,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003355590331452729,
      "loss": 4.9284,
      "mean_token_accuracy": 0.23809717893600463,
      "num_tokens": 83430247.0,
      "step": 37380
    },
    {
      "entropy": 5.655669784545898,
      "epoch": 4.19868591003538,
      "grad_norm": 1.125,
      "learning_rate": 0.00033552036503365945,
      "loss": 4.8363,
      "mean_token_accuracy": 0.2356538414955139,
      "num_tokens": 83442294.0,
      "step": 37385
    },
    {
      "entropy": 5.581719589233399,
      "epoch": 4.199247486943337,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003354816949950541,
      "loss": 4.881,
      "mean_token_accuracy": 0.23822848051786422,
      "num_tokens": 83453388.0,
      "step": 37390
    },
    {
      "entropy": 5.5489531517028805,
      "epoch": 4.199809063851294,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033544302303068814,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23528313040733337,
      "num_tokens": 83465151.0,
      "step": 37395
    },
    {
      "entropy": 5.652197265625,
      "epoch": 4.200370640759252,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003354043491417929,
      "loss": 4.9137,
      "mean_token_accuracy": 0.23272707760334016,
      "num_tokens": 83474731.0,
      "step": 37400
    },
    {
      "entropy": 5.630314016342163,
      "epoch": 4.2009322176672095,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033536567332959966,
      "loss": 4.8958,
      "mean_token_accuracy": 0.2333659842610359,
      "num_tokens": 83486456.0,
      "step": 37405
    },
    {
      "entropy": 5.65318078994751,
      "epoch": 4.201493794575167,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033532699559534,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22968435138463975,
      "num_tokens": 83496503.0,
      "step": 37410
    },
    {
      "entropy": 5.754896831512451,
      "epoch": 4.202055371483125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033528831594024553,
      "loss": 5.0005,
      "mean_token_accuracy": 0.23038333803415298,
      "num_tokens": 83508198.0,
      "step": 37415
    },
    {
      "entropy": 5.714136075973511,
      "epoch": 4.202616948391082,
      "grad_norm": 1.125,
      "learning_rate": 0.00033524963436554753,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22796969562768937,
      "num_tokens": 83520255.0,
      "step": 37420
    },
    {
      "entropy": 5.697954273223877,
      "epoch": 4.20317852529904,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033521095087247794,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2253190979361534,
      "num_tokens": 83532083.0,
      "step": 37425
    },
    {
      "entropy": 5.706497049331665,
      "epoch": 4.203740102206997,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003351722654622683,
      "loss": 5.0156,
      "mean_token_accuracy": 0.22399199306964873,
      "num_tokens": 83542891.0,
      "step": 37430
    },
    {
      "entropy": 5.656608915328979,
      "epoch": 4.204301679114955,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003351335781361503,
      "loss": 4.9086,
      "mean_token_accuracy": 0.23566219359636306,
      "num_tokens": 83554238.0,
      "step": 37435
    },
    {
      "entropy": 5.677778482437134,
      "epoch": 4.204863256022913,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033509488889535595,
      "loss": 4.981,
      "mean_token_accuracy": 0.22300977110862732,
      "num_tokens": 83565681.0,
      "step": 37440
    },
    {
      "entropy": 5.579102277755737,
      "epoch": 4.20542483293087,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033505619774111696,
      "loss": 4.7864,
      "mean_token_accuracy": 0.240905325114727,
      "num_tokens": 83576844.0,
      "step": 37445
    },
    {
      "entropy": 5.652520990371704,
      "epoch": 4.205986409838827,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033501750467466535,
      "loss": 4.9004,
      "mean_token_accuracy": 0.24073131531476974,
      "num_tokens": 83589501.0,
      "step": 37450
    },
    {
      "entropy": 5.69446325302124,
      "epoch": 4.206547986746785,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003349788096972331,
      "loss": 4.9561,
      "mean_token_accuracy": 0.23695818185806275,
      "num_tokens": 83600305.0,
      "step": 37455
    },
    {
      "entropy": 5.702171564102173,
      "epoch": 4.207109563654742,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003349401128100523,
      "loss": 4.9861,
      "mean_token_accuracy": 0.22058844864368438,
      "num_tokens": 83611908.0,
      "step": 37460
    },
    {
      "entropy": 5.640995264053345,
      "epoch": 4.2076711405627,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033490141401435516,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22630503475666047,
      "num_tokens": 83624156.0,
      "step": 37465
    },
    {
      "entropy": 5.625569915771484,
      "epoch": 4.208232717470658,
      "grad_norm": 1.25,
      "learning_rate": 0.0003348627133113736,
      "loss": 4.84,
      "mean_token_accuracy": 0.23651403039693833,
      "num_tokens": 83634655.0,
      "step": 37470
    },
    {
      "entropy": 5.5982917785644535,
      "epoch": 4.208794294378615,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033482401070234006,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23615521788597107,
      "num_tokens": 83645015.0,
      "step": 37475
    },
    {
      "entropy": 5.624991798400879,
      "epoch": 4.209355871286573,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000334785306188487,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2334776282310486,
      "num_tokens": 83656540.0,
      "step": 37480
    },
    {
      "entropy": 5.6714599609375,
      "epoch": 4.20991744819453,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003347465997710464,
      "loss": 4.9103,
      "mean_token_accuracy": 0.23519267588853837,
      "num_tokens": 83666798.0,
      "step": 37485
    },
    {
      "entropy": 5.619138193130493,
      "epoch": 4.2104790251024875,
      "grad_norm": 1.171875,
      "learning_rate": 0.000334707891451251,
      "loss": 4.8853,
      "mean_token_accuracy": 0.23125506043434144,
      "num_tokens": 83677058.0,
      "step": 37490
    },
    {
      "entropy": 5.557712507247925,
      "epoch": 4.211040602010446,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003346691812303331,
      "loss": 4.8228,
      "mean_token_accuracy": 0.23474613726139068,
      "num_tokens": 83688126.0,
      "step": 37495
    },
    {
      "entropy": 5.681519269943237,
      "epoch": 4.211602178918403,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003346304691095254,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21965523958206176,
      "num_tokens": 83700298.0,
      "step": 37500
    },
    {
      "entropy": 5.637903642654419,
      "epoch": 4.21216375582636,
      "grad_norm": 1.25,
      "learning_rate": 0.0003345917550900604,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2314053788781166,
      "num_tokens": 83711037.0,
      "step": 37505
    },
    {
      "entropy": 5.685969066619873,
      "epoch": 4.212725332734318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033455303917317075,
      "loss": 4.9043,
      "mean_token_accuracy": 0.23162678629159927,
      "num_tokens": 83722306.0,
      "step": 37510
    },
    {
      "entropy": 5.679928779602051,
      "epoch": 4.213286909642275,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033451432136008937,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2316527009010315,
      "num_tokens": 83733306.0,
      "step": 37515
    },
    {
      "entropy": 5.572290658950806,
      "epoch": 4.2138484865502335,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003344756016520488,
      "loss": 4.8741,
      "mean_token_accuracy": 0.232879239320755,
      "num_tokens": 83745475.0,
      "step": 37520
    },
    {
      "entropy": 5.639034748077393,
      "epoch": 4.214410063458191,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033443688005028213,
      "loss": 4.8597,
      "mean_token_accuracy": 0.231907020509243,
      "num_tokens": 83756992.0,
      "step": 37525
    },
    {
      "entropy": 5.635828638076783,
      "epoch": 4.214971640366148,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000334398156556022,
      "loss": 4.878,
      "mean_token_accuracy": 0.22745184153318404,
      "num_tokens": 83768881.0,
      "step": 37530
    },
    {
      "entropy": 5.786276054382324,
      "epoch": 4.215533217274106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003343594311705016,
      "loss": 4.9692,
      "mean_token_accuracy": 0.23176799565553666,
      "num_tokens": 83781144.0,
      "step": 37535
    },
    {
      "entropy": 5.7225548267364506,
      "epoch": 4.216094794182063,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033432070389495386,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2233831837773323,
      "num_tokens": 83792449.0,
      "step": 37540
    },
    {
      "entropy": 5.621266937255859,
      "epoch": 4.2166563710900204,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003342819747306119,
      "loss": 4.8323,
      "mean_token_accuracy": 0.23563109934329987,
      "num_tokens": 83802889.0,
      "step": 37545
    },
    {
      "entropy": 5.70510892868042,
      "epoch": 4.2172179479979786,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033424324367870896,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2257535070180893,
      "num_tokens": 83813527.0,
      "step": 37550
    },
    {
      "entropy": 5.569364929199219,
      "epoch": 4.217779524905936,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033420451074047807,
      "loss": 4.8698,
      "mean_token_accuracy": 0.23322071731090546,
      "num_tokens": 83824963.0,
      "step": 37555
    },
    {
      "entropy": 5.655016851425171,
      "epoch": 4.218341101813893,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003341657759171526,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22549538761377336,
      "num_tokens": 83836188.0,
      "step": 37560
    },
    {
      "entropy": 5.694364023208618,
      "epoch": 4.218902678721851,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003341270392099659,
      "loss": 4.9284,
      "mean_token_accuracy": 0.23603588491678237,
      "num_tokens": 83848297.0,
      "step": 37565
    },
    {
      "entropy": 5.62755560874939,
      "epoch": 4.219464255629808,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033408830062015135,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2255405828356743,
      "num_tokens": 83859074.0,
      "step": 37570
    },
    {
      "entropy": 5.6045817852020265,
      "epoch": 4.220025832537766,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033404956014894237,
      "loss": 4.8761,
      "mean_token_accuracy": 0.23078409284353257,
      "num_tokens": 83869921.0,
      "step": 37575
    },
    {
      "entropy": 5.647591400146484,
      "epoch": 4.220587409445724,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033401081779757254,
      "loss": 4.9235,
      "mean_token_accuracy": 0.23061399012804032,
      "num_tokens": 83881994.0,
      "step": 37580
    },
    {
      "entropy": 5.669177007675171,
      "epoch": 4.221148986353681,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003339720735672753,
      "loss": 4.931,
      "mean_token_accuracy": 0.22602382004261018,
      "num_tokens": 83893530.0,
      "step": 37585
    },
    {
      "entropy": 5.67723240852356,
      "epoch": 4.221710563261639,
      "grad_norm": 1.125,
      "learning_rate": 0.00033393332745928453,
      "loss": 4.885,
      "mean_token_accuracy": 0.22969469130039216,
      "num_tokens": 83904929.0,
      "step": 37590
    },
    {
      "entropy": 5.621298933029175,
      "epoch": 4.222272140169596,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033389457947483357,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23240747153759003,
      "num_tokens": 83915888.0,
      "step": 37595
    },
    {
      "entropy": 5.611521291732788,
      "epoch": 4.222833717077553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033385582961515645,
      "loss": 4.9324,
      "mean_token_accuracy": 0.23298579603433608,
      "num_tokens": 83927481.0,
      "step": 37600
    },
    {
      "entropy": 5.6972421169281,
      "epoch": 4.2233952939855115,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033381707788148683,
      "loss": 4.9781,
      "mean_token_accuracy": 0.23098000138998032,
      "num_tokens": 83938644.0,
      "step": 37605
    },
    {
      "entropy": 5.617418813705444,
      "epoch": 4.223956870893469,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033377832427505864,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2426986038684845,
      "num_tokens": 83949165.0,
      "step": 37610
    },
    {
      "entropy": 5.668229055404663,
      "epoch": 4.224518447801427,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033373956879710593,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22731857746839523,
      "num_tokens": 83961112.0,
      "step": 37615
    },
    {
      "entropy": 5.6911139488220215,
      "epoch": 4.225080024709384,
      "grad_norm": 1.125,
      "learning_rate": 0.00033370081144886246,
      "loss": 4.9602,
      "mean_token_accuracy": 0.23210545182228087,
      "num_tokens": 83972451.0,
      "step": 37620
    },
    {
      "entropy": 5.572857475280761,
      "epoch": 4.225641601617341,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033366205223156237,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23507682532072066,
      "num_tokens": 83984082.0,
      "step": 37625
    },
    {
      "entropy": 5.718644380569458,
      "epoch": 4.226203178525299,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033362329114643987,
      "loss": 5.0477,
      "mean_token_accuracy": 0.2076541393995285,
      "num_tokens": 83995853.0,
      "step": 37630
    },
    {
      "entropy": 5.682709455490112,
      "epoch": 4.226764755433257,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033358452819472895,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2278224527835846,
      "num_tokens": 84006440.0,
      "step": 37635
    },
    {
      "entropy": 5.771839380264282,
      "epoch": 4.227326332341214,
      "grad_norm": 1.0,
      "learning_rate": 0.00033354576337766396,
      "loss": 5.0606,
      "mean_token_accuracy": 0.22384627014398575,
      "num_tokens": 84018968.0,
      "step": 37640
    },
    {
      "entropy": 5.676143217086792,
      "epoch": 4.227887909249172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003335069966964792,
      "loss": 4.9191,
      "mean_token_accuracy": 0.23349919617176057,
      "num_tokens": 84030607.0,
      "step": 37645
    },
    {
      "entropy": 5.638049030303955,
      "epoch": 4.228449486157129,
      "grad_norm": 1.09375,
      "learning_rate": 0.000333468228152409,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23202400505542756,
      "num_tokens": 84042598.0,
      "step": 37650
    },
    {
      "entropy": 5.593410205841065,
      "epoch": 4.229011063065086,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033342945774668766,
      "loss": 4.8224,
      "mean_token_accuracy": 0.23475658446550368,
      "num_tokens": 84052708.0,
      "step": 37655
    },
    {
      "entropy": 5.702109336853027,
      "epoch": 4.2295726399730444,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033339068548054974,
      "loss": 5.0068,
      "mean_token_accuracy": 0.22740055620670319,
      "num_tokens": 84063573.0,
      "step": 37660
    },
    {
      "entropy": 5.750990915298462,
      "epoch": 4.230134216881002,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033335191135522974,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22360191196203233,
      "num_tokens": 84074132.0,
      "step": 37665
    },
    {
      "entropy": 5.555358409881592,
      "epoch": 4.23069579378896,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033331313537196235,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2562327668070793,
      "num_tokens": 84086424.0,
      "step": 37670
    },
    {
      "entropy": 5.692412185668945,
      "epoch": 4.231257370696917,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000333274357531982,
      "loss": 4.9609,
      "mean_token_accuracy": 0.2274410158395767,
      "num_tokens": 84098413.0,
      "step": 37675
    },
    {
      "entropy": 5.522421836853027,
      "epoch": 4.231818947604874,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003332355778365235,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2304209589958191,
      "num_tokens": 84109156.0,
      "step": 37680
    },
    {
      "entropy": 5.648580646514892,
      "epoch": 4.232380524512832,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003331967962868217,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23170194327831267,
      "num_tokens": 84120677.0,
      "step": 37685
    },
    {
      "entropy": 5.657796573638916,
      "epoch": 4.2329421014207895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003331580128841112,
      "loss": 4.9241,
      "mean_token_accuracy": 0.22625960409641266,
      "num_tokens": 84132607.0,
      "step": 37690
    },
    {
      "entropy": 5.688043308258057,
      "epoch": 4.233503678328747,
      "grad_norm": 1.25,
      "learning_rate": 0.00033311922762962716,
      "loss": 4.9304,
      "mean_token_accuracy": 0.23075832724571227,
      "num_tokens": 84143958.0,
      "step": 37695
    },
    {
      "entropy": 5.783470106124878,
      "epoch": 4.234065255236705,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003330804405246043,
      "loss": 5.028,
      "mean_token_accuracy": 0.22813902497291566,
      "num_tokens": 84155154.0,
      "step": 37700
    },
    {
      "entropy": 5.681801986694336,
      "epoch": 4.234626832144662,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003330416515702777,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2217710182070732,
      "num_tokens": 84166100.0,
      "step": 37705
    },
    {
      "entropy": 5.70458025932312,
      "epoch": 4.23518840905262,
      "grad_norm": 1.25,
      "learning_rate": 0.00033300286076788237,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22962362915277482,
      "num_tokens": 84177249.0,
      "step": 37710
    },
    {
      "entropy": 5.699748611450195,
      "epoch": 4.235749985960577,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003329640681186535,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22830138653516768,
      "num_tokens": 84187922.0,
      "step": 37715
    },
    {
      "entropy": 5.676024341583252,
      "epoch": 4.236311562868535,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033292527362382626,
      "loss": 4.9308,
      "mean_token_accuracy": 0.232828289270401,
      "num_tokens": 84198929.0,
      "step": 37720
    },
    {
      "entropy": 5.59347825050354,
      "epoch": 4.236873139776493,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00033288647728463576,
      "loss": 4.9344,
      "mean_token_accuracy": 0.23133157193660736,
      "num_tokens": 84209503.0,
      "step": 37725
    },
    {
      "entropy": 5.674815082550049,
      "epoch": 4.23743471668445,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033284767910231745,
      "loss": 4.992,
      "mean_token_accuracy": 0.2210989147424698,
      "num_tokens": 84220667.0,
      "step": 37730
    },
    {
      "entropy": 5.710185813903808,
      "epoch": 4.237996293592407,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003328088790781065,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2296345502138138,
      "num_tokens": 84232072.0,
      "step": 37735
    },
    {
      "entropy": 5.7236706733703615,
      "epoch": 4.238557870500365,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033277007721323854,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2297132521867752,
      "num_tokens": 84243765.0,
      "step": 37740
    },
    {
      "entropy": 5.711098337173462,
      "epoch": 4.2391194474083225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033273127350894884,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2235092341899872,
      "num_tokens": 84254869.0,
      "step": 37745
    },
    {
      "entropy": 5.656642150878906,
      "epoch": 4.23968102431628,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033269246796647296,
      "loss": 5.0044,
      "mean_token_accuracy": 0.22384048402309417,
      "num_tokens": 84264992.0,
      "step": 37750
    },
    {
      "entropy": 5.605458545684814,
      "epoch": 4.240242601224238,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003326536605870466,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23481839448213576,
      "num_tokens": 84275648.0,
      "step": 37755
    },
    {
      "entropy": 5.75646448135376,
      "epoch": 4.240804178132195,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033261485137190533,
      "loss": 4.921,
      "mean_token_accuracy": 0.23267664909362792,
      "num_tokens": 84287217.0,
      "step": 37760
    },
    {
      "entropy": 5.676743745803833,
      "epoch": 4.241365755040153,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033257604032228485,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2179018422961235,
      "num_tokens": 84298764.0,
      "step": 37765
    },
    {
      "entropy": 5.720919942855835,
      "epoch": 4.24192733194811,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033253722743942086,
      "loss": 4.9968,
      "mean_token_accuracy": 0.2256320208311081,
      "num_tokens": 84310016.0,
      "step": 37770
    },
    {
      "entropy": 5.661870288848877,
      "epoch": 4.2424889088560676,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003324984127245492,
      "loss": 4.9397,
      "mean_token_accuracy": 0.22888969630002975,
      "num_tokens": 84321836.0,
      "step": 37775
    },
    {
      "entropy": 5.627084541320801,
      "epoch": 4.243050485764026,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003324595961789058,
      "loss": 4.8347,
      "mean_token_accuracy": 0.23419626653194428,
      "num_tokens": 84332434.0,
      "step": 37780
    },
    {
      "entropy": 5.572172212600708,
      "epoch": 4.243612062671983,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033242077780372664,
      "loss": 4.8655,
      "mean_token_accuracy": 0.24056329876184462,
      "num_tokens": 84343997.0,
      "step": 37785
    },
    {
      "entropy": 5.636068487167359,
      "epoch": 4.24417363957994,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033238195760024766,
      "loss": 4.9076,
      "mean_token_accuracy": 0.23408612310886384,
      "num_tokens": 84354157.0,
      "step": 37790
    },
    {
      "entropy": 5.59657130241394,
      "epoch": 4.244735216487898,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033234313556970485,
      "loss": 4.8235,
      "mean_token_accuracy": 0.23938903510570525,
      "num_tokens": 84365888.0,
      "step": 37795
    },
    {
      "entropy": 5.675789451599121,
      "epoch": 4.245296793395855,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00033230431171333435,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2342504158616066,
      "num_tokens": 84376121.0,
      "step": 37800
    },
    {
      "entropy": 5.586598014831543,
      "epoch": 4.2458583703038135,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003322654860323723,
      "loss": 4.8026,
      "mean_token_accuracy": 0.2348696455359459,
      "num_tokens": 84387318.0,
      "step": 37805
    },
    {
      "entropy": 5.643711566925049,
      "epoch": 4.246419947211771,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033222665852805514,
      "loss": 4.926,
      "mean_token_accuracy": 0.2234077498316765,
      "num_tokens": 84398952.0,
      "step": 37810
    },
    {
      "entropy": 5.656999397277832,
      "epoch": 4.246981524119728,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00033218782920161884,
      "loss": 4.934,
      "mean_token_accuracy": 0.22343915998935698,
      "num_tokens": 84410010.0,
      "step": 37815
    },
    {
      "entropy": 5.670069599151612,
      "epoch": 4.247543101027686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033214899805429984,
      "loss": 4.8927,
      "mean_token_accuracy": 0.23520984798669814,
      "num_tokens": 84421879.0,
      "step": 37820
    },
    {
      "entropy": 5.7362898826599125,
      "epoch": 4.248104677935643,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033211016508733463,
      "loss": 5.017,
      "mean_token_accuracy": 0.22325387895107268,
      "num_tokens": 84432811.0,
      "step": 37825
    },
    {
      "entropy": 5.681524467468262,
      "epoch": 4.2486662548436005,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033207133030195964,
      "loss": 4.985,
      "mean_token_accuracy": 0.23098772019147873,
      "num_tokens": 84444330.0,
      "step": 37830
    },
    {
      "entropy": 5.685081481933594,
      "epoch": 4.249227831751559,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003320324936994113,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2336876004934311,
      "num_tokens": 84455827.0,
      "step": 37835
    },
    {
      "entropy": 5.656625652313233,
      "epoch": 4.249789408659516,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033199365528092625,
      "loss": 4.8579,
      "mean_token_accuracy": 0.2335599109530449,
      "num_tokens": 84465608.0,
      "step": 37840
    },
    {
      "entropy": 5.649042797088623,
      "epoch": 4.250350985567474,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00033195481504774113,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2336771547794342,
      "num_tokens": 84476142.0,
      "step": 37845
    },
    {
      "entropy": 5.680358409881592,
      "epoch": 4.250912562475431,
      "grad_norm": 1.25,
      "learning_rate": 0.00033191597300109263,
      "loss": 4.958,
      "mean_token_accuracy": 0.22207680940628052,
      "num_tokens": 84486496.0,
      "step": 37850
    },
    {
      "entropy": 5.6240315437316895,
      "epoch": 4.251474139383388,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033187712914221744,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2380757823586464,
      "num_tokens": 84497529.0,
      "step": 37855
    },
    {
      "entropy": 5.604462957382202,
      "epoch": 4.2520357162913465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003318382834723524,
      "loss": 4.854,
      "mean_token_accuracy": 0.23306032717227937,
      "num_tokens": 84508016.0,
      "step": 37860
    },
    {
      "entropy": 5.768501615524292,
      "epoch": 4.252597293199304,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033179943599273445,
      "loss": 5.0485,
      "mean_token_accuracy": 0.22566299587488176,
      "num_tokens": 84519787.0,
      "step": 37865
    },
    {
      "entropy": 5.644883823394776,
      "epoch": 4.253158870107261,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003317605867046003,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22660932689905167,
      "num_tokens": 84530908.0,
      "step": 37870
    },
    {
      "entropy": 5.77052001953125,
      "epoch": 4.253720447015219,
      "grad_norm": 1.1875,
      "learning_rate": 0.00033172173560918716,
      "loss": 5.0513,
      "mean_token_accuracy": 0.21849098354578017,
      "num_tokens": 84543502.0,
      "step": 37875
    },
    {
      "entropy": 5.616072416305542,
      "epoch": 4.254282023923176,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033168288270773197,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22835788428783416,
      "num_tokens": 84554500.0,
      "step": 37880
    },
    {
      "entropy": 5.7173905849456785,
      "epoch": 4.254843600831133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033164402800147177,
      "loss": 5.0749,
      "mean_token_accuracy": 0.22321690917015075,
      "num_tokens": 84566704.0,
      "step": 37885
    },
    {
      "entropy": 5.664034271240235,
      "epoch": 4.2554051777390915,
      "grad_norm": 1.25,
      "learning_rate": 0.0003316051714916436,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22951393723487853,
      "num_tokens": 84578264.0,
      "step": 37890
    },
    {
      "entropy": 5.603421258926391,
      "epoch": 4.255966754647049,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00033156631317948505,
      "loss": 4.7259,
      "mean_token_accuracy": 0.2477874457836151,
      "num_tokens": 84588513.0,
      "step": 37895
    },
    {
      "entropy": 5.692032670974731,
      "epoch": 4.256528331555007,
      "grad_norm": 1.140625,
      "learning_rate": 0.000331527453066233,
      "loss": 4.8922,
      "mean_token_accuracy": 0.24095731377601623,
      "num_tokens": 84600384.0,
      "step": 37900
    },
    {
      "entropy": 5.692369270324707,
      "epoch": 4.257089908462964,
      "grad_norm": 1.265625,
      "learning_rate": 0.000331488591153125,
      "loss": 4.9538,
      "mean_token_accuracy": 0.22826341539621353,
      "num_tokens": 84610827.0,
      "step": 37905
    },
    {
      "entropy": 5.647129058837891,
      "epoch": 4.257651485370921,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003314497274413983,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23677753061056137,
      "num_tokens": 84621130.0,
      "step": 37910
    },
    {
      "entropy": 5.68380389213562,
      "epoch": 4.258213062278879,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033141086193229043,
      "loss": 4.9646,
      "mean_token_accuracy": 0.23180609345436096,
      "num_tokens": 84633428.0,
      "step": 37915
    },
    {
      "entropy": 5.668544816970825,
      "epoch": 4.258774639186837,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003313719946270388,
      "loss": 4.9401,
      "mean_token_accuracy": 0.227055099606514,
      "num_tokens": 84644519.0,
      "step": 37920
    },
    {
      "entropy": 5.7208575248718265,
      "epoch": 4.259336216094794,
      "grad_norm": 1.265625,
      "learning_rate": 0.00033133312552688094,
      "loss": 5.0037,
      "mean_token_accuracy": 0.22537384331226348,
      "num_tokens": 84654752.0,
      "step": 37925
    },
    {
      "entropy": 5.671205282211304,
      "epoch": 4.259897793002752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033129425463305465,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23885730355978013,
      "num_tokens": 84666384.0,
      "step": 37930
    },
    {
      "entropy": 5.767321586608887,
      "epoch": 4.260459369910709,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003312553819467974,
      "loss": 5.0364,
      "mean_token_accuracy": 0.22542980462312698,
      "num_tokens": 84676630.0,
      "step": 37935
    },
    {
      "entropy": 5.585523414611816,
      "epoch": 4.261020946818666,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033121650746934693,
      "loss": 4.892,
      "mean_token_accuracy": 0.23488040119409562,
      "num_tokens": 84689407.0,
      "step": 37940
    },
    {
      "entropy": 5.64865517616272,
      "epoch": 4.2615825237266245,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003311776312019412,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2239397644996643,
      "num_tokens": 84699237.0,
      "step": 37945
    },
    {
      "entropy": 5.671033430099487,
      "epoch": 4.262144100634582,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003311387531458177,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23196094781160354,
      "num_tokens": 84710074.0,
      "step": 37950
    },
    {
      "entropy": 5.670414924621582,
      "epoch": 4.26270567754254,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003310998733022146,
      "loss": 5.0667,
      "mean_token_accuracy": 0.21384451240301133,
      "num_tokens": 84721315.0,
      "step": 37955
    },
    {
      "entropy": 5.73902587890625,
      "epoch": 4.263267254450497,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003310609916723698,
      "loss": 4.9758,
      "mean_token_accuracy": 0.23489325642585754,
      "num_tokens": 84732654.0,
      "step": 37960
    },
    {
      "entropy": 5.7319409370422365,
      "epoch": 4.263828831358454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003310221082575213,
      "loss": 5.0012,
      "mean_token_accuracy": 0.23487063348293305,
      "num_tokens": 84745568.0,
      "step": 37965
    },
    {
      "entropy": 5.684169912338257,
      "epoch": 4.264390408266412,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033098322305890715,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2211553707718849,
      "num_tokens": 84756443.0,
      "step": 37970
    },
    {
      "entropy": 5.7223427295684814,
      "epoch": 4.26495198517437,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033094433607776545,
      "loss": 4.9443,
      "mean_token_accuracy": 0.24143864065408707,
      "num_tokens": 84766789.0,
      "step": 37975
    },
    {
      "entropy": 5.706956386566162,
      "epoch": 4.265513562082327,
      "grad_norm": 1.171875,
      "learning_rate": 0.00033090544731533434,
      "loss": 5.0937,
      "mean_token_accuracy": 0.2187914326786995,
      "num_tokens": 84778186.0,
      "step": 37980
    },
    {
      "entropy": 5.6581014633178714,
      "epoch": 4.266075138990285,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003308665567728522,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22366230487823485,
      "num_tokens": 84788431.0,
      "step": 37985
    },
    {
      "entropy": 5.705530118942261,
      "epoch": 4.266636715898242,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033082766445155715,
      "loss": 5.0182,
      "mean_token_accuracy": 0.23200305849313735,
      "num_tokens": 84799881.0,
      "step": 37990
    },
    {
      "entropy": 5.7115209102630615,
      "epoch": 4.2671982928062,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033078877035268766,
      "loss": 4.9739,
      "mean_token_accuracy": 0.230252642929554,
      "num_tokens": 84810710.0,
      "step": 37995
    },
    {
      "entropy": 5.73603081703186,
      "epoch": 4.267759869714157,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033074987447748205,
      "loss": 4.8984,
      "mean_token_accuracy": 0.23947446942329406,
      "num_tokens": 84821932.0,
      "step": 38000
    },
    {
      "entropy": 5.6135234355926515,
      "epoch": 4.268321446622115,
      "grad_norm": 1.15625,
      "learning_rate": 0.00033071097682717883,
      "loss": 4.8913,
      "mean_token_accuracy": 0.23574347645044327,
      "num_tokens": 84832662.0,
      "step": 38005
    },
    {
      "entropy": 5.704634523391723,
      "epoch": 4.268883023530073,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003306720774030165,
      "loss": 4.9421,
      "mean_token_accuracy": 0.23519121408462523,
      "num_tokens": 84843199.0,
      "step": 38010
    },
    {
      "entropy": 5.6393890380859375,
      "epoch": 4.26944460043803,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003306331762062336,
      "loss": 4.907,
      "mean_token_accuracy": 0.22878590524196624,
      "num_tokens": 84853367.0,
      "step": 38015
    },
    {
      "entropy": 5.648377323150635,
      "epoch": 4.270006177345987,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003305942732380689,
      "loss": 4.9185,
      "mean_token_accuracy": 0.23490276485681533,
      "num_tokens": 84863527.0,
      "step": 38020
    },
    {
      "entropy": 5.599504041671753,
      "epoch": 4.270567754253945,
      "grad_norm": 1.125,
      "learning_rate": 0.0003305553684997609,
      "loss": 4.8665,
      "mean_token_accuracy": 0.23219102770090103,
      "num_tokens": 84874929.0,
      "step": 38025
    },
    {
      "entropy": 5.647496128082276,
      "epoch": 4.2711293311619025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033051646199254843,
      "loss": 4.8802,
      "mean_token_accuracy": 0.23311600983142852,
      "num_tokens": 84885182.0,
      "step": 38030
    },
    {
      "entropy": 5.7058816909790036,
      "epoch": 4.271690908069861,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003304775537176702,
      "loss": 4.9356,
      "mean_token_accuracy": 0.22447499781847,
      "num_tokens": 84896115.0,
      "step": 38035
    },
    {
      "entropy": 5.645781660079956,
      "epoch": 4.272252484977818,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003304386436763652,
      "loss": 4.9569,
      "mean_token_accuracy": 0.23130519688129425,
      "num_tokens": 84908174.0,
      "step": 38040
    },
    {
      "entropy": 5.708269786834717,
      "epoch": 4.272814061885775,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003303997318698723,
      "loss": 4.9655,
      "mean_token_accuracy": 0.2244805097579956,
      "num_tokens": 84920893.0,
      "step": 38045
    },
    {
      "entropy": 5.699571037292481,
      "epoch": 4.273375638793733,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003303608182994304,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2274586394429207,
      "num_tokens": 84931860.0,
      "step": 38050
    },
    {
      "entropy": 5.645113754272461,
      "epoch": 4.27393721570169,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003303219029662787,
      "loss": 4.884,
      "mean_token_accuracy": 0.245127634704113,
      "num_tokens": 84942533.0,
      "step": 38055
    },
    {
      "entropy": 5.562093496322632,
      "epoch": 4.274498792609648,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033028298587165606,
      "loss": 4.8002,
      "mean_token_accuracy": 0.2359966367483139,
      "num_tokens": 84954018.0,
      "step": 38060
    },
    {
      "entropy": 5.7111962795257565,
      "epoch": 4.275060369517606,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003302440670168017,
      "loss": 5.0588,
      "mean_token_accuracy": 0.22145013362169266,
      "num_tokens": 84966150.0,
      "step": 38065
    },
    {
      "entropy": 5.737330102920533,
      "epoch": 4.275621946425563,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003302051464029548,
      "loss": 4.9219,
      "mean_token_accuracy": 0.22842443734407425,
      "num_tokens": 84977668.0,
      "step": 38070
    },
    {
      "entropy": 5.6702745914459225,
      "epoch": 4.27618352333352,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00033016622403135473,
      "loss": 4.866,
      "mean_token_accuracy": 0.22796558290719987,
      "num_tokens": 84988735.0,
      "step": 38075
    },
    {
      "entropy": 5.632670783996582,
      "epoch": 4.276745100241478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033012729990324063,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22691508531570434,
      "num_tokens": 84999573.0,
      "step": 38080
    },
    {
      "entropy": 5.595296144485474,
      "epoch": 4.2773066771494355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003300883740198519,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22868525236845016,
      "num_tokens": 85010582.0,
      "step": 38085
    },
    {
      "entropy": 5.6009117603302006,
      "epoch": 4.277868254057394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033004944638242805,
      "loss": 4.8639,
      "mean_token_accuracy": 0.23750175386667252,
      "num_tokens": 85021563.0,
      "step": 38090
    },
    {
      "entropy": 5.750196409225464,
      "epoch": 4.278429830965351,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003300105169922084,
      "loss": 5.1561,
      "mean_token_accuracy": 0.2141709879040718,
      "num_tokens": 85033231.0,
      "step": 38095
    },
    {
      "entropy": 5.610163688659668,
      "epoch": 4.278991407873308,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003299715858504326,
      "loss": 4.8264,
      "mean_token_accuracy": 0.23597215116024017,
      "num_tokens": 85044110.0,
      "step": 38100
    },
    {
      "entropy": 5.640511465072632,
      "epoch": 4.279552984781266,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032993265295834013,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23157468140125276,
      "num_tokens": 85055425.0,
      "step": 38105
    },
    {
      "entropy": 5.712062835693359,
      "epoch": 4.280114561689223,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003298937183171707,
      "loss": 4.9808,
      "mean_token_accuracy": 0.23086508363485336,
      "num_tokens": 85067358.0,
      "step": 38110
    },
    {
      "entropy": 5.751607084274292,
      "epoch": 4.2806761385971805,
      "grad_norm": 1.09375,
      "learning_rate": 0.000329854781928164,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22861648947000504,
      "num_tokens": 85078967.0,
      "step": 38115
    },
    {
      "entropy": 5.637835502624512,
      "epoch": 4.281237715505139,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003298158437925598,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23065239936113358,
      "num_tokens": 85088515.0,
      "step": 38120
    },
    {
      "entropy": 5.5992614269256595,
      "epoch": 4.281799292413096,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003297769039115978,
      "loss": 4.8708,
      "mean_token_accuracy": 0.23302016109228135,
      "num_tokens": 85099535.0,
      "step": 38125
    },
    {
      "entropy": 5.780270433425903,
      "epoch": 4.282360869321053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032973796228651797,
      "loss": 5.1016,
      "mean_token_accuracy": 0.2157081037759781,
      "num_tokens": 85110818.0,
      "step": 38130
    },
    {
      "entropy": 5.678874444961548,
      "epoch": 4.282922446229011,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032969901891856027,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2251797467470169,
      "num_tokens": 85120362.0,
      "step": 38135
    },
    {
      "entropy": 5.645415878295898,
      "epoch": 4.283484023136968,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032966007380896443,
      "loss": 4.9317,
      "mean_token_accuracy": 0.22910252064466477,
      "num_tokens": 85131358.0,
      "step": 38140
    },
    {
      "entropy": 5.692908811569214,
      "epoch": 4.2840456000449265,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003296211269589708,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23477453738451004,
      "num_tokens": 85143395.0,
      "step": 38145
    },
    {
      "entropy": 5.6457947254180905,
      "epoch": 4.284607176952884,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003295821783698192,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23853930383920668,
      "num_tokens": 85155654.0,
      "step": 38150
    },
    {
      "entropy": 5.63053936958313,
      "epoch": 4.285168753860841,
      "grad_norm": 1.28125,
      "learning_rate": 0.00032954322804274995,
      "loss": 5.0374,
      "mean_token_accuracy": 0.21962418556213378,
      "num_tokens": 85168285.0,
      "step": 38155
    },
    {
      "entropy": 5.686619520187378,
      "epoch": 4.285730330768799,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032950427597900315,
      "loss": 5.0124,
      "mean_token_accuracy": 0.22409551441669465,
      "num_tokens": 85180058.0,
      "step": 38160
    },
    {
      "entropy": 5.678700876235962,
      "epoch": 4.286291907676756,
      "grad_norm": 1.203125,
      "learning_rate": 0.000329465322179819,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2263276234269142,
      "num_tokens": 85190990.0,
      "step": 38165
    },
    {
      "entropy": 5.767411470413208,
      "epoch": 4.2868534845847135,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032942636664643794,
      "loss": 4.9973,
      "mean_token_accuracy": 0.22513622790575027,
      "num_tokens": 85201136.0,
      "step": 38170
    },
    {
      "entropy": 5.702991676330567,
      "epoch": 4.287415061492672,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032938740938010024,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2297070026397705,
      "num_tokens": 85212424.0,
      "step": 38175
    },
    {
      "entropy": 5.729250907897949,
      "epoch": 4.287976638400629,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003293484503820463,
      "loss": 5.0059,
      "mean_token_accuracy": 0.22471451312303542,
      "num_tokens": 85223767.0,
      "step": 38180
    },
    {
      "entropy": 5.757835817337036,
      "epoch": 4.288538215308587,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032930948965351664,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2288326010107994,
      "num_tokens": 85234374.0,
      "step": 38185
    },
    {
      "entropy": 5.631551361083984,
      "epoch": 4.289099792216544,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003292705271957518,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2313947394490242,
      "num_tokens": 85245322.0,
      "step": 38190
    },
    {
      "entropy": 5.745029020309448,
      "epoch": 4.289661369124501,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032923156300999223,
      "loss": 5.0156,
      "mean_token_accuracy": 0.22503956854343415,
      "num_tokens": 85257114.0,
      "step": 38195
    },
    {
      "entropy": 5.717083072662353,
      "epoch": 4.2902229460324595,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032919259709747874,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2292761519551277,
      "num_tokens": 85267564.0,
      "step": 38200
    },
    {
      "entropy": 5.643003225326538,
      "epoch": 4.290784522940417,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032915362945945195,
      "loss": 4.8837,
      "mean_token_accuracy": 0.23021963685750962,
      "num_tokens": 85277763.0,
      "step": 38205
    },
    {
      "entropy": 5.704035949707031,
      "epoch": 4.291346099848374,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032911466009715254,
      "loss": 5.0547,
      "mean_token_accuracy": 0.22600134313106537,
      "num_tokens": 85288504.0,
      "step": 38210
    },
    {
      "entropy": 5.69498291015625,
      "epoch": 4.291907676756332,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032907568901182136,
      "loss": 4.9125,
      "mean_token_accuracy": 0.23989814519882202,
      "num_tokens": 85299928.0,
      "step": 38215
    },
    {
      "entropy": 5.665208339691162,
      "epoch": 4.292469253664289,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003290367162046993,
      "loss": 4.9243,
      "mean_token_accuracy": 0.23272018134593964,
      "num_tokens": 85311052.0,
      "step": 38220
    },
    {
      "entropy": 5.692674207687378,
      "epoch": 4.293030830572247,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032899774167702723,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2286960780620575,
      "num_tokens": 85322023.0,
      "step": 38225
    },
    {
      "entropy": 5.613058567047119,
      "epoch": 4.2935924074802045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003289587654300461,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22893297672271729,
      "num_tokens": 85334160.0,
      "step": 38230
    },
    {
      "entropy": 5.742772054672241,
      "epoch": 4.294153984388162,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000328919787464997,
      "loss": 5.065,
      "mean_token_accuracy": 0.21868728399276732,
      "num_tokens": 85345378.0,
      "step": 38235
    },
    {
      "entropy": 5.690897798538208,
      "epoch": 4.29471556129612,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000328880807783121,
      "loss": 4.927,
      "mean_token_accuracy": 0.23221334367990493,
      "num_tokens": 85357610.0,
      "step": 38240
    },
    {
      "entropy": 5.700384998321534,
      "epoch": 4.295277138204077,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032884182638565914,
      "loss": 4.9071,
      "mean_token_accuracy": 0.2265010416507721,
      "num_tokens": 85368749.0,
      "step": 38245
    },
    {
      "entropy": 5.680515384674072,
      "epoch": 4.295838715112034,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032880284327385255,
      "loss": 4.9621,
      "mean_token_accuracy": 0.23067781031131745,
      "num_tokens": 85379488.0,
      "step": 38250
    },
    {
      "entropy": 5.648318099975586,
      "epoch": 4.296400292019992,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003287638584489426,
      "loss": 4.9155,
      "mean_token_accuracy": 0.23141598105430602,
      "num_tokens": 85390581.0,
      "step": 38255
    },
    {
      "entropy": 5.615615367889404,
      "epoch": 4.29696186892795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003287248719121706,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22422979176044464,
      "num_tokens": 85402430.0,
      "step": 38260
    },
    {
      "entropy": 5.641195011138916,
      "epoch": 4.297523445835907,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032868588366477785,
      "loss": 4.823,
      "mean_token_accuracy": 0.23447318524122238,
      "num_tokens": 85414932.0,
      "step": 38265
    },
    {
      "entropy": 5.6777793884277346,
      "epoch": 4.298085022743865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032864689370800564,
      "loss": 4.8639,
      "mean_token_accuracy": 0.24093710482120514,
      "num_tokens": 85426484.0,
      "step": 38270
    },
    {
      "entropy": 5.6333106517791744,
      "epoch": 4.298646599651822,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003286079020430957,
      "loss": 4.894,
      "mean_token_accuracy": 0.22916826009750366,
      "num_tokens": 85436864.0,
      "step": 38275
    },
    {
      "entropy": 5.583886814117432,
      "epoch": 4.29920817655978,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032856890867128926,
      "loss": 4.8486,
      "mean_token_accuracy": 0.2374309405684471,
      "num_tokens": 85446303.0,
      "step": 38280
    },
    {
      "entropy": 5.643993473052978,
      "epoch": 4.2997697534677375,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003285299135938279,
      "loss": 4.9233,
      "mean_token_accuracy": 0.23461612761020662,
      "num_tokens": 85457218.0,
      "step": 38285
    },
    {
      "entropy": 5.678148508071899,
      "epoch": 4.300331330375695,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032849091681195347,
      "loss": 4.9305,
      "mean_token_accuracy": 0.2299821585416794,
      "num_tokens": 85469085.0,
      "step": 38290
    },
    {
      "entropy": 5.645841503143311,
      "epoch": 4.300892907283653,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032845191832690747,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23288535326719284,
      "num_tokens": 85478966.0,
      "step": 38295
    },
    {
      "entropy": 5.613875532150269,
      "epoch": 4.30145448419161,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003284129181399317,
      "loss": 4.8782,
      "mean_token_accuracy": 0.24126698225736617,
      "num_tokens": 85489742.0,
      "step": 38300
    },
    {
      "entropy": 5.582136583328247,
      "epoch": 4.302016061099567,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003283739162522678,
      "loss": 4.8141,
      "mean_token_accuracy": 0.24592140763998033,
      "num_tokens": 85500766.0,
      "step": 38305
    },
    {
      "entropy": 5.65252833366394,
      "epoch": 4.302577638007525,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003283349126651578,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22838716208934784,
      "num_tokens": 85512206.0,
      "step": 38310
    },
    {
      "entropy": 5.653087568283081,
      "epoch": 4.303139214915483,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032829590737984345,
      "loss": 4.9135,
      "mean_token_accuracy": 0.22968345433473586,
      "num_tokens": 85523881.0,
      "step": 38315
    },
    {
      "entropy": 5.608842849731445,
      "epoch": 4.30370079182344,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032825690039756675,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22977137714624404,
      "num_tokens": 85535130.0,
      "step": 38320
    },
    {
      "entropy": 5.722570848464966,
      "epoch": 4.304262368731398,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032821789171956966,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22737012356519698,
      "num_tokens": 85547261.0,
      "step": 38325
    },
    {
      "entropy": 5.690861082077026,
      "epoch": 4.304823945639355,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003281788813470943,
      "loss": 4.9516,
      "mean_token_accuracy": 0.22678851336240768,
      "num_tokens": 85558300.0,
      "step": 38330
    },
    {
      "entropy": 5.607230997085571,
      "epoch": 4.305385522547313,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003281398692813827,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22664233297109604,
      "num_tokens": 85569370.0,
      "step": 38335
    },
    {
      "entropy": 5.609987020492554,
      "epoch": 4.30594709945527,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000328100855523677,
      "loss": 4.8991,
      "mean_token_accuracy": 0.234385347366333,
      "num_tokens": 85579527.0,
      "step": 38340
    },
    {
      "entropy": 5.68495020866394,
      "epoch": 4.306508676363228,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032806184007521955,
      "loss": 4.9284,
      "mean_token_accuracy": 0.2273807033896446,
      "num_tokens": 85590791.0,
      "step": 38345
    },
    {
      "entropy": 5.630884075164795,
      "epoch": 4.307070253271186,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003280228229372525,
      "loss": 4.8357,
      "mean_token_accuracy": 0.2350939095020294,
      "num_tokens": 85601210.0,
      "step": 38350
    },
    {
      "entropy": 5.65572304725647,
      "epoch": 4.307631830179143,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003279838041110182,
      "loss": 4.9464,
      "mean_token_accuracy": 0.22638535648584365,
      "num_tokens": 85611811.0,
      "step": 38355
    },
    {
      "entropy": 5.654628801345825,
      "epoch": 4.3081934070871,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032794478359775904,
      "loss": 5.0444,
      "mean_token_accuracy": 0.2192263573408127,
      "num_tokens": 85622046.0,
      "step": 38360
    },
    {
      "entropy": 5.70190200805664,
      "epoch": 4.308754983995058,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032790576139871743,
      "loss": 4.9303,
      "mean_token_accuracy": 0.23599915504455565,
      "num_tokens": 85633094.0,
      "step": 38365
    },
    {
      "entropy": 5.652146673202514,
      "epoch": 4.3093165609030155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003278667375151359,
      "loss": 4.8578,
      "mean_token_accuracy": 0.228016696870327,
      "num_tokens": 85644224.0,
      "step": 38370
    },
    {
      "entropy": 5.642113113403321,
      "epoch": 4.309878137810974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032782771194825685,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22471849173307418,
      "num_tokens": 85654950.0,
      "step": 38375
    },
    {
      "entropy": 5.662839317321778,
      "epoch": 4.310439714718931,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032778868469932304,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22848953753709794,
      "num_tokens": 85665250.0,
      "step": 38380
    },
    {
      "entropy": 5.7209837436676025,
      "epoch": 4.311001291626888,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032774965576957705,
      "loss": 5.0105,
      "mean_token_accuracy": 0.22580251544713975,
      "num_tokens": 85676856.0,
      "step": 38385
    },
    {
      "entropy": 5.828666019439697,
      "epoch": 4.311562868534846,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003277106251602615,
      "loss": 5.0998,
      "mean_token_accuracy": 0.22184984833002092,
      "num_tokens": 85689733.0,
      "step": 38390
    },
    {
      "entropy": 5.688221979141235,
      "epoch": 4.312124445442803,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032767159287261914,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2368386283516884,
      "num_tokens": 85699678.0,
      "step": 38395
    },
    {
      "entropy": 5.703643131256103,
      "epoch": 4.312686022350761,
      "grad_norm": 1.28125,
      "learning_rate": 0.000327632558907893,
      "loss": 5.0402,
      "mean_token_accuracy": 0.22128933668136597,
      "num_tokens": 85712130.0,
      "step": 38400
    },
    {
      "entropy": 5.6810445308685305,
      "epoch": 4.313247599258719,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032759352326732573,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23623427450656892,
      "num_tokens": 85723025.0,
      "step": 38405
    },
    {
      "entropy": 5.626816749572754,
      "epoch": 4.313809176166676,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032755448595216027,
      "loss": 4.8863,
      "mean_token_accuracy": 0.23321671634912491,
      "num_tokens": 85733551.0,
      "step": 38410
    },
    {
      "entropy": 5.698029041290283,
      "epoch": 4.314370753074634,
      "grad_norm": 1.125,
      "learning_rate": 0.00032751544696363945,
      "loss": 4.9652,
      "mean_token_accuracy": 0.22637950032949447,
      "num_tokens": 85745633.0,
      "step": 38415
    },
    {
      "entropy": 5.65946044921875,
      "epoch": 4.314932329982591,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032747640630300664,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2359657868742943,
      "num_tokens": 85757522.0,
      "step": 38420
    },
    {
      "entropy": 5.739404106140137,
      "epoch": 4.3154939068905485,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032743736397150455,
      "loss": 4.9127,
      "mean_token_accuracy": 0.2405530259013176,
      "num_tokens": 85768666.0,
      "step": 38425
    },
    {
      "entropy": 5.615839958190918,
      "epoch": 4.316055483798507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003273983199703765,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23429313600063323,
      "num_tokens": 85780527.0,
      "step": 38430
    },
    {
      "entropy": 5.587776231765747,
      "epoch": 4.316617060706464,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003273592743008656,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2254778876900673,
      "num_tokens": 85792166.0,
      "step": 38435
    },
    {
      "entropy": 5.746067476272583,
      "epoch": 4.317178637614421,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032732022696421514,
      "loss": 5.0198,
      "mean_token_accuracy": 0.22626525461673735,
      "num_tokens": 85803706.0,
      "step": 38440
    },
    {
      "entropy": 5.721667337417602,
      "epoch": 4.317740214522379,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003272811779616684,
      "loss": 4.9465,
      "mean_token_accuracy": 0.23717185407876967,
      "num_tokens": 85815515.0,
      "step": 38445
    },
    {
      "entropy": 5.685748672485351,
      "epoch": 4.318301791430336,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032724212729446864,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23284097164869308,
      "num_tokens": 85826359.0,
      "step": 38450
    },
    {
      "entropy": 5.685171222686767,
      "epoch": 4.3188633683382935,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003272030749638593,
      "loss": 4.955,
      "mean_token_accuracy": 0.2328107699751854,
      "num_tokens": 85837977.0,
      "step": 38455
    },
    {
      "entropy": 5.714051008224487,
      "epoch": 4.319424945246252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003271640209710837,
      "loss": 4.9898,
      "mean_token_accuracy": 0.22978084534406662,
      "num_tokens": 85850356.0,
      "step": 38460
    },
    {
      "entropy": 5.694748258590698,
      "epoch": 4.319986522154209,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003271249653173855,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22434878647327422,
      "num_tokens": 85863101.0,
      "step": 38465
    },
    {
      "entropy": 5.691779232025146,
      "epoch": 4.320548099062167,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032708590800400826,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2261966124176979,
      "num_tokens": 85873374.0,
      "step": 38470
    },
    {
      "entropy": 5.734991598129272,
      "epoch": 4.321109675970124,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003270468490321955,
      "loss": 5.0316,
      "mean_token_accuracy": 0.22385576516389846,
      "num_tokens": 85886167.0,
      "step": 38475
    },
    {
      "entropy": 5.693853664398193,
      "epoch": 4.321671252878081,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032700778840319086,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22711348980665208,
      "num_tokens": 85897040.0,
      "step": 38480
    },
    {
      "entropy": 5.652790689468384,
      "epoch": 4.3222328297860395,
      "grad_norm": 1.0625,
      "learning_rate": 0.000326968726118238,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22732454687356948,
      "num_tokens": 85908386.0,
      "step": 38485
    },
    {
      "entropy": 5.638051176071167,
      "epoch": 4.322794406693997,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032692966217858074,
      "loss": 4.8889,
      "mean_token_accuracy": 0.24642884582281113,
      "num_tokens": 85920313.0,
      "step": 38490
    },
    {
      "entropy": 5.708235168457032,
      "epoch": 4.323355983601954,
      "grad_norm": 1.125,
      "learning_rate": 0.0003268905965854629,
      "loss": 4.9895,
      "mean_token_accuracy": 0.22628020048141478,
      "num_tokens": 85931273.0,
      "step": 38495
    },
    {
      "entropy": 5.691198539733887,
      "epoch": 4.323917560509912,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032685152934012844,
      "loss": 4.9912,
      "mean_token_accuracy": 0.22587350904941558,
      "num_tokens": 85942952.0,
      "step": 38500
    },
    {
      "entropy": 5.637888050079345,
      "epoch": 4.324479137417869,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003268124604438211,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22721425890922547,
      "num_tokens": 85953331.0,
      "step": 38505
    },
    {
      "entropy": 5.6575017929077145,
      "epoch": 4.3250407143258265,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032677338989778494,
      "loss": 4.9351,
      "mean_token_accuracy": 0.23452587872743608,
      "num_tokens": 85965142.0,
      "step": 38510
    },
    {
      "entropy": 5.7168203830719,
      "epoch": 4.325602291233785,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000326734317703264,
      "loss": 4.9849,
      "mean_token_accuracy": 0.23487377017736435,
      "num_tokens": 85976258.0,
      "step": 38515
    },
    {
      "entropy": 5.668901157379151,
      "epoch": 4.326163868141742,
      "grad_norm": 1.25,
      "learning_rate": 0.0003266952438615023,
      "loss": 4.9507,
      "mean_token_accuracy": 0.2346853107213974,
      "num_tokens": 85987976.0,
      "step": 38520
    },
    {
      "entropy": 5.6182972431182865,
      "epoch": 4.3267254450497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032665616837374397,
      "loss": 4.8252,
      "mean_token_accuracy": 0.2330508217215538,
      "num_tokens": 85999283.0,
      "step": 38525
    },
    {
      "entropy": 5.60768494606018,
      "epoch": 4.327287021957657,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003266170912412332,
      "loss": 4.8742,
      "mean_token_accuracy": 0.23685434311628342,
      "num_tokens": 86010484.0,
      "step": 38530
    },
    {
      "entropy": 5.684122562408447,
      "epoch": 4.327848598865614,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003265780124652143,
      "loss": 4.9922,
      "mean_token_accuracy": 0.22394923716783524,
      "num_tokens": 86021582.0,
      "step": 38535
    },
    {
      "entropy": 5.6729203224182125,
      "epoch": 4.3284101757735725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032653893204693136,
      "loss": 4.9284,
      "mean_token_accuracy": 0.2199084833264351,
      "num_tokens": 86033847.0,
      "step": 38540
    },
    {
      "entropy": 5.719325113296509,
      "epoch": 4.32897175268153,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003264998499876289,
      "loss": 4.9061,
      "mean_token_accuracy": 0.23209196776151658,
      "num_tokens": 86044870.0,
      "step": 38545
    },
    {
      "entropy": 5.59068603515625,
      "epoch": 4.329533329589487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003264607662885512,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23472687155008315,
      "num_tokens": 86055807.0,
      "step": 38550
    },
    {
      "entropy": 5.655610513687134,
      "epoch": 4.330094906497445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032642168095094274,
      "loss": 5.0223,
      "mean_token_accuracy": 0.2235068693757057,
      "num_tokens": 86066869.0,
      "step": 38555
    },
    {
      "entropy": 5.782654523849487,
      "epoch": 4.330656483405402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003263825939760481,
      "loss": 5.0196,
      "mean_token_accuracy": 0.22212929278612137,
      "num_tokens": 86078870.0,
      "step": 38560
    },
    {
      "entropy": 5.615726947784424,
      "epoch": 4.33121806031336,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032634350536511167,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22339271605014802,
      "num_tokens": 86090940.0,
      "step": 38565
    },
    {
      "entropy": 5.614997625350952,
      "epoch": 4.3317796372213175,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032630441511937813,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23741527646780014,
      "num_tokens": 86101171.0,
      "step": 38570
    },
    {
      "entropy": 5.637591123580933,
      "epoch": 4.332341214129275,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032626532324009204,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22430407553911208,
      "num_tokens": 86111614.0,
      "step": 38575
    },
    {
      "entropy": 5.649464845657349,
      "epoch": 4.332902791037233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003262262297284983,
      "loss": 4.9203,
      "mean_token_accuracy": 0.23877224922180176,
      "num_tokens": 86124131.0,
      "step": 38580
    },
    {
      "entropy": 5.681895780563354,
      "epoch": 4.33346436794519,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032618713458584147,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22958499938249588,
      "num_tokens": 86135584.0,
      "step": 38585
    },
    {
      "entropy": 5.656424474716187,
      "epoch": 4.334025944853147,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003261480378133665,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23662806451320648,
      "num_tokens": 86146163.0,
      "step": 38590
    },
    {
      "entropy": 5.6911242485046385,
      "epoch": 4.334587521761105,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003261089394123181,
      "loss": 5.0004,
      "mean_token_accuracy": 0.2256777375936508,
      "num_tokens": 86157199.0,
      "step": 38595
    },
    {
      "entropy": 5.706472253799438,
      "epoch": 4.335149098669063,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003260698393839413,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2323920965194702,
      "num_tokens": 86167523.0,
      "step": 38600
    },
    {
      "entropy": 5.7026666641235355,
      "epoch": 4.335710675577021,
      "grad_norm": 1.125,
      "learning_rate": 0.00032603073772948096,
      "loss": 4.8915,
      "mean_token_accuracy": 0.23307790607213974,
      "num_tokens": 86179341.0,
      "step": 38605
    },
    {
      "entropy": 5.700310897827149,
      "epoch": 4.336272252484978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032599163445018204,
      "loss": 4.9427,
      "mean_token_accuracy": 0.23143519163131715,
      "num_tokens": 86190154.0,
      "step": 38610
    },
    {
      "entropy": 5.633745861053467,
      "epoch": 4.336833829392935,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003259525295472899,
      "loss": 4.921,
      "mean_token_accuracy": 0.22710034996271133,
      "num_tokens": 86201598.0,
      "step": 38615
    },
    {
      "entropy": 5.6680464267730715,
      "epoch": 4.337395406300893,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032591342302204935,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22871873378753663,
      "num_tokens": 86212033.0,
      "step": 38620
    },
    {
      "entropy": 5.633868169784546,
      "epoch": 4.3379569832088505,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003258743148757057,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22249553650617598,
      "num_tokens": 86222931.0,
      "step": 38625
    },
    {
      "entropy": 5.6485700607299805,
      "epoch": 4.338518560116808,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032583520510950405,
      "loss": 4.9078,
      "mean_token_accuracy": 0.2277293846011162,
      "num_tokens": 86234022.0,
      "step": 38630
    },
    {
      "entropy": 5.5756463527679445,
      "epoch": 4.339080137024766,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032579609372468984,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23291218727827073,
      "num_tokens": 86244267.0,
      "step": 38635
    },
    {
      "entropy": 5.626357316970825,
      "epoch": 4.339641713932723,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032575698072250827,
      "loss": 4.8807,
      "mean_token_accuracy": 0.23762170374393463,
      "num_tokens": 86254200.0,
      "step": 38640
    },
    {
      "entropy": 5.669282531738281,
      "epoch": 4.34020329084068,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003257178661042047,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2189281851053238,
      "num_tokens": 86266523.0,
      "step": 38645
    },
    {
      "entropy": 5.725520992279053,
      "epoch": 4.340764867748638,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032567874987102463,
      "loss": 4.9641,
      "mean_token_accuracy": 0.22823361903429032,
      "num_tokens": 86276949.0,
      "step": 38650
    },
    {
      "entropy": 5.719264554977417,
      "epoch": 4.341326444656596,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003256396320242136,
      "loss": 4.9278,
      "mean_token_accuracy": 0.22694727182388305,
      "num_tokens": 86287412.0,
      "step": 38655
    },
    {
      "entropy": 5.684763336181641,
      "epoch": 4.341888021564554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000325600512565017,
      "loss": 4.9125,
      "mean_token_accuracy": 0.24035253375768661,
      "num_tokens": 86298540.0,
      "step": 38660
    },
    {
      "entropy": 5.611533689498901,
      "epoch": 4.342449598472511,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032556139149468046,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2385365590453148,
      "num_tokens": 86309351.0,
      "step": 38665
    },
    {
      "entropy": 5.757372903823852,
      "epoch": 4.343011175380468,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032552226881444956,
      "loss": 4.9852,
      "mean_token_accuracy": 0.22562878727912902,
      "num_tokens": 86320873.0,
      "step": 38670
    },
    {
      "entropy": 5.729717779159546,
      "epoch": 4.343572752288426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032548314452557006,
      "loss": 5.0093,
      "mean_token_accuracy": 0.22489226162433623,
      "num_tokens": 86332889.0,
      "step": 38675
    },
    {
      "entropy": 5.6310834884643555,
      "epoch": 4.344134329196383,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032544401862928766,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23872973918914794,
      "num_tokens": 86344155.0,
      "step": 38680
    },
    {
      "entropy": 5.626909017562866,
      "epoch": 4.344695906104341,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032540489112684804,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23228947222232818,
      "num_tokens": 86355718.0,
      "step": 38685
    },
    {
      "entropy": 5.7237996578216555,
      "epoch": 4.345257483012299,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032536576201949724,
      "loss": 5.0165,
      "mean_token_accuracy": 0.22342369109392166,
      "num_tokens": 86367174.0,
      "step": 38690
    },
    {
      "entropy": 5.695788908004761,
      "epoch": 4.345819059920256,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003253266313084811,
      "loss": 4.9735,
      "mean_token_accuracy": 0.22381363064050674,
      "num_tokens": 86379086.0,
      "step": 38695
    },
    {
      "entropy": 5.749718284606933,
      "epoch": 4.346380636828213,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003252874989950454,
      "loss": 4.9904,
      "mean_token_accuracy": 0.23021087944507598,
      "num_tokens": 86390113.0,
      "step": 38700
    },
    {
      "entropy": 5.786137676239013,
      "epoch": 4.346942213736171,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032524836508043625,
      "loss": 5.0482,
      "mean_token_accuracy": 0.2221962332725525,
      "num_tokens": 86400824.0,
      "step": 38705
    },
    {
      "entropy": 5.668839740753174,
      "epoch": 4.3475037906441285,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032520922956589973,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22860902398824692,
      "num_tokens": 86412972.0,
      "step": 38710
    },
    {
      "entropy": 5.696154403686523,
      "epoch": 4.348065367552087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032517009245268175,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2273327514529228,
      "num_tokens": 86425332.0,
      "step": 38715
    },
    {
      "entropy": 5.658307838439941,
      "epoch": 4.348626944460044,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032513095374202863,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23694006949663163,
      "num_tokens": 86435615.0,
      "step": 38720
    },
    {
      "entropy": 5.731348037719727,
      "epoch": 4.349188521368001,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032509181343518646,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2307240590453148,
      "num_tokens": 86446875.0,
      "step": 38725
    },
    {
      "entropy": 5.6948216438293455,
      "epoch": 4.349750098275959,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00032505267153340163,
      "loss": 4.9149,
      "mean_token_accuracy": 0.23098453134298325,
      "num_tokens": 86458284.0,
      "step": 38730
    },
    {
      "entropy": 5.776831102371216,
      "epoch": 4.350311675183916,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003250135280379202,
      "loss": 5.0456,
      "mean_token_accuracy": 0.21547371447086333,
      "num_tokens": 86470977.0,
      "step": 38735
    },
    {
      "entropy": 5.700163984298706,
      "epoch": 4.350873252091874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003249743829499888,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22644670605659484,
      "num_tokens": 86482578.0,
      "step": 38740
    },
    {
      "entropy": 5.671737098693848,
      "epoch": 4.351434828999832,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003249352362708535,
      "loss": 4.8571,
      "mean_token_accuracy": 0.23486244976520537,
      "num_tokens": 86493123.0,
      "step": 38745
    },
    {
      "entropy": 5.746749782562256,
      "epoch": 4.351996405907789,
      "grad_norm": 1.1875,
      "learning_rate": 0.000324896088001761,
      "loss": 5.0842,
      "mean_token_accuracy": 0.22161634564399718,
      "num_tokens": 86503865.0,
      "step": 38750
    },
    {
      "entropy": 5.63921971321106,
      "epoch": 4.352557982815747,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003248569381439577,
      "loss": 4.9229,
      "mean_token_accuracy": 0.2261818006634712,
      "num_tokens": 86515006.0,
      "step": 38755
    },
    {
      "entropy": 5.648454570770264,
      "epoch": 4.353119559723704,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003248177866986901,
      "loss": 4.9273,
      "mean_token_accuracy": 0.22141261249780655,
      "num_tokens": 86526256.0,
      "step": 38760
    },
    {
      "entropy": 5.672713661193848,
      "epoch": 4.3536811366316615,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003247786336672049,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23465896099805833,
      "num_tokens": 86537050.0,
      "step": 38765
    },
    {
      "entropy": 5.649761295318603,
      "epoch": 4.35424271353962,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032473947905074866,
      "loss": 4.8676,
      "mean_token_accuracy": 0.23692428171634675,
      "num_tokens": 86547678.0,
      "step": 38770
    },
    {
      "entropy": 5.62708158493042,
      "epoch": 4.354804290447577,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032470032285056814,
      "loss": 4.861,
      "mean_token_accuracy": 0.23692874014377593,
      "num_tokens": 86558747.0,
      "step": 38775
    },
    {
      "entropy": 5.61347713470459,
      "epoch": 4.355365867355534,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032466116506791004,
      "loss": 4.9447,
      "mean_token_accuracy": 0.22724704295396805,
      "num_tokens": 86569500.0,
      "step": 38780
    },
    {
      "entropy": 5.607458305358887,
      "epoch": 4.355927444263492,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032462200570402124,
      "loss": 4.9545,
      "mean_token_accuracy": 0.22372222691774368,
      "num_tokens": 86581312.0,
      "step": 38785
    },
    {
      "entropy": 5.693431854248047,
      "epoch": 4.356489021171449,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003245828447601484,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2307392567396164,
      "num_tokens": 86592526.0,
      "step": 38790
    },
    {
      "entropy": 5.663235807418824,
      "epoch": 4.357050598079407,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003245436822375387,
      "loss": 4.9651,
      "mean_token_accuracy": 0.23125898838043213,
      "num_tokens": 86604284.0,
      "step": 38795
    },
    {
      "entropy": 5.619144105911255,
      "epoch": 4.357612174987365,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003245045181374389,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22944531589746475,
      "num_tokens": 86615567.0,
      "step": 38800
    },
    {
      "entropy": 5.613752603530884,
      "epoch": 4.358173751895322,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000324465352461096,
      "loss": 4.8658,
      "mean_token_accuracy": 0.23711239099502562,
      "num_tokens": 86626188.0,
      "step": 38805
    },
    {
      "entropy": 5.731815814971924,
      "epoch": 4.35873532880328,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003244261852097572,
      "loss": 4.9992,
      "mean_token_accuracy": 0.2305176317691803,
      "num_tokens": 86638468.0,
      "step": 38810
    },
    {
      "entropy": 5.7242859363555905,
      "epoch": 4.359296905711237,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032438701638466943,
      "loss": 4.9487,
      "mean_token_accuracy": 0.23149904161691665,
      "num_tokens": 86649119.0,
      "step": 38815
    },
    {
      "entropy": 5.677277565002441,
      "epoch": 4.359858482619194,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032434784598707987,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2308792307972908,
      "num_tokens": 86660924.0,
      "step": 38820
    },
    {
      "entropy": 5.674508857727051,
      "epoch": 4.3604200595271525,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003243086740182358,
      "loss": 4.9373,
      "mean_token_accuracy": 0.23652611076831817,
      "num_tokens": 86672655.0,
      "step": 38825
    },
    {
      "entropy": 5.65248064994812,
      "epoch": 4.36098163643511,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032426950047938436,
      "loss": 4.8801,
      "mean_token_accuracy": 0.23836199343204498,
      "num_tokens": 86684146.0,
      "step": 38830
    },
    {
      "entropy": 5.746495342254638,
      "epoch": 4.361543213343067,
      "grad_norm": 1.1875,
      "learning_rate": 0.000324230325371773,
      "loss": 5.0855,
      "mean_token_accuracy": 0.21987925916910173,
      "num_tokens": 86695435.0,
      "step": 38835
    },
    {
      "entropy": 5.807717561721802,
      "epoch": 4.362104790251025,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003241911486966489,
      "loss": 5.0999,
      "mean_token_accuracy": 0.22024573981761933,
      "num_tokens": 86706387.0,
      "step": 38840
    },
    {
      "entropy": 5.67833137512207,
      "epoch": 4.362666367158982,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003241519704552596,
      "loss": 4.8625,
      "mean_token_accuracy": 0.23106464743614197,
      "num_tokens": 86717661.0,
      "step": 38845
    },
    {
      "entropy": 5.67284197807312,
      "epoch": 4.36322794406694,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032411279064885253,
      "loss": 4.9204,
      "mean_token_accuracy": 0.2294123038649559,
      "num_tokens": 86729045.0,
      "step": 38850
    },
    {
      "entropy": 5.725610303878784,
      "epoch": 4.363789520974898,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032407360927867506,
      "loss": 5.0647,
      "mean_token_accuracy": 0.22228609025478363,
      "num_tokens": 86740763.0,
      "step": 38855
    },
    {
      "entropy": 5.700789928436279,
      "epoch": 4.364351097882855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003240344263459749,
      "loss": 4.9725,
      "mean_token_accuracy": 0.2335651695728302,
      "num_tokens": 86752369.0,
      "step": 38860
    },
    {
      "entropy": 5.679198312759399,
      "epoch": 4.364912674790813,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032399524185199957,
      "loss": 4.977,
      "mean_token_accuracy": 0.22433386445045472,
      "num_tokens": 86765947.0,
      "step": 38865
    },
    {
      "entropy": 5.651057720184326,
      "epoch": 4.36547425169877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032395605579799673,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23348111510276795,
      "num_tokens": 86777202.0,
      "step": 38870
    },
    {
      "entropy": 5.712373924255371,
      "epoch": 4.366035828606727,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032391686818521414,
      "loss": 4.9311,
      "mean_token_accuracy": 0.22789551615715026,
      "num_tokens": 86787838.0,
      "step": 38875
    },
    {
      "entropy": 5.769761562347412,
      "epoch": 4.3665974055146854,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032387767901489946,
      "loss": 5.0515,
      "mean_token_accuracy": 0.22297896295785904,
      "num_tokens": 86799875.0,
      "step": 38880
    },
    {
      "entropy": 5.694175767898559,
      "epoch": 4.367158982422643,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032383848828830054,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22863409370183946,
      "num_tokens": 86811018.0,
      "step": 38885
    },
    {
      "entropy": 5.613268852233887,
      "epoch": 4.3677205593306,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032379929600666525,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2293835774064064,
      "num_tokens": 86821556.0,
      "step": 38890
    },
    {
      "entropy": 5.570349884033203,
      "epoch": 4.368282136238558,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032376010217124133,
      "loss": 4.7991,
      "mean_token_accuracy": 0.2419011726975441,
      "num_tokens": 86832041.0,
      "step": 38895
    },
    {
      "entropy": 5.700942468643189,
      "epoch": 4.368843713146515,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032372090678327694,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23297591656446456,
      "num_tokens": 86844723.0,
      "step": 38900
    },
    {
      "entropy": 5.739451694488525,
      "epoch": 4.369405290054473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032368170984401997,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22899492233991622,
      "num_tokens": 86856211.0,
      "step": 38905
    },
    {
      "entropy": 5.683565950393676,
      "epoch": 4.3699668669624305,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032364251135471846,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22470356822013854,
      "num_tokens": 86867701.0,
      "step": 38910
    },
    {
      "entropy": 5.74428071975708,
      "epoch": 4.370528443870388,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003236033113166205,
      "loss": 5.0611,
      "mean_token_accuracy": 0.22261494547128677,
      "num_tokens": 86879838.0,
      "step": 38915
    },
    {
      "entropy": 5.6987813949584964,
      "epoch": 4.371090020778346,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032356410973097423,
      "loss": 5.0009,
      "mean_token_accuracy": 0.22628870606422424,
      "num_tokens": 86891101.0,
      "step": 38920
    },
    {
      "entropy": 5.720050239562989,
      "epoch": 4.371651597686303,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032352490659902795,
      "loss": 4.9234,
      "mean_token_accuracy": 0.23738712221384048,
      "num_tokens": 86902796.0,
      "step": 38925
    },
    {
      "entropy": 5.693366241455078,
      "epoch": 4.37221317459426,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003234857019220298,
      "loss": 4.9762,
      "mean_token_accuracy": 0.226865154504776,
      "num_tokens": 86914381.0,
      "step": 38930
    },
    {
      "entropy": 5.753331041336059,
      "epoch": 4.372774751502218,
      "grad_norm": 1.140625,
      "learning_rate": 0.000323446495701228,
      "loss": 5.0843,
      "mean_token_accuracy": 0.22148468643426894,
      "num_tokens": 86926082.0,
      "step": 38935
    },
    {
      "entropy": 5.720544528961182,
      "epoch": 4.373336328410176,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003234072879378711,
      "loss": 4.8898,
      "mean_token_accuracy": 0.2338349252939224,
      "num_tokens": 86936644.0,
      "step": 38940
    },
    {
      "entropy": 5.726983451843262,
      "epoch": 4.373897905318134,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003233680786332074,
      "loss": 4.98,
      "mean_token_accuracy": 0.22804564088582993,
      "num_tokens": 86947586.0,
      "step": 38945
    },
    {
      "entropy": 5.727723026275635,
      "epoch": 4.374459482226091,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003233288677884853,
      "loss": 4.9413,
      "mean_token_accuracy": 0.23273568451404572,
      "num_tokens": 86958265.0,
      "step": 38950
    },
    {
      "entropy": 5.7075440883636475,
      "epoch": 4.375021059134048,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003232896554049533,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22593257427215577,
      "num_tokens": 86969119.0,
      "step": 38955
    },
    {
      "entropy": 5.667564630508423,
      "epoch": 4.375582636042006,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032325044148385986,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2218627944588661,
      "num_tokens": 86980232.0,
      "step": 38960
    },
    {
      "entropy": 5.647676277160644,
      "epoch": 4.3761442129499635,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032321122602645374,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23099450021982193,
      "num_tokens": 86993590.0,
      "step": 38965
    },
    {
      "entropy": 5.671650981903076,
      "epoch": 4.376705789857921,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003231720090339834,
      "loss": 4.9636,
      "mean_token_accuracy": 0.22757624536752702,
      "num_tokens": 87004905.0,
      "step": 38970
    },
    {
      "entropy": 5.6374622821807865,
      "epoch": 4.377267366765879,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032313279050769767,
      "loss": 4.924,
      "mean_token_accuracy": 0.23641120940446853,
      "num_tokens": 87016920.0,
      "step": 38975
    },
    {
      "entropy": 5.7788842678070065,
      "epoch": 4.377828943673836,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003230935704488452,
      "loss": 5.0617,
      "mean_token_accuracy": 0.22334562689065934,
      "num_tokens": 87027740.0,
      "step": 38980
    },
    {
      "entropy": 5.740040826797485,
      "epoch": 4.378390520581794,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032305434885867483,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22457153499126434,
      "num_tokens": 87039742.0,
      "step": 38985
    },
    {
      "entropy": 5.617259311676025,
      "epoch": 4.378952097489751,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032301512573843535,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22585297971963883,
      "num_tokens": 87049750.0,
      "step": 38990
    },
    {
      "entropy": 5.701855611801148,
      "epoch": 4.3795136743977086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003229759010893756,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2372233286499977,
      "num_tokens": 87062890.0,
      "step": 38995
    },
    {
      "entropy": 5.700309610366821,
      "epoch": 4.380075251305667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003229366749127446,
      "loss": 4.969,
      "mean_token_accuracy": 0.22609617114067077,
      "num_tokens": 87074450.0,
      "step": 39000
    },
    {
      "epoch": 4.380075251305667,
      "eval_entropy": 5.448645541186427,
      "eval_loss": 5.0601420402526855,
      "eval_mean_token_accuracy": 0.23097824255715088,
      "eval_num_tokens": 87074450.0,
      "eval_runtime": 18.2583,
      "eval_samples_per_second": 1698.573,
      "eval_steps_per_second": 212.342,
      "step": 39000
    },
    {
      "entropy": 5.812345552444458,
      "epoch": 4.380636828213624,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003228974472097913,
      "loss": 5.054,
      "mean_token_accuracy": 0.22925143241882323,
      "num_tokens": 87085931.0,
      "step": 39005
    },
    {
      "entropy": 5.691954708099365,
      "epoch": 4.381198405121581,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032285821798176456,
      "loss": 5.007,
      "mean_token_accuracy": 0.2255108177661896,
      "num_tokens": 87096655.0,
      "step": 39010
    },
    {
      "entropy": 5.733165502548218,
      "epoch": 4.381759982029539,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032281898722991376,
      "loss": 5.0199,
      "mean_token_accuracy": 0.22411002367734909,
      "num_tokens": 87108359.0,
      "step": 39015
    },
    {
      "entropy": 5.765463066101074,
      "epoch": 4.382321558937496,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00032277975495548775,
      "loss": 5.0488,
      "mean_token_accuracy": 0.22149242013692855,
      "num_tokens": 87119028.0,
      "step": 39020
    },
    {
      "entropy": 5.65228271484375,
      "epoch": 4.382883135845454,
      "grad_norm": 1.125,
      "learning_rate": 0.0003227405211597359,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2306696355342865,
      "num_tokens": 87130574.0,
      "step": 39025
    },
    {
      "entropy": 5.673408460617066,
      "epoch": 4.383444712753412,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032270128584390735,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23003991544246674,
      "num_tokens": 87141052.0,
      "step": 39030
    },
    {
      "entropy": 5.697818899154663,
      "epoch": 4.384006289661369,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003226620490092513,
      "loss": 4.9772,
      "mean_token_accuracy": 0.22502797693014145,
      "num_tokens": 87151323.0,
      "step": 39035
    },
    {
      "entropy": 5.663007020950317,
      "epoch": 4.384567866569327,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032262281065701716,
      "loss": 4.8649,
      "mean_token_accuracy": 0.22471119165420533,
      "num_tokens": 87162614.0,
      "step": 39040
    },
    {
      "entropy": 5.662981271743774,
      "epoch": 4.385129443477284,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003225835707884542,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23049423694610596,
      "num_tokens": 87172712.0,
      "step": 39045
    },
    {
      "entropy": 5.710241794586182,
      "epoch": 4.3856910203852415,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003225443294048119,
      "loss": 4.9332,
      "mean_token_accuracy": 0.23029544949531555,
      "num_tokens": 87183612.0,
      "step": 39050
    },
    {
      "entropy": 5.668526649475098,
      "epoch": 4.3862525972932,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032250508650733976,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22911507785320281,
      "num_tokens": 87194707.0,
      "step": 39055
    },
    {
      "entropy": 5.611012649536133,
      "epoch": 4.386814174201157,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003224658420972871,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2372429296374321,
      "num_tokens": 87206451.0,
      "step": 39060
    },
    {
      "entropy": 5.675416278839111,
      "epoch": 4.387375751109114,
      "grad_norm": 1.125,
      "learning_rate": 0.0003224265961759038,
      "loss": 5.0484,
      "mean_token_accuracy": 0.22669326812028884,
      "num_tokens": 87217542.0,
      "step": 39065
    },
    {
      "entropy": 5.625747203826904,
      "epoch": 4.387937328017072,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003223873487444392,
      "loss": 4.8792,
      "mean_token_accuracy": 0.23653120994567872,
      "num_tokens": 87228399.0,
      "step": 39070
    },
    {
      "entropy": 5.700998783111572,
      "epoch": 4.388498904925029,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003223480998041431,
      "loss": 4.974,
      "mean_token_accuracy": 0.22767748236656188,
      "num_tokens": 87239875.0,
      "step": 39075
    },
    {
      "entropy": 5.708790397644043,
      "epoch": 4.389060481832987,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032230884935626505,
      "loss": 4.9848,
      "mean_token_accuracy": 0.22672781795263292,
      "num_tokens": 87252919.0,
      "step": 39080
    },
    {
      "entropy": 5.625307178497314,
      "epoch": 4.389622058740945,
      "grad_norm": 1.140625,
      "learning_rate": 0.000322269597402055,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2314080998301506,
      "num_tokens": 87264458.0,
      "step": 39085
    },
    {
      "entropy": 5.720674848556518,
      "epoch": 4.390183635648902,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003222303439427625,
      "loss": 4.9553,
      "mean_token_accuracy": 0.23435897380113602,
      "num_tokens": 87274935.0,
      "step": 39090
    },
    {
      "entropy": 5.641097545623779,
      "epoch": 4.39074521255686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032219108897963767,
      "loss": 4.8875,
      "mean_token_accuracy": 0.23632501512765886,
      "num_tokens": 87285870.0,
      "step": 39095
    },
    {
      "entropy": 5.648829936981201,
      "epoch": 4.391306789464817,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003221518325139302,
      "loss": 4.8826,
      "mean_token_accuracy": 0.23014523833990097,
      "num_tokens": 87297374.0,
      "step": 39100
    },
    {
      "entropy": 5.715194606781006,
      "epoch": 4.3918683663727744,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003221125745468901,
      "loss": 4.9112,
      "mean_token_accuracy": 0.22761713713407516,
      "num_tokens": 87307597.0,
      "step": 39105
    },
    {
      "entropy": 5.633510875701904,
      "epoch": 4.3924299432807326,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003220733150797673,
      "loss": 4.856,
      "mean_token_accuracy": 0.2338206097483635,
      "num_tokens": 87318459.0,
      "step": 39110
    },
    {
      "entropy": 5.757958030700683,
      "epoch": 4.39299152018869,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003220340541138119,
      "loss": 5.1009,
      "mean_token_accuracy": 0.21639736592769623,
      "num_tokens": 87330613.0,
      "step": 39115
    },
    {
      "entropy": 5.745051670074463,
      "epoch": 4.393553097096647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032199479165027407,
      "loss": 5.0124,
      "mean_token_accuracy": 0.22604095041751862,
      "num_tokens": 87341988.0,
      "step": 39120
    },
    {
      "entropy": 5.68265585899353,
      "epoch": 4.394114674004605,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032195552769040376,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22985781878232955,
      "num_tokens": 87353532.0,
      "step": 39125
    },
    {
      "entropy": 5.716541576385498,
      "epoch": 4.394676250912562,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003219162622354513,
      "loss": 4.9731,
      "mean_token_accuracy": 0.23125041425228118,
      "num_tokens": 87364754.0,
      "step": 39130
    },
    {
      "entropy": 5.620986413955689,
      "epoch": 4.39523782782052,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032187699528666674,
      "loss": 4.9376,
      "mean_token_accuracy": 0.22477574795484542,
      "num_tokens": 87375983.0,
      "step": 39135
    },
    {
      "entropy": 5.67942795753479,
      "epoch": 4.395799404728478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003218377268453006,
      "loss": 4.9181,
      "mean_token_accuracy": 0.23906583189964295,
      "num_tokens": 87388169.0,
      "step": 39140
    },
    {
      "entropy": 5.760351037979126,
      "epoch": 4.396360981636435,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000321798456912603,
      "loss": 4.9737,
      "mean_token_accuracy": 0.234090593457222,
      "num_tokens": 87399341.0,
      "step": 39145
    },
    {
      "entropy": 5.672761249542236,
      "epoch": 4.396922558544393,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032175918548982437,
      "loss": 4.9054,
      "mean_token_accuracy": 0.23419155180454254,
      "num_tokens": 87410941.0,
      "step": 39150
    },
    {
      "entropy": 5.615921258926392,
      "epoch": 4.39748413545235,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032171991257821513,
      "loss": 4.8864,
      "mean_token_accuracy": 0.23135495334863662,
      "num_tokens": 87421833.0,
      "step": 39155
    },
    {
      "entropy": 5.584819793701172,
      "epoch": 4.398045712360307,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003216806381790258,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23483153581619262,
      "num_tokens": 87431648.0,
      "step": 39160
    },
    {
      "entropy": 5.710288715362549,
      "epoch": 4.3986072892682655,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003216413622935069,
      "loss": 5.0146,
      "mean_token_accuracy": 0.22409615367650987,
      "num_tokens": 87444207.0,
      "step": 39165
    },
    {
      "entropy": 5.659712791442871,
      "epoch": 4.399168866176223,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032160208492290886,
      "loss": 4.8501,
      "mean_token_accuracy": 0.2379395991563797,
      "num_tokens": 87455919.0,
      "step": 39170
    },
    {
      "entropy": 5.675922012329101,
      "epoch": 4.399730443084181,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032156280606848237,
      "loss": 4.8916,
      "mean_token_accuracy": 0.228200700879097,
      "num_tokens": 87466979.0,
      "step": 39175
    },
    {
      "entropy": 5.69632625579834,
      "epoch": 4.400292019992138,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003215235257314781,
      "loss": 4.9894,
      "mean_token_accuracy": 0.2221953734755516,
      "num_tokens": 87478193.0,
      "step": 39180
    },
    {
      "entropy": 5.664364671707153,
      "epoch": 4.400853596900095,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032148424391314665,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2261005699634552,
      "num_tokens": 87488961.0,
      "step": 39185
    },
    {
      "entropy": 5.755137825012207,
      "epoch": 4.401415173808053,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032144496061473895,
      "loss": 5.0099,
      "mean_token_accuracy": 0.2276647701859474,
      "num_tokens": 87500508.0,
      "step": 39190
    },
    {
      "entropy": 5.683781003952026,
      "epoch": 4.401976750716011,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003214056758375056,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22847432345151902,
      "num_tokens": 87510755.0,
      "step": 39195
    },
    {
      "entropy": 5.665392255783081,
      "epoch": 4.402538327623968,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003213663895826976,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22987408190965652,
      "num_tokens": 87522425.0,
      "step": 39200
    },
    {
      "entropy": 5.731524896621704,
      "epoch": 4.403099904531926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003213271018515657,
      "loss": 4.9755,
      "mean_token_accuracy": 0.23359668701887132,
      "num_tokens": 87534209.0,
      "step": 39205
    },
    {
      "entropy": 5.635415124893188,
      "epoch": 4.403661481439883,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000321287812645361,
      "loss": 4.855,
      "mean_token_accuracy": 0.23122673332691193,
      "num_tokens": 87544647.0,
      "step": 39210
    },
    {
      "entropy": 5.637995910644531,
      "epoch": 4.40422305834784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00032124852196533434,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23823147863149643,
      "num_tokens": 87555924.0,
      "step": 39215
    },
    {
      "entropy": 5.629217720031738,
      "epoch": 4.404784635255798,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003212092298127368,
      "loss": 4.9066,
      "mean_token_accuracy": 0.23727430403232574,
      "num_tokens": 87566998.0,
      "step": 39220
    },
    {
      "entropy": 5.632915496826172,
      "epoch": 4.405346212163756,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032116993618881945,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2294981837272644,
      "num_tokens": 87579516.0,
      "step": 39225
    },
    {
      "entropy": 5.761600065231323,
      "epoch": 4.405907789071714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032113064109483345,
      "loss": 4.9783,
      "mean_token_accuracy": 0.22926330268383027,
      "num_tokens": 87591512.0,
      "step": 39230
    },
    {
      "entropy": 5.723795080184937,
      "epoch": 4.406469365979671,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032109134453202993,
      "loss": 4.9193,
      "mean_token_accuracy": 0.22371701747179032,
      "num_tokens": 87602783.0,
      "step": 39235
    },
    {
      "entropy": 5.652433252334594,
      "epoch": 4.407030942887628,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032105204650166016,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22768781930208207,
      "num_tokens": 87613585.0,
      "step": 39240
    },
    {
      "entropy": 5.732391023635865,
      "epoch": 4.407592519795586,
      "grad_norm": 1.25,
      "learning_rate": 0.00032101274700497537,
      "loss": 4.9912,
      "mean_token_accuracy": 0.2280968487262726,
      "num_tokens": 87624003.0,
      "step": 39245
    },
    {
      "entropy": 5.6719671249389645,
      "epoch": 4.4081540967035435,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003209734460432269,
      "loss": 4.892,
      "mean_token_accuracy": 0.23440418839454652,
      "num_tokens": 87634429.0,
      "step": 39250
    },
    {
      "entropy": 5.650047016143799,
      "epoch": 4.408715673611501,
      "grad_norm": 1.15625,
      "learning_rate": 0.000320934143617666,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2297268971800804,
      "num_tokens": 87645887.0,
      "step": 39255
    },
    {
      "entropy": 5.716287565231323,
      "epoch": 4.409277250519459,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003208948397295441,
      "loss": 4.9212,
      "mean_token_accuracy": 0.2241521567106247,
      "num_tokens": 87657439.0,
      "step": 39260
    },
    {
      "entropy": 5.642221307754516,
      "epoch": 4.409838827427416,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032085553438011285,
      "loss": 4.8875,
      "mean_token_accuracy": 0.22920449525117875,
      "num_tokens": 87668461.0,
      "step": 39265
    },
    {
      "entropy": 5.780731964111328,
      "epoch": 4.410400404335373,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003208162275706235,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22578854411840438,
      "num_tokens": 87680595.0,
      "step": 39270
    },
    {
      "entropy": 5.659033012390137,
      "epoch": 4.410961981243331,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003207769193023277,
      "loss": 4.901,
      "mean_token_accuracy": 0.229743854701519,
      "num_tokens": 87691563.0,
      "step": 39275
    },
    {
      "entropy": 5.732366132736206,
      "epoch": 4.411523558151289,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032073760957647705,
      "loss": 4.9761,
      "mean_token_accuracy": 0.225150203704834,
      "num_tokens": 87702052.0,
      "step": 39280
    },
    {
      "entropy": 5.6593011856079105,
      "epoch": 4.412085135059247,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003206982983943232,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22462423294782638,
      "num_tokens": 87713123.0,
      "step": 39285
    },
    {
      "entropy": 5.721936941146851,
      "epoch": 4.412646711967204,
      "grad_norm": 1.125,
      "learning_rate": 0.0003206589857571178,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21912646740674974,
      "num_tokens": 87723294.0,
      "step": 39290
    },
    {
      "entropy": 5.699203300476074,
      "epoch": 4.413208288875161,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032061967166611255,
      "loss": 4.9193,
      "mean_token_accuracy": 0.23380226641893387,
      "num_tokens": 87733751.0,
      "step": 39295
    },
    {
      "entropy": 5.681936550140381,
      "epoch": 4.413769865783119,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00032058035612255934,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2375865876674652,
      "num_tokens": 87744584.0,
      "step": 39300
    },
    {
      "entropy": 5.708238315582276,
      "epoch": 4.4143314426910765,
      "grad_norm": 1.125,
      "learning_rate": 0.0003205410391277098,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2320846676826477,
      "num_tokens": 87755420.0,
      "step": 39305
    },
    {
      "entropy": 5.738642835617066,
      "epoch": 4.414893019599034,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032050172068281594,
      "loss": 4.9487,
      "mean_token_accuracy": 0.23002977520227433,
      "num_tokens": 87767213.0,
      "step": 39310
    },
    {
      "entropy": 5.701729011535645,
      "epoch": 4.415454596506992,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003204624007891297,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22884394973516464,
      "num_tokens": 87778058.0,
      "step": 39315
    },
    {
      "entropy": 5.678818082809448,
      "epoch": 4.416016173414949,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000320423079447903,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22440277636051179,
      "num_tokens": 87788279.0,
      "step": 39320
    },
    {
      "entropy": 5.658033609390259,
      "epoch": 4.416577750322907,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032038375666038776,
      "loss": 4.902,
      "mean_token_accuracy": 0.22684034556150437,
      "num_tokens": 87799643.0,
      "step": 39325
    },
    {
      "entropy": 5.626031970977783,
      "epoch": 4.417139327230864,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003203444324278361,
      "loss": 4.879,
      "mean_token_accuracy": 0.23540543615818024,
      "num_tokens": 87810493.0,
      "step": 39330
    },
    {
      "entropy": 5.6690271377563475,
      "epoch": 4.4177009041388215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003203051067515002,
      "loss": 4.9487,
      "mean_token_accuracy": 0.22684314399957656,
      "num_tokens": 87821876.0,
      "step": 39335
    },
    {
      "entropy": 5.674629020690918,
      "epoch": 4.41826248104678,
      "grad_norm": 1.25,
      "learning_rate": 0.0003202657796326321,
      "loss": 4.8788,
      "mean_token_accuracy": 0.24067425429821016,
      "num_tokens": 87832058.0,
      "step": 39340
    },
    {
      "entropy": 5.701448535919189,
      "epoch": 4.418824057954737,
      "grad_norm": 1.15625,
      "learning_rate": 0.000320226451072484,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22518128603696824,
      "num_tokens": 87842549.0,
      "step": 39345
    },
    {
      "entropy": 5.628860902786255,
      "epoch": 4.419385634862694,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032018712107230815,
      "loss": 4.9193,
      "mean_token_accuracy": 0.23172427862882614,
      "num_tokens": 87853507.0,
      "step": 39350
    },
    {
      "entropy": 5.709858655929565,
      "epoch": 4.419947211770652,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003201477896333568,
      "loss": 4.9601,
      "mean_token_accuracy": 0.22858740240335465,
      "num_tokens": 87864546.0,
      "step": 39355
    },
    {
      "entropy": 5.725544834136963,
      "epoch": 4.420508788678609,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003201084567568824,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23715697675943376,
      "num_tokens": 87876553.0,
      "step": 39360
    },
    {
      "entropy": 5.681301212310791,
      "epoch": 4.4210703655865675,
      "grad_norm": 1.265625,
      "learning_rate": 0.00032006912244413727,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22798556089401245,
      "num_tokens": 87886257.0,
      "step": 39365
    },
    {
      "entropy": 5.663320398330688,
      "epoch": 4.421631942494525,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003200297866963737,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23691389262676238,
      "num_tokens": 87897704.0,
      "step": 39370
    },
    {
      "entropy": 5.643827867507935,
      "epoch": 4.422193519402482,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003199904495148443,
      "loss": 4.931,
      "mean_token_accuracy": 0.23427506983280183,
      "num_tokens": 87908892.0,
      "step": 39375
    },
    {
      "entropy": 5.664021110534668,
      "epoch": 4.42275509631044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003199511109008016,
      "loss": 4.9875,
      "mean_token_accuracy": 0.23856417685747147,
      "num_tokens": 87921502.0,
      "step": 39380
    },
    {
      "entropy": 5.73260498046875,
      "epoch": 4.423316673218397,
      "grad_norm": 1.25,
      "learning_rate": 0.00031991177085549807,
      "loss": 4.9557,
      "mean_token_accuracy": 0.23391447812318802,
      "num_tokens": 87933250.0,
      "step": 39385
    },
    {
      "entropy": 5.660473346710205,
      "epoch": 4.4238782501263545,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031987242938018626,
      "loss": 4.8549,
      "mean_token_accuracy": 0.24110978692770005,
      "num_tokens": 87945715.0,
      "step": 39390
    },
    {
      "entropy": 5.604368400573731,
      "epoch": 4.424439827034313,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031983308647611896,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23402344584465026,
      "num_tokens": 87956973.0,
      "step": 39395
    },
    {
      "entropy": 5.7431882381439205,
      "epoch": 4.42500140394227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003197937421445488,
      "loss": 5.0035,
      "mean_token_accuracy": 0.22725330293178558,
      "num_tokens": 87968341.0,
      "step": 39400
    },
    {
      "entropy": 5.629709053039551,
      "epoch": 4.425562980850227,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003197543963867285,
      "loss": 4.9499,
      "mean_token_accuracy": 0.23089676797389985,
      "num_tokens": 87978980.0,
      "step": 39405
    },
    {
      "entropy": 5.68694109916687,
      "epoch": 4.426124557758185,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003197150492039108,
      "loss": 4.8946,
      "mean_token_accuracy": 0.237947416305542,
      "num_tokens": 87990612.0,
      "step": 39410
    },
    {
      "entropy": 5.69690351486206,
      "epoch": 4.426686134666142,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003196757005973487,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22908110916614532,
      "num_tokens": 88001052.0,
      "step": 39415
    },
    {
      "entropy": 5.6836048603057865,
      "epoch": 4.4272477115741005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031963635056829487,
      "loss": 4.9481,
      "mean_token_accuracy": 0.2290861949324608,
      "num_tokens": 88013102.0,
      "step": 39420
    },
    {
      "entropy": 5.736581182479858,
      "epoch": 4.427809288482058,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031959699911800237,
      "loss": 5.0142,
      "mean_token_accuracy": 0.23125474900007248,
      "num_tokens": 88023952.0,
      "step": 39425
    },
    {
      "entropy": 5.680573177337647,
      "epoch": 4.428370865390015,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003195576462477241,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22795908898115158,
      "num_tokens": 88035148.0,
      "step": 39430
    },
    {
      "entropy": 5.651772928237915,
      "epoch": 4.428932442297973,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031951829195871315,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22257012724876404,
      "num_tokens": 88046294.0,
      "step": 39435
    },
    {
      "entropy": 5.660462665557861,
      "epoch": 4.42949401920593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031947893625222244,
      "loss": 4.9251,
      "mean_token_accuracy": 0.22932168692350388,
      "num_tokens": 88058225.0,
      "step": 39440
    },
    {
      "entropy": 5.752375364303589,
      "epoch": 4.430055596113887,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031943957912950523,
      "loss": 4.9645,
      "mean_token_accuracy": 0.2227581188082695,
      "num_tokens": 88069994.0,
      "step": 39445
    },
    {
      "entropy": 5.741655445098877,
      "epoch": 4.4306171730218455,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003194002205918145,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2285333126783371,
      "num_tokens": 88081912.0,
      "step": 39450
    },
    {
      "entropy": 5.681422090530395,
      "epoch": 4.431178749929803,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003193608606404036,
      "loss": 4.9214,
      "mean_token_accuracy": 0.23155568540096283,
      "num_tokens": 88092457.0,
      "step": 39455
    },
    {
      "entropy": 5.7679572105407715,
      "epoch": 4.43174032683776,
      "grad_norm": 1.25,
      "learning_rate": 0.0003193214992765257,
      "loss": 5.0896,
      "mean_token_accuracy": 0.2145327091217041,
      "num_tokens": 88104407.0,
      "step": 39460
    },
    {
      "entropy": 5.661266231536866,
      "epoch": 4.432301903745718,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031928213650143407,
      "loss": 4.8827,
      "mean_token_accuracy": 0.2346656247973442,
      "num_tokens": 88114653.0,
      "step": 39465
    },
    {
      "entropy": 5.7578307628631595,
      "epoch": 4.432863480653675,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000319242772316382,
      "loss": 4.93,
      "mean_token_accuracy": 0.23817052841186523,
      "num_tokens": 88125477.0,
      "step": 39470
    },
    {
      "entropy": 5.624803733825684,
      "epoch": 4.433425057561633,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003192034067226228,
      "loss": 4.8458,
      "mean_token_accuracy": 0.24476187974214553,
      "num_tokens": 88136419.0,
      "step": 39475
    },
    {
      "entropy": 5.632326078414917,
      "epoch": 4.433986634469591,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00031916403972141017,
      "loss": 4.8632,
      "mean_token_accuracy": 0.23451082706451415,
      "num_tokens": 88146578.0,
      "step": 39480
    },
    {
      "entropy": 5.521237897872925,
      "epoch": 4.434548211377548,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003191246713139973,
      "loss": 4.7829,
      "mean_token_accuracy": 0.23884371817111968,
      "num_tokens": 88157351.0,
      "step": 39485
    },
    {
      "entropy": 5.72218952178955,
      "epoch": 4.435109788285506,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031908530150163777,
      "loss": 4.9681,
      "mean_token_accuracy": 0.2249475747346878,
      "num_tokens": 88168460.0,
      "step": 39490
    },
    {
      "entropy": 5.759441757202149,
      "epoch": 4.435671365193463,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031904593028558517,
      "loss": 4.9586,
      "mean_token_accuracy": 0.23126134425401687,
      "num_tokens": 88178994.0,
      "step": 39495
    },
    {
      "entropy": 5.680372476577759,
      "epoch": 4.43623294210142,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031900655766709305,
      "loss": 4.8706,
      "mean_token_accuracy": 0.2306483969092369,
      "num_tokens": 88188916.0,
      "step": 39500
    },
    {
      "entropy": 5.62461404800415,
      "epoch": 4.4367945190093785,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003189671836474151,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22891703844070435,
      "num_tokens": 88200833.0,
      "step": 39505
    },
    {
      "entropy": 5.709565591812134,
      "epoch": 4.437356095917336,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003189278082278049,
      "loss": 5.0311,
      "mean_token_accuracy": 0.22117566913366318,
      "num_tokens": 88212022.0,
      "step": 39510
    },
    {
      "entropy": 5.673894929885864,
      "epoch": 4.437917672825294,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031888843140951627,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23525281846523285,
      "num_tokens": 88222327.0,
      "step": 39515
    },
    {
      "entropy": 5.589035129547119,
      "epoch": 4.438479249733251,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003188490531938031,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23626897633075714,
      "num_tokens": 88233141.0,
      "step": 39520
    },
    {
      "entropy": 5.761027240753174,
      "epoch": 4.439040826641208,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031880967358191884,
      "loss": 5.017,
      "mean_token_accuracy": 0.21916008293628692,
      "num_tokens": 88245663.0,
      "step": 39525
    },
    {
      "entropy": 5.711053562164307,
      "epoch": 4.439602403549166,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031877029257511774,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2293349325656891,
      "num_tokens": 88256291.0,
      "step": 39530
    },
    {
      "entropy": 5.742602920532226,
      "epoch": 4.440163980457124,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031873091017465356,
      "loss": 5.0038,
      "mean_token_accuracy": 0.2223600447177887,
      "num_tokens": 88268382.0,
      "step": 39535
    },
    {
      "entropy": 5.623960161209107,
      "epoch": 4.440725557365081,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003186915263817802,
      "loss": 4.8833,
      "mean_token_accuracy": 0.23980448246002198,
      "num_tokens": 88279638.0,
      "step": 39540
    },
    {
      "entropy": 5.68126916885376,
      "epoch": 4.441287134273039,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031865214119775173,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22941793650388717,
      "num_tokens": 88290599.0,
      "step": 39545
    },
    {
      "entropy": 5.686835289001465,
      "epoch": 4.441848711180996,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000318612754623822,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23361109793186188,
      "num_tokens": 88301572.0,
      "step": 39550
    },
    {
      "entropy": 5.726101446151733,
      "epoch": 4.442410288088954,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031857336666124537,
      "loss": 4.9975,
      "mean_token_accuracy": 0.23030732274055482,
      "num_tokens": 88312978.0,
      "step": 39555
    },
    {
      "entropy": 5.739376974105835,
      "epoch": 4.442971864996911,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003185339773112758,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2248788967728615,
      "num_tokens": 88324023.0,
      "step": 39560
    },
    {
      "entropy": 5.668859577178955,
      "epoch": 4.443533441904869,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003184945865751675,
      "loss": 4.8819,
      "mean_token_accuracy": 0.2247696503996849,
      "num_tokens": 88335366.0,
      "step": 39565
    },
    {
      "entropy": 5.676552486419678,
      "epoch": 4.444095018812827,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031845519445417475,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22518337666988372,
      "num_tokens": 88347786.0,
      "step": 39570
    },
    {
      "entropy": 5.74381275177002,
      "epoch": 4.444656595720784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003184158009495518,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22735944241285325,
      "num_tokens": 88358720.0,
      "step": 39575
    },
    {
      "entropy": 5.730683612823486,
      "epoch": 4.445218172628741,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003183764060625528,
      "loss": 4.9824,
      "mean_token_accuracy": 0.22478198409080505,
      "num_tokens": 88370152.0,
      "step": 39580
    },
    {
      "entropy": 5.6692534446716305,
      "epoch": 4.445779749536699,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031833700979443227,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22657172977924347,
      "num_tokens": 88381385.0,
      "step": 39585
    },
    {
      "entropy": 5.693360710144043,
      "epoch": 4.4463413264446565,
      "grad_norm": 1.328125,
      "learning_rate": 0.00031829761214644453,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23364294618368148,
      "num_tokens": 88391450.0,
      "step": 39590
    },
    {
      "entropy": 5.590979528427124,
      "epoch": 4.446902903352614,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000318258213119844,
      "loss": 4.846,
      "mean_token_accuracy": 0.2443210244178772,
      "num_tokens": 88402310.0,
      "step": 39595
    },
    {
      "entropy": 5.6438017845153805,
      "epoch": 4.447464480260572,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003182188127158852,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22070925682783127,
      "num_tokens": 88412286.0,
      "step": 39600
    },
    {
      "entropy": 5.734720706939697,
      "epoch": 4.448026057168529,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003181794109358227,
      "loss": 4.974,
      "mean_token_accuracy": 0.23224829584360124,
      "num_tokens": 88423335.0,
      "step": 39605
    },
    {
      "entropy": 5.624710178375244,
      "epoch": 4.448587634076487,
      "grad_norm": 1.125,
      "learning_rate": 0.0003181400077809109,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23271337151527405,
      "num_tokens": 88434328.0,
      "step": 39610
    },
    {
      "entropy": 5.687139511108398,
      "epoch": 4.449149210984444,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031810060325240466,
      "loss": 4.9281,
      "mean_token_accuracy": 0.230561363697052,
      "num_tokens": 88445611.0,
      "step": 39615
    },
    {
      "entropy": 5.713623428344727,
      "epoch": 4.449710787892402,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003180611973515583,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2303130254149437,
      "num_tokens": 88456207.0,
      "step": 39620
    },
    {
      "entropy": 5.7746437072753904,
      "epoch": 4.45027236480036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031802179007962687,
      "loss": 5.0749,
      "mean_token_accuracy": 0.21835861951112748,
      "num_tokens": 88467322.0,
      "step": 39625
    },
    {
      "entropy": 5.654308557510376,
      "epoch": 4.450833941708317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003179823814378649,
      "loss": 4.9824,
      "mean_token_accuracy": 0.23045240789651872,
      "num_tokens": 88479703.0,
      "step": 39630
    },
    {
      "entropy": 5.63208589553833,
      "epoch": 4.451395518616274,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031794297142752727,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2360176131129265,
      "num_tokens": 88490284.0,
      "step": 39635
    },
    {
      "entropy": 5.669963121414185,
      "epoch": 4.451957095524232,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003179035600498687,
      "loss": 4.9534,
      "mean_token_accuracy": 0.2305993840098381,
      "num_tokens": 88502232.0,
      "step": 39640
    },
    {
      "entropy": 5.618154430389405,
      "epoch": 4.4525186724321895,
      "grad_norm": 1.125,
      "learning_rate": 0.00031786414730614423,
      "loss": 4.8891,
      "mean_token_accuracy": 0.23285411447286605,
      "num_tokens": 88513788.0,
      "step": 39645
    },
    {
      "entropy": 5.692682933807373,
      "epoch": 4.453080249340147,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003178247331976086,
      "loss": 4.9533,
      "mean_token_accuracy": 0.22896423637866975,
      "num_tokens": 88524942.0,
      "step": 39650
    },
    {
      "entropy": 5.606480979919434,
      "epoch": 4.453641826248105,
      "grad_norm": 1.25,
      "learning_rate": 0.00031778531772551696,
      "loss": 4.8026,
      "mean_token_accuracy": 0.23553079813718797,
      "num_tokens": 88535692.0,
      "step": 39655
    },
    {
      "entropy": 5.659510660171509,
      "epoch": 4.454203403156062,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003177459008911241,
      "loss": 4.902,
      "mean_token_accuracy": 0.23064474612474442,
      "num_tokens": 88547007.0,
      "step": 39660
    },
    {
      "entropy": 5.746747350692749,
      "epoch": 4.45476498006402,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031770648269568524,
      "loss": 4.9997,
      "mean_token_accuracy": 0.2316354975104332,
      "num_tokens": 88558197.0,
      "step": 39665
    },
    {
      "entropy": 5.751490831375122,
      "epoch": 4.455326556971977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003176670631404554,
      "loss": 5.009,
      "mean_token_accuracy": 0.22427302449941636,
      "num_tokens": 88569488.0,
      "step": 39670
    },
    {
      "entropy": 5.699948406219482,
      "epoch": 4.4558881338799345,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031762764222668976,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2236274853348732,
      "num_tokens": 88582054.0,
      "step": 39675
    },
    {
      "entropy": 5.683641576766968,
      "epoch": 4.456449710787893,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031758821995564343,
      "loss": 4.8597,
      "mean_token_accuracy": 0.2311691164970398,
      "num_tokens": 88592528.0,
      "step": 39680
    },
    {
      "entropy": 5.719804048538208,
      "epoch": 4.45701128769585,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003175487963285716,
      "loss": 4.9368,
      "mean_token_accuracy": 0.23474412113428117,
      "num_tokens": 88604136.0,
      "step": 39685
    },
    {
      "entropy": 5.762092161178589,
      "epoch": 4.457572864603807,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003175093713467297,
      "loss": 4.9952,
      "mean_token_accuracy": 0.22763127386569976,
      "num_tokens": 88615629.0,
      "step": 39690
    },
    {
      "entropy": 5.614289617538452,
      "epoch": 4.458134441511765,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031746994501137283,
      "loss": 4.8441,
      "mean_token_accuracy": 0.23704473972320556,
      "num_tokens": 88626100.0,
      "step": 39695
    },
    {
      "entropy": 5.645095252990723,
      "epoch": 4.458696018419722,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003174305173237565,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2255966395139694,
      "num_tokens": 88638948.0,
      "step": 39700
    },
    {
      "entropy": 5.711699056625366,
      "epoch": 4.4592575953276805,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000317391088285136,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22904917299747468,
      "num_tokens": 88650488.0,
      "step": 39705
    },
    {
      "entropy": 5.804946947097778,
      "epoch": 4.459819172235638,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003173516578967668,
      "loss": 5.0227,
      "mean_token_accuracy": 0.2216411828994751,
      "num_tokens": 88661584.0,
      "step": 39710
    },
    {
      "entropy": 5.678458118438721,
      "epoch": 4.460380749143595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031731222615990443,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2269735500216484,
      "num_tokens": 88673051.0,
      "step": 39715
    },
    {
      "entropy": 5.701056385040284,
      "epoch": 4.460942326051553,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031727279307580436,
      "loss": 5.0298,
      "mean_token_accuracy": 0.2237112045288086,
      "num_tokens": 88684122.0,
      "step": 39720
    },
    {
      "entropy": 5.746242904663086,
      "epoch": 4.46150390295951,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031723335864572216,
      "loss": 5.0377,
      "mean_token_accuracy": 0.22344363182783128,
      "num_tokens": 88694620.0,
      "step": 39725
    },
    {
      "entropy": 5.794932699203491,
      "epoch": 4.4620654798674675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031719392287091343,
      "loss": 5.02,
      "mean_token_accuracy": 0.22518071681261062,
      "num_tokens": 88706137.0,
      "step": 39730
    },
    {
      "entropy": 5.658501720428466,
      "epoch": 4.462627056775426,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003171544857526338,
      "loss": 4.883,
      "mean_token_accuracy": 0.22849163562059402,
      "num_tokens": 88717326.0,
      "step": 39735
    },
    {
      "entropy": 5.648787164688111,
      "epoch": 4.463188633683383,
      "grad_norm": 1.25,
      "learning_rate": 0.0003171150472921391,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23146621882915497,
      "num_tokens": 88727125.0,
      "step": 39740
    },
    {
      "entropy": 5.647704696655273,
      "epoch": 4.463750210591341,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031707560749068487,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22921959906816483,
      "num_tokens": 88737098.0,
      "step": 39745
    },
    {
      "entropy": 5.710072088241577,
      "epoch": 4.464311787499298,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000317036166349527,
      "loss": 4.9365,
      "mean_token_accuracy": 0.23454747647047042,
      "num_tokens": 88748454.0,
      "step": 39750
    },
    {
      "entropy": 5.690549421310425,
      "epoch": 4.464873364407255,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003169967238699213,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22848209738731384,
      "num_tokens": 88760165.0,
      "step": 39755
    },
    {
      "entropy": 5.6671287536621096,
      "epoch": 4.4654349413152135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031695728005312363,
      "loss": 4.96,
      "mean_token_accuracy": 0.22670480161905288,
      "num_tokens": 88772124.0,
      "step": 39760
    },
    {
      "entropy": 5.632981586456299,
      "epoch": 4.465996518223171,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003169178349003898,
      "loss": 4.8644,
      "mean_token_accuracy": 0.23353685140609742,
      "num_tokens": 88782429.0,
      "step": 39765
    },
    {
      "entropy": 5.678928184509277,
      "epoch": 4.466558095131128,
      "grad_norm": 1.1875,
      "learning_rate": 0.000316878388412976,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22825498431921004,
      "num_tokens": 88794060.0,
      "step": 39770
    },
    {
      "entropy": 5.750347423553467,
      "epoch": 4.467119672039086,
      "grad_norm": 1.0625,
      "learning_rate": 0.000316838940592138,
      "loss": 4.985,
      "mean_token_accuracy": 0.2336868315935135,
      "num_tokens": 88806174.0,
      "step": 39775
    },
    {
      "entropy": 5.620849180221557,
      "epoch": 4.467681248947043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031679949143913183,
      "loss": 4.8321,
      "mean_token_accuracy": 0.23275772333145142,
      "num_tokens": 88817587.0,
      "step": 39780
    },
    {
      "entropy": 5.6946789741516115,
      "epoch": 4.468242825855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003167600409552137,
      "loss": 4.9119,
      "mean_token_accuracy": 0.23307379335165024,
      "num_tokens": 88829435.0,
      "step": 39785
    },
    {
      "entropy": 5.70127124786377,
      "epoch": 4.4688044027629585,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003167205891416397,
      "loss": 5.0121,
      "mean_token_accuracy": 0.22397609204053878,
      "num_tokens": 88840244.0,
      "step": 39790
    },
    {
      "entropy": 5.797454929351806,
      "epoch": 4.469365979670916,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000316681135999666,
      "loss": 5.0164,
      "mean_token_accuracy": 0.23462532609701156,
      "num_tokens": 88852085.0,
      "step": 39795
    },
    {
      "entropy": 5.734954786300659,
      "epoch": 4.469927556578874,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003166416815305486,
      "loss": 5.0196,
      "mean_token_accuracy": 0.22927509099245072,
      "num_tokens": 88864605.0,
      "step": 39800
    },
    {
      "entropy": 5.739845895767212,
      "epoch": 4.470489133486831,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031660222573554406,
      "loss": 4.9222,
      "mean_token_accuracy": 0.2328273296356201,
      "num_tokens": 88876209.0,
      "step": 39805
    },
    {
      "entropy": 5.640303087234497,
      "epoch": 4.471050710394788,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003165627686159085,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2313024714589119,
      "num_tokens": 88886346.0,
      "step": 39810
    },
    {
      "entropy": 5.624478673934936,
      "epoch": 4.471612287302746,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003165233101728982,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22697805613279343,
      "num_tokens": 88898190.0,
      "step": 39815
    },
    {
      "entropy": 5.689992141723633,
      "epoch": 4.472173864210704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031648385040776965,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21492781341075898,
      "num_tokens": 88909957.0,
      "step": 39820
    },
    {
      "entropy": 5.8488060474395756,
      "epoch": 4.472735441118661,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003164443893217793,
      "loss": 5.1013,
      "mean_token_accuracy": 0.2234713613986969,
      "num_tokens": 88922993.0,
      "step": 39825
    },
    {
      "entropy": 5.609824752807617,
      "epoch": 4.473297018026619,
      "grad_norm": 1.125,
      "learning_rate": 0.0003164049269161834,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23830590695142745,
      "num_tokens": 88934506.0,
      "step": 39830
    },
    {
      "entropy": 5.654167127609253,
      "epoch": 4.473858594934576,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031636546319223845,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2300452098250389,
      "num_tokens": 88946963.0,
      "step": 39835
    },
    {
      "entropy": 5.7170671939849855,
      "epoch": 4.474420171842533,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003163259981512013,
      "loss": 4.9982,
      "mean_token_accuracy": 0.23114436417818068,
      "num_tokens": 88956995.0,
      "step": 39840
    },
    {
      "entropy": 5.77328634262085,
      "epoch": 4.4749817487504915,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003162865317943283,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2291094556450844,
      "num_tokens": 88967507.0,
      "step": 39845
    },
    {
      "entropy": 5.779266738891602,
      "epoch": 4.475543325658449,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031624706412287614,
      "loss": 5.0009,
      "mean_token_accuracy": 0.2308438763022423,
      "num_tokens": 88979943.0,
      "step": 39850
    },
    {
      "entropy": 5.71986951828003,
      "epoch": 4.476104902566407,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003162075951381014,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2244747757911682,
      "num_tokens": 88990840.0,
      "step": 39855
    },
    {
      "entropy": 5.722035646438599,
      "epoch": 4.476666479474364,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031616812484126084,
      "loss": 5.0042,
      "mean_token_accuracy": 0.22155797034502028,
      "num_tokens": 89002245.0,
      "step": 39860
    },
    {
      "entropy": 5.793744134902954,
      "epoch": 4.477228056382321,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003161286532336113,
      "loss": 5.0002,
      "mean_token_accuracy": 0.23101490885019302,
      "num_tokens": 89013520.0,
      "step": 39865
    },
    {
      "entropy": 5.761450386047363,
      "epoch": 4.477789633290279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031608918031640946,
      "loss": 4.9242,
      "mean_token_accuracy": 0.230235056579113,
      "num_tokens": 89025377.0,
      "step": 39870
    },
    {
      "entropy": 5.662507438659668,
      "epoch": 4.478351210198237,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003160497060909121,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2254226863384247,
      "num_tokens": 89035299.0,
      "step": 39875
    },
    {
      "entropy": 5.685638523101806,
      "epoch": 4.478912787106194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031601023055837624,
      "loss": 4.9825,
      "mean_token_accuracy": 0.23114589154720305,
      "num_tokens": 89046255.0,
      "step": 39880
    },
    {
      "entropy": 5.745644998550415,
      "epoch": 4.479474364014152,
      "grad_norm": 1.40625,
      "learning_rate": 0.00031597075372005875,
      "loss": 5.0323,
      "mean_token_accuracy": 0.22452756017446518,
      "num_tokens": 89057873.0,
      "step": 39885
    },
    {
      "entropy": 5.691717195510864,
      "epoch": 4.480035940922109,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031593127557721657,
      "loss": 4.855,
      "mean_token_accuracy": 0.23227598071098327,
      "num_tokens": 89068830.0,
      "step": 39890
    },
    {
      "entropy": 5.647245311737061,
      "epoch": 4.480597517830067,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003158917961311067,
      "loss": 4.9342,
      "mean_token_accuracy": 0.23469128459692,
      "num_tokens": 89080328.0,
      "step": 39895
    },
    {
      "entropy": 5.702776765823364,
      "epoch": 4.481159094738024,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003158523153829862,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22767811119556428,
      "num_tokens": 89090118.0,
      "step": 39900
    },
    {
      "entropy": 5.7423289775848385,
      "epoch": 4.481720671645982,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003158128333341121,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2300456389784813,
      "num_tokens": 89103338.0,
      "step": 39905
    },
    {
      "entropy": 5.727916145324707,
      "epoch": 4.48228224855394,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031577334998574155,
      "loss": 4.904,
      "mean_token_accuracy": 0.23579777777194977,
      "num_tokens": 89113721.0,
      "step": 39910
    },
    {
      "entropy": 5.793037223815918,
      "epoch": 4.482843825461897,
      "grad_norm": 1.25,
      "learning_rate": 0.0003157338653391317,
      "loss": 5.0103,
      "mean_token_accuracy": 0.22751339375972748,
      "num_tokens": 89125425.0,
      "step": 39915
    },
    {
      "entropy": 5.689334106445313,
      "epoch": 4.483405402369854,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003156943793955398,
      "loss": 4.9848,
      "mean_token_accuracy": 0.2281741976737976,
      "num_tokens": 89135407.0,
      "step": 39920
    },
    {
      "entropy": 5.718850946426391,
      "epoch": 4.483966979277812,
      "grad_norm": 1.34375,
      "learning_rate": 0.00031565489215622304,
      "loss": 4.9882,
      "mean_token_accuracy": 0.22552642226219177,
      "num_tokens": 89147563.0,
      "step": 39925
    },
    {
      "entropy": 5.753605222702026,
      "epoch": 4.4845285561857695,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003156154036224388,
      "loss": 4.9803,
      "mean_token_accuracy": 0.22992780506610871,
      "num_tokens": 89158336.0,
      "step": 39930
    },
    {
      "entropy": 5.637319564819336,
      "epoch": 4.485090133093728,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003155759137954444,
      "loss": 4.8712,
      "mean_token_accuracy": 0.23666571974754333,
      "num_tokens": 89168894.0,
      "step": 39935
    },
    {
      "entropy": 5.768155193328857,
      "epoch": 4.485651710001685,
      "grad_norm": 1.3125,
      "learning_rate": 0.00031553642267649715,
      "loss": 4.9792,
      "mean_token_accuracy": 0.2299123927950859,
      "num_tokens": 89180701.0,
      "step": 39940
    },
    {
      "entropy": 5.7198573589324955,
      "epoch": 4.486213286909642,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003154969302668544,
      "loss": 4.9295,
      "mean_token_accuracy": 0.23663345128297805,
      "num_tokens": 89191192.0,
      "step": 39945
    },
    {
      "entropy": 5.68983793258667,
      "epoch": 4.4867748638176,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003154574365677737,
      "loss": 4.8865,
      "mean_token_accuracy": 0.23132859617471696,
      "num_tokens": 89202377.0,
      "step": 39950
    },
    {
      "entropy": 5.601609754562378,
      "epoch": 4.487336440725557,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003154179415805126,
      "loss": 4.8441,
      "mean_token_accuracy": 0.24028964936733246,
      "num_tokens": 89212869.0,
      "step": 39955
    },
    {
      "entropy": 5.673832416534424,
      "epoch": 4.487898017633515,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003153784453063285,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2273809254169464,
      "num_tokens": 89224558.0,
      "step": 39960
    },
    {
      "entropy": 5.62257490158081,
      "epoch": 4.488459594541473,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003153389477464791,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23530466258525848,
      "num_tokens": 89235645.0,
      "step": 39965
    },
    {
      "entropy": 5.647004079818726,
      "epoch": 4.48902117144943,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003152994489022219,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22912724912166596,
      "num_tokens": 89247063.0,
      "step": 39970
    },
    {
      "entropy": 5.681287622451782,
      "epoch": 4.489582748357387,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031525994877481465,
      "loss": 4.9836,
      "mean_token_accuracy": 0.2283946231007576,
      "num_tokens": 89258297.0,
      "step": 39975
    },
    {
      "entropy": 5.662408065795899,
      "epoch": 4.490144325265345,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000315220447365515,
      "loss": 4.9279,
      "mean_token_accuracy": 0.23177847415208816,
      "num_tokens": 89269245.0,
      "step": 39980
    },
    {
      "entropy": 5.734433937072754,
      "epoch": 4.4907059021733025,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031518094467558074,
      "loss": 4.999,
      "mean_token_accuracy": 0.22556760758161545,
      "num_tokens": 89281959.0,
      "step": 39985
    },
    {
      "entropy": 5.698241138458252,
      "epoch": 4.491267479081261,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031514144070626956,
      "loss": 4.8764,
      "mean_token_accuracy": 0.24105250537395478,
      "num_tokens": 89292656.0,
      "step": 39990
    },
    {
      "entropy": 5.71643557548523,
      "epoch": 4.491829055989218,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031510193545883944,
      "loss": 4.99,
      "mean_token_accuracy": 0.22612998485565186,
      "num_tokens": 89304198.0,
      "step": 39995
    },
    {
      "entropy": 5.6899556636810305,
      "epoch": 4.492390632897175,
      "grad_norm": 1.28125,
      "learning_rate": 0.000315062428934548,
      "loss": 4.959,
      "mean_token_accuracy": 0.2370431140065193,
      "num_tokens": 89315443.0,
      "step": 40000
    },
    {
      "entropy": 5.659947681427002,
      "epoch": 4.492952209805133,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031502292113465334,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22994639873504638,
      "num_tokens": 89326597.0,
      "step": 40005
    },
    {
      "entropy": 5.722170114517212,
      "epoch": 4.49351378671309,
      "grad_norm": 1.359375,
      "learning_rate": 0.00031498341206041337,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22885742038488388,
      "num_tokens": 89338385.0,
      "step": 40010
    },
    {
      "entropy": 5.668883514404297,
      "epoch": 4.4940753636210475,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000314943901713086,
      "loss": 4.8842,
      "mean_token_accuracy": 0.23692445307970048,
      "num_tokens": 89350950.0,
      "step": 40015
    },
    {
      "entropy": 5.672875213623047,
      "epoch": 4.494636940529006,
      "grad_norm": 1.25,
      "learning_rate": 0.0003149043900939294,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22571140229701997,
      "num_tokens": 89361154.0,
      "step": 40020
    },
    {
      "entropy": 5.691471862792969,
      "epoch": 4.495198517436963,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031486487720420146,
      "loss": 4.9698,
      "mean_token_accuracy": 0.22920708060264589,
      "num_tokens": 89373188.0,
      "step": 40025
    },
    {
      "entropy": 5.721014547348022,
      "epoch": 4.49576009434492,
      "grad_norm": 1.28125,
      "learning_rate": 0.00031482536304516047,
      "loss": 4.9144,
      "mean_token_accuracy": 0.23555326759815215,
      "num_tokens": 89383651.0,
      "step": 40030
    },
    {
      "entropy": 5.653041076660156,
      "epoch": 4.496321671252878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003147858476180644,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2327425628900528,
      "num_tokens": 89395385.0,
      "step": 40035
    },
    {
      "entropy": 5.641970777511597,
      "epoch": 4.496883248160835,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031474633092417155,
      "loss": 4.9268,
      "mean_token_accuracy": 0.23269826769828797,
      "num_tokens": 89406071.0,
      "step": 40040
    },
    {
      "entropy": 5.71780743598938,
      "epoch": 4.4974448250687935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031470681296474006,
      "loss": 5.0508,
      "mean_token_accuracy": 0.22650967687368392,
      "num_tokens": 89417044.0,
      "step": 40045
    },
    {
      "entropy": 5.682605457305908,
      "epoch": 4.498006401976751,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003146672937410283,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2273021847009659,
      "num_tokens": 89428253.0,
      "step": 40050
    },
    {
      "entropy": 5.7022469997406,
      "epoch": 4.498567978884708,
      "grad_norm": 1.25,
      "learning_rate": 0.00031462777325429447,
      "loss": 4.9549,
      "mean_token_accuracy": 0.22446168214082718,
      "num_tokens": 89438760.0,
      "step": 40055
    },
    {
      "entropy": 5.651180553436279,
      "epoch": 4.499129555792666,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000314588251505797,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2410467103123665,
      "num_tokens": 89450335.0,
      "step": 40060
    },
    {
      "entropy": 5.7290033340454105,
      "epoch": 4.499691132700623,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00031454872849679425,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22913734912872313,
      "num_tokens": 89461077.0,
      "step": 40065
    },
    {
      "entropy": 5.751163911819458,
      "epoch": 4.500252709608581,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003145092042285446,
      "loss": 4.9685,
      "mean_token_accuracy": 0.2247541978955269,
      "num_tokens": 89472281.0,
      "step": 40070
    },
    {
      "entropy": 5.690649080276489,
      "epoch": 4.500814286516539,
      "grad_norm": 1.125,
      "learning_rate": 0.0003144696787023066,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2290804296731949,
      "num_tokens": 89483030.0,
      "step": 40075
    },
    {
      "entropy": 5.617540788650513,
      "epoch": 4.501375863424496,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031443015191933864,
      "loss": 4.83,
      "mean_token_accuracy": 0.2318356767296791,
      "num_tokens": 89493355.0,
      "step": 40080
    },
    {
      "entropy": 5.689609813690185,
      "epoch": 4.501937440332454,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031439062388089953,
      "loss": 4.9888,
      "mean_token_accuracy": 0.22844127118587493,
      "num_tokens": 89504253.0,
      "step": 40085
    },
    {
      "entropy": 5.621167612075806,
      "epoch": 4.502499017240411,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003143510945882475,
      "loss": 4.9742,
      "mean_token_accuracy": 0.22824725210666658,
      "num_tokens": 89514720.0,
      "step": 40090
    },
    {
      "entropy": 5.707033157348633,
      "epoch": 4.503060594148368,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031431156404264143,
      "loss": 4.9356,
      "mean_token_accuracy": 0.23312594890594482,
      "num_tokens": 89526583.0,
      "step": 40095
    },
    {
      "entropy": 5.683926153182983,
      "epoch": 4.5036221710563265,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031427203224533996,
      "loss": 4.9644,
      "mean_token_accuracy": 0.23062622249126435,
      "num_tokens": 89537791.0,
      "step": 40100
    },
    {
      "entropy": 5.745060014724731,
      "epoch": 4.504183747964284,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031423249919760167,
      "loss": 5.0069,
      "mean_token_accuracy": 0.2285539448261261,
      "num_tokens": 89549480.0,
      "step": 40105
    },
    {
      "entropy": 5.700106906890869,
      "epoch": 4.504745324872241,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003141929649006854,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22995727360248566,
      "num_tokens": 89559531.0,
      "step": 40110
    },
    {
      "entropy": 5.709388494491577,
      "epoch": 4.505306901780199,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031415342935585,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22287449538707732,
      "num_tokens": 89570238.0,
      "step": 40115
    },
    {
      "entropy": 5.683377456665039,
      "epoch": 4.505868478688156,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003141138925643542,
      "loss": 4.9636,
      "mean_token_accuracy": 0.229656483232975,
      "num_tokens": 89581300.0,
      "step": 40120
    },
    {
      "entropy": 5.757976293563843,
      "epoch": 4.506430055596114,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003140743545274569,
      "loss": 4.9972,
      "mean_token_accuracy": 0.2265973448753357,
      "num_tokens": 89592461.0,
      "step": 40125
    },
    {
      "entropy": 5.78132963180542,
      "epoch": 4.5069916325040715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031403481524641696,
      "loss": 5.0795,
      "mean_token_accuracy": 0.21739933341741563,
      "num_tokens": 89603380.0,
      "step": 40130
    },
    {
      "entropy": 5.713700532913208,
      "epoch": 4.507553209412029,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031399527472249336,
      "loss": 4.9567,
      "mean_token_accuracy": 0.2245760068297386,
      "num_tokens": 89614268.0,
      "step": 40135
    },
    {
      "entropy": 5.741205644607544,
      "epoch": 4.508114786319987,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031395573295694507,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22884212881326677,
      "num_tokens": 89625480.0,
      "step": 40140
    },
    {
      "entropy": 5.706895923614502,
      "epoch": 4.508676363227944,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031391618995103116,
      "loss": 4.9755,
      "mean_token_accuracy": 0.22781564742326738,
      "num_tokens": 89637437.0,
      "step": 40145
    },
    {
      "entropy": 5.687348222732544,
      "epoch": 4.509237940135901,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003138766457060107,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2364622637629509,
      "num_tokens": 89649196.0,
      "step": 40150
    },
    {
      "entropy": 5.587875843048096,
      "epoch": 4.509799517043859,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003138371002231427,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23389423489570618,
      "num_tokens": 89659536.0,
      "step": 40155
    },
    {
      "entropy": 5.7498533725738525,
      "epoch": 4.510361093951817,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003137975535036864,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22717299461364746,
      "num_tokens": 89671021.0,
      "step": 40160
    },
    {
      "entropy": 5.660739850997925,
      "epoch": 4.510922670859774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031375800554890094,
      "loss": 4.8609,
      "mean_token_accuracy": 0.23546816855669023,
      "num_tokens": 89681689.0,
      "step": 40165
    },
    {
      "entropy": 5.697380638122558,
      "epoch": 4.511484247767732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031371845636004556,
      "loss": 4.9424,
      "mean_token_accuracy": 0.22932549118995665,
      "num_tokens": 89692353.0,
      "step": 40170
    },
    {
      "entropy": 5.64848484992981,
      "epoch": 4.512045824675689,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003136789059383795,
      "loss": 4.9572,
      "mean_token_accuracy": 0.23104816824197769,
      "num_tokens": 89702852.0,
      "step": 40175
    },
    {
      "entropy": 5.706516981124878,
      "epoch": 4.512607401583647,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003136393542851622,
      "loss": 4.9972,
      "mean_token_accuracy": 0.22661093771457672,
      "num_tokens": 89716125.0,
      "step": 40180
    },
    {
      "entropy": 5.677619743347168,
      "epoch": 4.5131689784916045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031359980140165274,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22626127451658248,
      "num_tokens": 89726611.0,
      "step": 40185
    },
    {
      "entropy": 5.684386444091797,
      "epoch": 4.513730555399562,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003135602472891106,
      "loss": 5.0178,
      "mean_token_accuracy": 0.22738903760910034,
      "num_tokens": 89738563.0,
      "step": 40190
    },
    {
      "entropy": 5.6555112361907955,
      "epoch": 4.51429213230752,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031352069194879527,
      "loss": 4.8914,
      "mean_token_accuracy": 0.23046004474163057,
      "num_tokens": 89748842.0,
      "step": 40195
    },
    {
      "entropy": 5.648071813583374,
      "epoch": 4.514853709215477,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003134811353819662,
      "loss": 4.9299,
      "mean_token_accuracy": 0.23298295289278032,
      "num_tokens": 89760747.0,
      "step": 40200
    },
    {
      "entropy": 5.672765684127808,
      "epoch": 4.515415286123434,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003134415775898828,
      "loss": 4.8298,
      "mean_token_accuracy": 0.22897313833236693,
      "num_tokens": 89773059.0,
      "step": 40205
    },
    {
      "entropy": 5.698870897293091,
      "epoch": 4.515976863031392,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003134020185738047,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2365595132112503,
      "num_tokens": 89783347.0,
      "step": 40210
    },
    {
      "entropy": 5.678024196624756,
      "epoch": 4.51653843993935,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031336245833499146,
      "loss": 4.9477,
      "mean_token_accuracy": 0.23097314536571503,
      "num_tokens": 89793094.0,
      "step": 40215
    },
    {
      "entropy": 5.66988787651062,
      "epoch": 4.517100016847307,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003133228968747026,
      "loss": 4.9427,
      "mean_token_accuracy": 0.22836648374795915,
      "num_tokens": 89804591.0,
      "step": 40220
    },
    {
      "entropy": 5.744131565093994,
      "epoch": 4.517661593755265,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003132833341941979,
      "loss": 4.922,
      "mean_token_accuracy": 0.23029368072748185,
      "num_tokens": 89815749.0,
      "step": 40225
    },
    {
      "entropy": 5.763515377044678,
      "epoch": 4.518223170663222,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000313243770294737,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22783951908349992,
      "num_tokens": 89826940.0,
      "step": 40230
    },
    {
      "entropy": 5.752949476242065,
      "epoch": 4.51878474757118,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031320420517757955,
      "loss": 5.0143,
      "mean_token_accuracy": 0.22627446204423904,
      "num_tokens": 89838955.0,
      "step": 40235
    },
    {
      "entropy": 5.629997634887696,
      "epoch": 4.519346324479137,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003131646388439854,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22120501548051835,
      "num_tokens": 89850177.0,
      "step": 40240
    },
    {
      "entropy": 5.637545490264893,
      "epoch": 4.519907901387095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003131250712952144,
      "loss": 4.947,
      "mean_token_accuracy": 0.22425841987133027,
      "num_tokens": 89861185.0,
      "step": 40245
    },
    {
      "entropy": 5.732206344604492,
      "epoch": 4.520469478295053,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031308550253252634,
      "loss": 4.9658,
      "mean_token_accuracy": 0.23036589324474335,
      "num_tokens": 89873000.0,
      "step": 40250
    },
    {
      "entropy": 5.689436960220337,
      "epoch": 4.52103105520301,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031304593255718113,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2292705610394478,
      "num_tokens": 89884223.0,
      "step": 40255
    },
    {
      "entropy": 5.780792379379273,
      "epoch": 4.521592632110968,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031300636137043857,
      "loss": 5.0407,
      "mean_token_accuracy": 0.22541432529687883,
      "num_tokens": 89895655.0,
      "step": 40260
    },
    {
      "entropy": 5.675997018814087,
      "epoch": 4.522154209018925,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031296678897355877,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22852724492549897,
      "num_tokens": 89906475.0,
      "step": 40265
    },
    {
      "entropy": 5.626355361938477,
      "epoch": 4.5227157859268825,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003129272153678017,
      "loss": 4.9584,
      "mean_token_accuracy": 0.23081908077001573,
      "num_tokens": 89917523.0,
      "step": 40270
    },
    {
      "entropy": 5.570516967773438,
      "epoch": 4.523277362834841,
      "grad_norm": 1.25,
      "learning_rate": 0.0003128876405544274,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23182060420513154,
      "num_tokens": 89928106.0,
      "step": 40275
    },
    {
      "entropy": 5.663627243041992,
      "epoch": 4.523838939742798,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031284806453469596,
      "loss": 4.8684,
      "mean_token_accuracy": 0.2391984134912491,
      "num_tokens": 89938638.0,
      "step": 40280
    },
    {
      "entropy": 5.69415602684021,
      "epoch": 4.524400516650755,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031280848730986735,
      "loss": 4.8876,
      "mean_token_accuracy": 0.23962705731391906,
      "num_tokens": 89949760.0,
      "step": 40285
    },
    {
      "entropy": 5.782716417312622,
      "epoch": 4.524962093558713,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003127689088812019,
      "loss": 5.0498,
      "mean_token_accuracy": 0.22386223822832108,
      "num_tokens": 89960578.0,
      "step": 40290
    },
    {
      "entropy": 5.638512992858887,
      "epoch": 4.52552367046667,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003127293292499598,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2313542202115059,
      "num_tokens": 89971370.0,
      "step": 40295
    },
    {
      "entropy": 5.676192808151245,
      "epoch": 4.526085247374628,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003126897484174012,
      "loss": 4.8489,
      "mean_token_accuracy": 0.2357883483171463,
      "num_tokens": 89982068.0,
      "step": 40300
    },
    {
      "entropy": 5.672941875457764,
      "epoch": 4.526646824282586,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003126501663847863,
      "loss": 4.8469,
      "mean_token_accuracy": 0.2384112611413002,
      "num_tokens": 89993108.0,
      "step": 40305
    },
    {
      "entropy": 5.679787731170654,
      "epoch": 4.527208401190543,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003126105831533756,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22446459382772446,
      "num_tokens": 90004604.0,
      "step": 40310
    },
    {
      "entropy": 5.725195074081421,
      "epoch": 4.527769978098501,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003125709987244293,
      "loss": 5.0086,
      "mean_token_accuracy": 0.22467595785856248,
      "num_tokens": 90015241.0,
      "step": 40315
    },
    {
      "entropy": 5.570701360702515,
      "epoch": 4.528331555006458,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031253141309920784,
      "loss": 4.808,
      "mean_token_accuracy": 0.2380654975771904,
      "num_tokens": 90026511.0,
      "step": 40320
    },
    {
      "entropy": 5.5838690280914305,
      "epoch": 4.5288931319144154,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031249182627897165,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23769066631793975,
      "num_tokens": 90038994.0,
      "step": 40325
    },
    {
      "entropy": 5.710984897613526,
      "epoch": 4.5294547088223736,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031245223826498116,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21956789046525954,
      "num_tokens": 90049183.0,
      "step": 40330
    },
    {
      "entropy": 5.677455759048462,
      "epoch": 4.530016285730331,
      "grad_norm": 1.171875,
      "learning_rate": 0.00031241264905849687,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23007967174053193,
      "num_tokens": 90060642.0,
      "step": 40335
    },
    {
      "entropy": 5.690221643447876,
      "epoch": 4.530577862638288,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00031237305866077926,
      "loss": 4.9846,
      "mean_token_accuracy": 0.22435180991888046,
      "num_tokens": 90069709.0,
      "step": 40340
    },
    {
      "entropy": 5.702293539047242,
      "epoch": 4.531139439546246,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00031233346707308906,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2279423952102661,
      "num_tokens": 90080219.0,
      "step": 40345
    },
    {
      "entropy": 5.698235511779785,
      "epoch": 4.531701016454203,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031229387429668675,
      "loss": 5.0474,
      "mean_token_accuracy": 0.22524502724409104,
      "num_tokens": 90091769.0,
      "step": 40350
    },
    {
      "entropy": 5.718071985244751,
      "epoch": 4.5322625933621605,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031225428033283305,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2333777964115143,
      "num_tokens": 90102273.0,
      "step": 40355
    },
    {
      "entropy": 5.687039756774903,
      "epoch": 4.532824170270119,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003122146851827886,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2242586374282837,
      "num_tokens": 90112311.0,
      "step": 40360
    },
    {
      "entropy": 5.598856830596924,
      "epoch": 4.533385747178076,
      "grad_norm": 1.203125,
      "learning_rate": 0.00031217508884781413,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23557349145412446,
      "num_tokens": 90123361.0,
      "step": 40365
    },
    {
      "entropy": 5.625381278991699,
      "epoch": 4.533947324086034,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003121354913291704,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23678320944309234,
      "num_tokens": 90134811.0,
      "step": 40370
    },
    {
      "entropy": 5.68803071975708,
      "epoch": 4.534508900993991,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003120958926281182,
      "loss": 4.9601,
      "mean_token_accuracy": 0.2284413680434227,
      "num_tokens": 90145218.0,
      "step": 40375
    },
    {
      "entropy": 5.759547567367553,
      "epoch": 4.535070477901948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031205629274591844,
      "loss": 5.0142,
      "mean_token_accuracy": 0.22876782566308976,
      "num_tokens": 90157217.0,
      "step": 40380
    },
    {
      "entropy": 5.7038157939910885,
      "epoch": 4.5356320548099065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031201669168383193,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22875692695379257,
      "num_tokens": 90167752.0,
      "step": 40385
    },
    {
      "entropy": 5.662690114974976,
      "epoch": 4.536193631717864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00031197708944311964,
      "loss": 4.92,
      "mean_token_accuracy": 0.23237161040306092,
      "num_tokens": 90178309.0,
      "step": 40390
    },
    {
      "entropy": 5.60667314529419,
      "epoch": 4.536755208625821,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003119374860250425,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2415331482887268,
      "num_tokens": 90189158.0,
      "step": 40395
    },
    {
      "entropy": 5.696336984634399,
      "epoch": 4.537316785533779,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003118978814308614,
      "loss": 4.9331,
      "mean_token_accuracy": 0.23051735311746596,
      "num_tokens": 90199611.0,
      "step": 40400
    },
    {
      "entropy": 5.722600841522217,
      "epoch": 4.537878362441736,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003118582756618375,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22638350278139113,
      "num_tokens": 90210852.0,
      "step": 40405
    },
    {
      "entropy": 5.727217292785644,
      "epoch": 4.5384399393496935,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003118186687192318,
      "loss": 4.9595,
      "mean_token_accuracy": 0.227121239900589,
      "num_tokens": 90221193.0,
      "step": 40410
    },
    {
      "entropy": 5.694554138183594,
      "epoch": 4.539001516257652,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031177906060430536,
      "loss": 4.9515,
      "mean_token_accuracy": 0.22803074568510057,
      "num_tokens": 90232811.0,
      "step": 40415
    },
    {
      "entropy": 5.702007722854614,
      "epoch": 4.539563093165609,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031173945131831936,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22804235368967057,
      "num_tokens": 90243566.0,
      "step": 40420
    },
    {
      "entropy": 5.739405012130737,
      "epoch": 4.540124670073567,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000311699840862535,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2383894920349121,
      "num_tokens": 90255292.0,
      "step": 40425
    },
    {
      "entropy": 5.710750675201416,
      "epoch": 4.540686246981524,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003116602292382135,
      "loss": 4.9727,
      "mean_token_accuracy": 0.23036620616912842,
      "num_tokens": 90266691.0,
      "step": 40430
    },
    {
      "entropy": 5.691594505310059,
      "epoch": 4.541247823889481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031162061644661606,
      "loss": 4.9519,
      "mean_token_accuracy": 0.23393805623054503,
      "num_tokens": 90278240.0,
      "step": 40435
    },
    {
      "entropy": 5.752673578262329,
      "epoch": 4.5418094007974394,
      "grad_norm": 1.15625,
      "learning_rate": 0.000311581002489004,
      "loss": 4.9698,
      "mean_token_accuracy": 0.23136602640151976,
      "num_tokens": 90288203.0,
      "step": 40440
    },
    {
      "entropy": 5.666491365432739,
      "epoch": 4.542370977705397,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003115413873666387,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22988309264183043,
      "num_tokens": 90299326.0,
      "step": 40445
    },
    {
      "entropy": 5.654356002807617,
      "epoch": 4.542932554613355,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003115017710807813,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23388026058673858,
      "num_tokens": 90309504.0,
      "step": 40450
    },
    {
      "entropy": 5.6698565006256105,
      "epoch": 4.543494131521312,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003114621536326934,
      "loss": 4.9263,
      "mean_token_accuracy": 0.23329586535692215,
      "num_tokens": 90320919.0,
      "step": 40455
    },
    {
      "entropy": 5.66638412475586,
      "epoch": 4.544055708429269,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003114225350236365,
      "loss": 4.8317,
      "mean_token_accuracy": 0.2401635840535164,
      "num_tokens": 90331864.0,
      "step": 40460
    },
    {
      "entropy": 5.658038711547851,
      "epoch": 4.544617285337227,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003113829152548719,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2292104333639145,
      "num_tokens": 90342563.0,
      "step": 40465
    },
    {
      "entropy": 5.6328905582427975,
      "epoch": 4.5451788622451845,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003113432943276611,
      "loss": 4.8596,
      "mean_token_accuracy": 0.22431473881006242,
      "num_tokens": 90353464.0,
      "step": 40470
    },
    {
      "entropy": 5.660415458679199,
      "epoch": 4.545740439153142,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031130367224326585,
      "loss": 4.9143,
      "mean_token_accuracy": 0.23029388934373857,
      "num_tokens": 90363255.0,
      "step": 40475
    },
    {
      "entropy": 5.730483198165894,
      "epoch": 4.5463020160611,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003112640490029475,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2266380473971367,
      "num_tokens": 90374478.0,
      "step": 40480
    },
    {
      "entropy": 5.702342081069946,
      "epoch": 4.546863592969057,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031122442460796774,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2309937834739685,
      "num_tokens": 90386916.0,
      "step": 40485
    },
    {
      "entropy": 5.618673944473267,
      "epoch": 4.547425169877014,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003111847990595883,
      "loss": 4.8248,
      "mean_token_accuracy": 0.23184494227170943,
      "num_tokens": 90397407.0,
      "step": 40490
    },
    {
      "entropy": 5.6959630966186525,
      "epoch": 4.547986746784972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003111451723590708,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2244402751326561,
      "num_tokens": 90408762.0,
      "step": 40495
    },
    {
      "entropy": 5.680000305175781,
      "epoch": 4.54854832369293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000311105544507677,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2213170722126961,
      "num_tokens": 90418942.0,
      "step": 40500
    },
    {
      "entropy": 5.796164178848267,
      "epoch": 4.549109900600888,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031106591550666865,
      "loss": 5.0821,
      "mean_token_accuracy": 0.22114600241184235,
      "num_tokens": 90430052.0,
      "step": 40505
    },
    {
      "entropy": 5.749198627471924,
      "epoch": 4.549671477508845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003110262853573076,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2316888228058815,
      "num_tokens": 90442631.0,
      "step": 40510
    },
    {
      "entropy": 5.73857102394104,
      "epoch": 4.550233054416802,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003109866540608556,
      "loss": 4.9553,
      "mean_token_accuracy": 0.22665724158287048,
      "num_tokens": 90454429.0,
      "step": 40515
    },
    {
      "entropy": 5.697719287872315,
      "epoch": 4.55079463132476,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031094702161857455,
      "loss": 5.0245,
      "mean_token_accuracy": 0.222223199903965,
      "num_tokens": 90466367.0,
      "step": 40520
    },
    {
      "entropy": 5.720803260803223,
      "epoch": 4.5513562082327175,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031090738803172646,
      "loss": 4.8886,
      "mean_token_accuracy": 0.23342798948287963,
      "num_tokens": 90478480.0,
      "step": 40525
    },
    {
      "entropy": 5.690645456314087,
      "epoch": 4.551917785140675,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00031086775330157324,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22598184049129486,
      "num_tokens": 90490501.0,
      "step": 40530
    },
    {
      "entropy": 5.693047046661377,
      "epoch": 4.552479362048633,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003108281174293768,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22868418842554092,
      "num_tokens": 90500647.0,
      "step": 40535
    },
    {
      "entropy": 5.650712490081787,
      "epoch": 4.55304093895659,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00031078848041639926,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23216028064489364,
      "num_tokens": 90510963.0,
      "step": 40540
    },
    {
      "entropy": 5.757444381713867,
      "epoch": 4.553602515864547,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031074884226390246,
      "loss": 4.9876,
      "mean_token_accuracy": 0.23213592320680618,
      "num_tokens": 90523475.0,
      "step": 40545
    },
    {
      "entropy": 5.688775587081909,
      "epoch": 4.554164092772505,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003107092029731488,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22944957613945008,
      "num_tokens": 90535348.0,
      "step": 40550
    },
    {
      "entropy": 5.61048698425293,
      "epoch": 4.5547256696804626,
      "grad_norm": 1.265625,
      "learning_rate": 0.00031066956254540025,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2309832289814949,
      "num_tokens": 90545841.0,
      "step": 40555
    },
    {
      "entropy": 5.672283887863159,
      "epoch": 4.555287246588421,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00031062992098191897,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22767737060785292,
      "num_tokens": 90557221.0,
      "step": 40560
    },
    {
      "entropy": 5.628676652908325,
      "epoch": 4.555848823496378,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003105902782839672,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23630910962820054,
      "num_tokens": 90568080.0,
      "step": 40565
    },
    {
      "entropy": 5.746728801727295,
      "epoch": 4.556410400404335,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003105506344528071,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22794990986585617,
      "num_tokens": 90579075.0,
      "step": 40570
    },
    {
      "entropy": 5.734940385818481,
      "epoch": 4.556971977312293,
      "grad_norm": 1.265625,
      "learning_rate": 0.000310510989489701,
      "loss": 5.0101,
      "mean_token_accuracy": 0.22940465658903123,
      "num_tokens": 90589937.0,
      "step": 40575
    },
    {
      "entropy": 5.642250394821167,
      "epoch": 4.55753355422025,
      "grad_norm": 1.1875,
      "learning_rate": 0.00031047134339591136,
      "loss": 4.9021,
      "mean_token_accuracy": 0.23865376114845277,
      "num_tokens": 90600262.0,
      "step": 40580
    },
    {
      "entropy": 5.7350249767303465,
      "epoch": 4.558095131128208,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031043169617270033,
      "loss": 5.025,
      "mean_token_accuracy": 0.22653991878032684,
      "num_tokens": 90612161.0,
      "step": 40585
    },
    {
      "entropy": 5.7431769371032715,
      "epoch": 4.558656708036166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031039204782133033,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22598339021205902,
      "num_tokens": 90624304.0,
      "step": 40590
    },
    {
      "entropy": 5.705441474914551,
      "epoch": 4.559218284944123,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003103523983430638,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2237721398472786,
      "num_tokens": 90636196.0,
      "step": 40595
    },
    {
      "entropy": 5.67054934501648,
      "epoch": 4.55977986185208,
      "grad_norm": 1.15625,
      "learning_rate": 0.00031031274773916326,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2334063395857811,
      "num_tokens": 90649024.0,
      "step": 40600
    },
    {
      "entropy": 5.664902257919311,
      "epoch": 4.560341438760038,
      "grad_norm": 1.109375,
      "learning_rate": 0.00031027309601089104,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22534520477056502,
      "num_tokens": 90661110.0,
      "step": 40605
    },
    {
      "entropy": 5.654627656936645,
      "epoch": 4.5609030156679955,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003102334431595098,
      "loss": 4.9195,
      "mean_token_accuracy": 0.230630761384964,
      "num_tokens": 90671691.0,
      "step": 40610
    },
    {
      "entropy": 5.725446701049805,
      "epoch": 4.561464592575954,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00031019378918628203,
      "loss": 4.9004,
      "mean_token_accuracy": 0.23381164222955703,
      "num_tokens": 90683147.0,
      "step": 40615
    },
    {
      "entropy": 5.710891819000244,
      "epoch": 4.562026169483911,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00031015413409247046,
      "loss": 4.9612,
      "mean_token_accuracy": 0.23031733930110931,
      "num_tokens": 90694614.0,
      "step": 40620
    },
    {
      "entropy": 5.707758569717408,
      "epoch": 4.562587746391868,
      "grad_norm": 1.234375,
      "learning_rate": 0.00031011447787933756,
      "loss": 5.0073,
      "mean_token_accuracy": 0.228548264503479,
      "num_tokens": 90705326.0,
      "step": 40625
    },
    {
      "entropy": 5.735672903060913,
      "epoch": 4.563149323299826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003100748205481461,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22590212672948837,
      "num_tokens": 90716140.0,
      "step": 40630
    },
    {
      "entropy": 5.6033485412597654,
      "epoch": 4.563710900207783,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003100351621001587,
      "loss": 4.7553,
      "mean_token_accuracy": 0.24577070027589798,
      "num_tokens": 90727181.0,
      "step": 40635
    },
    {
      "entropy": 5.613107347488404,
      "epoch": 4.5642724771157415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003099955025366382,
      "loss": 4.9362,
      "mean_token_accuracy": 0.23404084593057634,
      "num_tokens": 90737931.0,
      "step": 40640
    },
    {
      "entropy": 5.617407131195068,
      "epoch": 4.564834054023699,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030995584185884733,
      "loss": 4.9418,
      "mean_token_accuracy": 0.23339106142520905,
      "num_tokens": 90749011.0,
      "step": 40645
    },
    {
      "entropy": 5.689836215972901,
      "epoch": 4.565395630931656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003099161800680488,
      "loss": 4.885,
      "mean_token_accuracy": 0.22896224856376649,
      "num_tokens": 90760413.0,
      "step": 40650
    },
    {
      "entropy": 5.66047306060791,
      "epoch": 4.565957207839614,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003098765171655056,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22688617557287216,
      "num_tokens": 90770973.0,
      "step": 40655
    },
    {
      "entropy": 5.710400056838989,
      "epoch": 4.566518784747571,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003098368531524806,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2325441747903824,
      "num_tokens": 90782204.0,
      "step": 40660
    },
    {
      "entropy": 5.722032117843628,
      "epoch": 4.567080361655528,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003097971880302367,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23323638141155242,
      "num_tokens": 90791906.0,
      "step": 40665
    },
    {
      "entropy": 5.677653741836548,
      "epoch": 4.5676419385634865,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030975752180003674,
      "loss": 4.8622,
      "mean_token_accuracy": 0.2357983708381653,
      "num_tokens": 90802153.0,
      "step": 40670
    },
    {
      "entropy": 5.653422832489014,
      "epoch": 4.568203515471444,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000309717854463144,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23178379833698273,
      "num_tokens": 90813129.0,
      "step": 40675
    },
    {
      "entropy": 5.71180830001831,
      "epoch": 4.568765092379401,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003096781860208211,
      "loss": 4.9757,
      "mean_token_accuracy": 0.22509945631027223,
      "num_tokens": 90823445.0,
      "step": 40680
    },
    {
      "entropy": 5.701697587966919,
      "epoch": 4.569326669287359,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003096385164743314,
      "loss": 4.9295,
      "mean_token_accuracy": 0.22542690932750703,
      "num_tokens": 90834823.0,
      "step": 40685
    },
    {
      "entropy": 5.594730567932129,
      "epoch": 4.569888246195316,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030959884582493793,
      "loss": 4.7932,
      "mean_token_accuracy": 0.23930739462375641,
      "num_tokens": 90845503.0,
      "step": 40690
    },
    {
      "entropy": 5.75843653678894,
      "epoch": 4.570449823103274,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003095591740739038,
      "loss": 4.9968,
      "mean_token_accuracy": 0.22445562183856965,
      "num_tokens": 90857562.0,
      "step": 40695
    },
    {
      "entropy": 5.73058500289917,
      "epoch": 4.571011400011232,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003095195012224921,
      "loss": 4.9878,
      "mean_token_accuracy": 0.2286472424864769,
      "num_tokens": 90868207.0,
      "step": 40700
    },
    {
      "entropy": 5.623605155944825,
      "epoch": 4.571572976919189,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003094798272719661,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22742082923650742,
      "num_tokens": 90879710.0,
      "step": 40705
    },
    {
      "entropy": 5.708581352233887,
      "epoch": 4.572134553827147,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030944015222358906,
      "loss": 4.9182,
      "mean_token_accuracy": 0.22997390180826188,
      "num_tokens": 90890702.0,
      "step": 40710
    },
    {
      "entropy": 5.668360185623169,
      "epoch": 4.572696130735104,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030940047607862415,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23105577975511551,
      "num_tokens": 90902171.0,
      "step": 40715
    },
    {
      "entropy": 5.692994832992554,
      "epoch": 4.573257707643061,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003093607988383348,
      "loss": 4.9008,
      "mean_token_accuracy": 0.228960120677948,
      "num_tokens": 90913735.0,
      "step": 40720
    },
    {
      "entropy": 5.691771936416626,
      "epoch": 4.5738192845510195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003093211205039842,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22639254927635194,
      "num_tokens": 90924137.0,
      "step": 40725
    },
    {
      "entropy": 5.651413917541504,
      "epoch": 4.574380861458977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003092814410768359,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23344521671533586,
      "num_tokens": 90935214.0,
      "step": 40730
    },
    {
      "entropy": 5.643724203109741,
      "epoch": 4.574942438366934,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030924176055815327,
      "loss": 4.8743,
      "mean_token_accuracy": 0.23702867925167084,
      "num_tokens": 90946797.0,
      "step": 40735
    },
    {
      "entropy": 5.633860874176025,
      "epoch": 4.575504015274892,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030920207894919957,
      "loss": 4.8986,
      "mean_token_accuracy": 0.23298589289188384,
      "num_tokens": 90958055.0,
      "step": 40740
    },
    {
      "entropy": 5.651974534988403,
      "epoch": 4.576065592182849,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003091623962512385,
      "loss": 4.8848,
      "mean_token_accuracy": 0.23526354879140854,
      "num_tokens": 90969516.0,
      "step": 40745
    },
    {
      "entropy": 5.599722957611084,
      "epoch": 4.576627169090807,
      "grad_norm": 1.25,
      "learning_rate": 0.00030912271246553346,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2300792708992958,
      "num_tokens": 90980536.0,
      "step": 40750
    },
    {
      "entropy": 5.683519268035889,
      "epoch": 4.577188745998765,
      "grad_norm": 1.484375,
      "learning_rate": 0.000309083027593348,
      "loss": 4.9704,
      "mean_token_accuracy": 0.23771483451128006,
      "num_tokens": 90992150.0,
      "step": 40755
    },
    {
      "entropy": 5.678859710693359,
      "epoch": 4.577750322906722,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030904334163594577,
      "loss": 4.9291,
      "mean_token_accuracy": 0.22791082561016082,
      "num_tokens": 91003323.0,
      "step": 40760
    },
    {
      "entropy": 5.634065008163452,
      "epoch": 4.57831189981468,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003090036545945904,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2342466726899147,
      "num_tokens": 91014238.0,
      "step": 40765
    },
    {
      "entropy": 5.686384868621826,
      "epoch": 4.578873476722637,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00030896396647054535,
      "loss": 4.9409,
      "mean_token_accuracy": 0.22766696512699128,
      "num_tokens": 91025375.0,
      "step": 40770
    },
    {
      "entropy": 5.629166984558106,
      "epoch": 4.579435053630594,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003089242772650744,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22669413089752197,
      "num_tokens": 91036498.0,
      "step": 40775
    },
    {
      "entropy": 5.646253871917724,
      "epoch": 4.579996630538552,
      "grad_norm": 1.25,
      "learning_rate": 0.00030888458697944144,
      "loss": 4.9022,
      "mean_token_accuracy": 0.2321141168475151,
      "num_tokens": 91046909.0,
      "step": 40780
    },
    {
      "entropy": 5.750164270401001,
      "epoch": 4.58055820744651,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030884489561491005,
      "loss": 5.0298,
      "mean_token_accuracy": 0.2231770470738411,
      "num_tokens": 91056052.0,
      "step": 40785
    },
    {
      "entropy": 5.853788614273071,
      "epoch": 4.581119784354467,
      "grad_norm": 1.15625,
      "learning_rate": 0.000308805203172744,
      "loss": 5.1136,
      "mean_token_accuracy": 0.21537838280200958,
      "num_tokens": 91067614.0,
      "step": 40790
    },
    {
      "entropy": 5.615201330184936,
      "epoch": 4.581681361262425,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030876550965420725,
      "loss": 4.8548,
      "mean_token_accuracy": 0.23368002474308014,
      "num_tokens": 91078539.0,
      "step": 40795
    },
    {
      "entropy": 5.696202135086059,
      "epoch": 4.582242938170382,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003087258150605635,
      "loss": 4.941,
      "mean_token_accuracy": 0.22995708733797074,
      "num_tokens": 91088909.0,
      "step": 40800
    },
    {
      "entropy": 5.645778703689575,
      "epoch": 4.58280451507834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003086861193930767,
      "loss": 4.9478,
      "mean_token_accuracy": 0.23182020783424379,
      "num_tokens": 91101073.0,
      "step": 40805
    },
    {
      "entropy": 5.688597536087036,
      "epoch": 4.5833660919862975,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003086464226530109,
      "loss": 4.9067,
      "mean_token_accuracy": 0.23258307129144667,
      "num_tokens": 91113518.0,
      "step": 40810
    },
    {
      "entropy": 5.7378839492797855,
      "epoch": 4.583927668894255,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030860672484162987,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22859539240598678,
      "num_tokens": 91124287.0,
      "step": 40815
    },
    {
      "entropy": 5.7605188369750975,
      "epoch": 4.584489245802213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003085670259601977,
      "loss": 4.9496,
      "mean_token_accuracy": 0.23203171640634537,
      "num_tokens": 91134828.0,
      "step": 40820
    },
    {
      "entropy": 5.695339822769165,
      "epoch": 4.58505082271017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00030852732600997846,
      "loss": 4.9715,
      "mean_token_accuracy": 0.2240214616060257,
      "num_tokens": 91146647.0,
      "step": 40825
    },
    {
      "entropy": 5.653254508972168,
      "epoch": 4.585612399618128,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030848762499223616,
      "loss": 4.9871,
      "mean_token_accuracy": 0.2268601343035698,
      "num_tokens": 91156769.0,
      "step": 40830
    },
    {
      "entropy": 5.720983505249023,
      "epoch": 4.586173976526085,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003084479229082348,
      "loss": 4.865,
      "mean_token_accuracy": 0.23836442828178406,
      "num_tokens": 91168876.0,
      "step": 40835
    },
    {
      "entropy": 5.824823713302612,
      "epoch": 4.586735553434043,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003084082197592387,
      "loss": 5.0676,
      "mean_token_accuracy": 0.22678182125091553,
      "num_tokens": 91178587.0,
      "step": 40840
    },
    {
      "entropy": 5.628231620788574,
      "epoch": 4.587297130342001,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030836851554651184,
      "loss": 4.8333,
      "mean_token_accuracy": 0.24113062173128127,
      "num_tokens": 91190210.0,
      "step": 40845
    },
    {
      "entropy": 5.681164932250977,
      "epoch": 4.587858707249958,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00030832881027131854,
      "loss": 4.962,
      "mean_token_accuracy": 0.22689961045980453,
      "num_tokens": 91200921.0,
      "step": 40850
    },
    {
      "entropy": 5.6778946876525875,
      "epoch": 4.588420284157915,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030828910393492306,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2321982204914093,
      "num_tokens": 91211553.0,
      "step": 40855
    },
    {
      "entropy": 5.733036088943481,
      "epoch": 4.588981861065873,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030824939653858953,
      "loss": 4.9501,
      "mean_token_accuracy": 0.23049245923757553,
      "num_tokens": 91222224.0,
      "step": 40860
    },
    {
      "entropy": 5.738907432556152,
      "epoch": 4.5895434379738305,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030820968808358223,
      "loss": 4.8976,
      "mean_token_accuracy": 0.231520314514637,
      "num_tokens": 91233645.0,
      "step": 40865
    },
    {
      "entropy": 5.718350172042847,
      "epoch": 4.590105014881788,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030816997857116573,
      "loss": 4.9437,
      "mean_token_accuracy": 0.23054262101650239,
      "num_tokens": 91245194.0,
      "step": 40870
    },
    {
      "entropy": 5.65810341835022,
      "epoch": 4.590666591789746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003081302680026042,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22758001387119292,
      "num_tokens": 91255124.0,
      "step": 40875
    },
    {
      "entropy": 5.736838388442993,
      "epoch": 4.591228168697703,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000308090556379162,
      "loss": 5.0471,
      "mean_token_accuracy": 0.22186787128448487,
      "num_tokens": 91267129.0,
      "step": 40880
    },
    {
      "entropy": 5.671534585952759,
      "epoch": 4.591789745605661,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030805084370210377,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22670916765928267,
      "num_tokens": 91277614.0,
      "step": 40885
    },
    {
      "entropy": 5.714290380477905,
      "epoch": 4.592351322513618,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00030801112997269386,
      "loss": 4.9125,
      "mean_token_accuracy": 0.22797329127788543,
      "num_tokens": 91289856.0,
      "step": 40890
    },
    {
      "entropy": 5.644074010848999,
      "epoch": 4.5929128994215755,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003079714151921967,
      "loss": 4.942,
      "mean_token_accuracy": 0.23320803195238113,
      "num_tokens": 91300593.0,
      "step": 40895
    },
    {
      "entropy": 5.693307590484619,
      "epoch": 4.593474476329534,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030793169936187695,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2392889752984047,
      "num_tokens": 91312934.0,
      "step": 40900
    },
    {
      "entropy": 5.628866672515869,
      "epoch": 4.594036053237491,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030789198248299906,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2300572156906128,
      "num_tokens": 91323856.0,
      "step": 40905
    },
    {
      "entropy": 5.728311252593994,
      "epoch": 4.594597630145448,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030785226455682773,
      "loss": 4.9626,
      "mean_token_accuracy": 0.2282801404595375,
      "num_tokens": 91334545.0,
      "step": 40910
    },
    {
      "entropy": 5.765360593795776,
      "epoch": 4.595159207053406,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003078125455846276,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2307359382510185,
      "num_tokens": 91346006.0,
      "step": 40915
    },
    {
      "entropy": 5.71982045173645,
      "epoch": 4.595720783961363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00030777282556766316,
      "loss": 4.9568,
      "mean_token_accuracy": 0.22193027287721634,
      "num_tokens": 91356710.0,
      "step": 40920
    },
    {
      "entropy": 5.686034488677978,
      "epoch": 4.596282360869321,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030773310450719937,
      "loss": 4.8714,
      "mean_token_accuracy": 0.23181068301200866,
      "num_tokens": 91367747.0,
      "step": 40925
    },
    {
      "entropy": 5.739812803268433,
      "epoch": 4.596843937777279,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003076933824045008,
      "loss": 4.9875,
      "mean_token_accuracy": 0.22842379063367843,
      "num_tokens": 91379079.0,
      "step": 40930
    },
    {
      "entropy": 5.693708610534668,
      "epoch": 4.597405514685236,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030765365926083214,
      "loss": 4.9982,
      "mean_token_accuracy": 0.22464492470026015,
      "num_tokens": 91390264.0,
      "step": 40935
    },
    {
      "entropy": 5.677793836593628,
      "epoch": 4.597967091593194,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003076139350774584,
      "loss": 4.874,
      "mean_token_accuracy": 0.23726072758436204,
      "num_tokens": 91401322.0,
      "step": 40940
    },
    {
      "entropy": 5.67651481628418,
      "epoch": 4.598528668501151,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003075742098556444,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22813715636730195,
      "num_tokens": 91412277.0,
      "step": 40945
    },
    {
      "entropy": 5.6240216255187985,
      "epoch": 4.5990902454091085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003075344835966548,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23053689002990724,
      "num_tokens": 91422941.0,
      "step": 40950
    },
    {
      "entropy": 5.646733808517456,
      "epoch": 4.599651822317067,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003074947563017547,
      "loss": 4.9035,
      "mean_token_accuracy": 0.23553151041269302,
      "num_tokens": 91434194.0,
      "step": 40955
    },
    {
      "entropy": 5.656481552124023,
      "epoch": 4.600213399225024,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003074550279722089,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22827233821153642,
      "num_tokens": 91445330.0,
      "step": 40960
    },
    {
      "entropy": 5.731833839416504,
      "epoch": 4.600774976132981,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003074152986092825,
      "loss": 4.938,
      "mean_token_accuracy": 0.22876531928777694,
      "num_tokens": 91456762.0,
      "step": 40965
    },
    {
      "entropy": 5.679109287261963,
      "epoch": 4.601336553040939,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003073755682142404,
      "loss": 4.8297,
      "mean_token_accuracy": 0.23037175834178925,
      "num_tokens": 91467272.0,
      "step": 40970
    },
    {
      "entropy": 5.739348602294922,
      "epoch": 4.601898129948896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030733583678834765,
      "loss": 5.0507,
      "mean_token_accuracy": 0.22063138484954833,
      "num_tokens": 91478188.0,
      "step": 40975
    },
    {
      "entropy": 5.660824775695801,
      "epoch": 4.602459706856854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030729610433286926,
      "loss": 4.9238,
      "mean_token_accuracy": 0.23059858679771422,
      "num_tokens": 91489759.0,
      "step": 40980
    },
    {
      "entropy": 5.704706859588623,
      "epoch": 4.603021283764812,
      "grad_norm": 1.25,
      "learning_rate": 0.00030725637084907037,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22958252131938933,
      "num_tokens": 91500705.0,
      "step": 40985
    },
    {
      "entropy": 5.647174739837647,
      "epoch": 4.603582860672769,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003072166363382162,
      "loss": 4.955,
      "mean_token_accuracy": 0.22842477709054948,
      "num_tokens": 91511832.0,
      "step": 40990
    },
    {
      "entropy": 5.731711483001709,
      "epoch": 4.604144437580727,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003071769008015717,
      "loss": 4.9453,
      "mean_token_accuracy": 0.23026622235774993,
      "num_tokens": 91522454.0,
      "step": 40995
    },
    {
      "entropy": 5.725698089599609,
      "epoch": 4.604706014488684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003071371642404024,
      "loss": 5.0202,
      "mean_token_accuracy": 0.22544444650411605,
      "num_tokens": 91533987.0,
      "step": 41000
    },
    {
      "entropy": 5.6946549892425535,
      "epoch": 4.605267591396641,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003070974266559732,
      "loss": 4.928,
      "mean_token_accuracy": 0.2295356810092926,
      "num_tokens": 91545000.0,
      "step": 41005
    },
    {
      "entropy": 5.797507905960083,
      "epoch": 4.6058291683045995,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003070576880495495,
      "loss": 4.9972,
      "mean_token_accuracy": 0.22597256898880005,
      "num_tokens": 91556481.0,
      "step": 41010
    },
    {
      "entropy": 5.681703186035156,
      "epoch": 4.606390745212557,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003070179484223966,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23855579346418382,
      "num_tokens": 91568668.0,
      "step": 41015
    },
    {
      "entropy": 5.62876148223877,
      "epoch": 4.606952322120515,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003069782077757797,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2369738221168518,
      "num_tokens": 91579884.0,
      "step": 41020
    },
    {
      "entropy": 5.74406099319458,
      "epoch": 4.607513899028472,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030693846611096437,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2334261879324913,
      "num_tokens": 91590488.0,
      "step": 41025
    },
    {
      "entropy": 5.724934053421021,
      "epoch": 4.608075475936429,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030689872342921595,
      "loss": 4.9121,
      "mean_token_accuracy": 0.22945201843976976,
      "num_tokens": 91602163.0,
      "step": 41030
    },
    {
      "entropy": 5.644447088241577,
      "epoch": 4.608637052844387,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030685897973179967,
      "loss": 4.8972,
      "mean_token_accuracy": 0.2341543674468994,
      "num_tokens": 91614251.0,
      "step": 41035
    },
    {
      "entropy": 5.6660631656646725,
      "epoch": 4.609198629752345,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030681923501998124,
      "loss": 4.9256,
      "mean_token_accuracy": 0.23266985267400742,
      "num_tokens": 91625102.0,
      "step": 41040
    },
    {
      "entropy": 5.773362302780152,
      "epoch": 4.609760206660302,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030677948929502596,
      "loss": 4.9814,
      "mean_token_accuracy": 0.22841348201036454,
      "num_tokens": 91636613.0,
      "step": 41045
    },
    {
      "entropy": 5.696863031387329,
      "epoch": 4.61032178356826,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030673974255819944,
      "loss": 4.8819,
      "mean_token_accuracy": 0.23310887962579727,
      "num_tokens": 91647648.0,
      "step": 41050
    },
    {
      "entropy": 5.831459426879883,
      "epoch": 4.610883360476217,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003066999948107672,
      "loss": 5.0441,
      "mean_token_accuracy": 0.22308216392993926,
      "num_tokens": 91659160.0,
      "step": 41055
    },
    {
      "entropy": 5.692764520645142,
      "epoch": 4.611444937384174,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003066602460539948,
      "loss": 4.8969,
      "mean_token_accuracy": 0.23525327891111375,
      "num_tokens": 91670973.0,
      "step": 41060
    },
    {
      "entropy": 5.73063063621521,
      "epoch": 4.6120065142921325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030662049628914787,
      "loss": 5.0341,
      "mean_token_accuracy": 0.2272852897644043,
      "num_tokens": 91682710.0,
      "step": 41065
    },
    {
      "entropy": 5.756023597717285,
      "epoch": 4.61256809120009,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003065807455174921,
      "loss": 4.9314,
      "mean_token_accuracy": 0.22493216693401336,
      "num_tokens": 91694008.0,
      "step": 41070
    },
    {
      "entropy": 5.625627565383911,
      "epoch": 4.613129668108048,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003065409937402932,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2264292061328888,
      "num_tokens": 91705289.0,
      "step": 41075
    },
    {
      "entropy": 5.695087432861328,
      "epoch": 4.613691245016005,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030650124095881674,
      "loss": 4.9157,
      "mean_token_accuracy": 0.2286624491214752,
      "num_tokens": 91717078.0,
      "step": 41080
    },
    {
      "entropy": 5.659264373779297,
      "epoch": 4.614252821923962,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030646148717432855,
      "loss": 4.9346,
      "mean_token_accuracy": 0.23230470567941666,
      "num_tokens": 91728971.0,
      "step": 41085
    },
    {
      "entropy": 5.665151500701905,
      "epoch": 4.61481439883192,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030642173238809453,
      "loss": 4.8673,
      "mean_token_accuracy": 0.2342691332101822,
      "num_tokens": 91739496.0,
      "step": 41090
    },
    {
      "entropy": 5.7210596084594725,
      "epoch": 4.615375975739878,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003063819766013802,
      "loss": 4.9461,
      "mean_token_accuracy": 0.22467200756072997,
      "num_tokens": 91749732.0,
      "step": 41095
    },
    {
      "entropy": 5.6386803150177,
      "epoch": 4.615937552647835,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003063422198154516,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22222595959901809,
      "num_tokens": 91758772.0,
      "step": 41100
    },
    {
      "entropy": 5.6695856094360355,
      "epoch": 4.616499129555793,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003063024620315746,
      "loss": 4.8566,
      "mean_token_accuracy": 0.23571252822875977,
      "num_tokens": 91770012.0,
      "step": 41105
    },
    {
      "entropy": 5.698201465606689,
      "epoch": 4.61706070646375,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003062627032510151,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22528381049633026,
      "num_tokens": 91781926.0,
      "step": 41110
    },
    {
      "entropy": 5.732100200653076,
      "epoch": 4.617622283371707,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000306222943475039,
      "loss": 5.038,
      "mean_token_accuracy": 0.22743409276008605,
      "num_tokens": 91794598.0,
      "step": 41115
    },
    {
      "entropy": 5.712809419631958,
      "epoch": 4.618183860279665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003061831827049123,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2284320577979088,
      "num_tokens": 91805898.0,
      "step": 41120
    },
    {
      "entropy": 5.7333903312683105,
      "epoch": 4.618745437187623,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00030614342094190095,
      "loss": 4.942,
      "mean_token_accuracy": 0.22724615186452865,
      "num_tokens": 91816369.0,
      "step": 41125
    },
    {
      "entropy": 5.566758632659912,
      "epoch": 4.619307014095581,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030610365818727097,
      "loss": 4.845,
      "mean_token_accuracy": 0.23104132413864137,
      "num_tokens": 91827223.0,
      "step": 41130
    },
    {
      "entropy": 5.686605453491211,
      "epoch": 4.619868591003538,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003060638944422884,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22357949763536453,
      "num_tokens": 91838898.0,
      "step": 41135
    },
    {
      "entropy": 5.734947729110718,
      "epoch": 4.620430167911495,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030602412970821944,
      "loss": 4.9147,
      "mean_token_accuracy": 0.22883488237857819,
      "num_tokens": 91850660.0,
      "step": 41140
    },
    {
      "entropy": 5.6917236328125,
      "epoch": 4.620991744819453,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003059843639863301,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22705120891332625,
      "num_tokens": 91860594.0,
      "step": 41145
    },
    {
      "entropy": 5.648229694366455,
      "epoch": 4.6215533217274105,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030594459727788665,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22608633190393448,
      "num_tokens": 91872209.0,
      "step": 41150
    },
    {
      "entropy": 5.733660173416138,
      "epoch": 4.622114898635368,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030590482958415524,
      "loss": 4.9574,
      "mean_token_accuracy": 0.23027956783771514,
      "num_tokens": 91885241.0,
      "step": 41155
    },
    {
      "entropy": 5.808947563171387,
      "epoch": 4.622676475543326,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030586506090640205,
      "loss": 5.0,
      "mean_token_accuracy": 0.23667083382606507,
      "num_tokens": 91896999.0,
      "step": 41160
    },
    {
      "entropy": 5.655390834808349,
      "epoch": 4.623238052451283,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003058252912458933,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22902296781539916,
      "num_tokens": 91908554.0,
      "step": 41165
    },
    {
      "entropy": 5.642778968811035,
      "epoch": 4.62379962935924,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003057855206038954,
      "loss": 4.9276,
      "mean_token_accuracy": 0.22972759306430818,
      "num_tokens": 91920013.0,
      "step": 41170
    },
    {
      "entropy": 5.711262655258179,
      "epoch": 4.624361206267198,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003057457489816745,
      "loss": 4.9611,
      "mean_token_accuracy": 0.23094866573810577,
      "num_tokens": 91930951.0,
      "step": 41175
    },
    {
      "entropy": 5.6934586524963375,
      "epoch": 4.624922783175156,
      "grad_norm": 1.125,
      "learning_rate": 0.0003057059763804971,
      "loss": 4.9361,
      "mean_token_accuracy": 0.23319024592638016,
      "num_tokens": 91941718.0,
      "step": 41180
    },
    {
      "entropy": 5.612945032119751,
      "epoch": 4.625484360083114,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003056662028016295,
      "loss": 4.8409,
      "mean_token_accuracy": 0.235760797560215,
      "num_tokens": 91952756.0,
      "step": 41185
    },
    {
      "entropy": 5.652544784545898,
      "epoch": 4.626045936991071,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030562642824633805,
      "loss": 4.9169,
      "mean_token_accuracy": 0.23120157271623612,
      "num_tokens": 91964200.0,
      "step": 41190
    },
    {
      "entropy": 5.699527931213379,
      "epoch": 4.626607513899028,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003055866527158893,
      "loss": 4.9663,
      "mean_token_accuracy": 0.22123746424913407,
      "num_tokens": 91975032.0,
      "step": 41195
    },
    {
      "entropy": 5.690155267715454,
      "epoch": 4.627169090806986,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003055468762115497,
      "loss": 4.874,
      "mean_token_accuracy": 0.23016518205404282,
      "num_tokens": 91986566.0,
      "step": 41200
    },
    {
      "entropy": 5.6599804878234865,
      "epoch": 4.6277306677149435,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030550709873458567,
      "loss": 4.8804,
      "mean_token_accuracy": 0.23696139305830002,
      "num_tokens": 91996944.0,
      "step": 41205
    },
    {
      "entropy": 5.692424488067627,
      "epoch": 4.628292244622902,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030546732028626373,
      "loss": 4.875,
      "mean_token_accuracy": 0.23614908903837203,
      "num_tokens": 92007692.0,
      "step": 41210
    },
    {
      "entropy": 5.680901527404785,
      "epoch": 4.628853821530859,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030542754086785055,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2368366003036499,
      "num_tokens": 92017514.0,
      "step": 41215
    },
    {
      "entropy": 5.683542108535766,
      "epoch": 4.629415398438816,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003053877604806127,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22953083068132402,
      "num_tokens": 92028311.0,
      "step": 41220
    },
    {
      "entropy": 5.598912382125855,
      "epoch": 4.629976975346774,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030534797912581664,
      "loss": 4.8406,
      "mean_token_accuracy": 0.23728302568197251,
      "num_tokens": 92038795.0,
      "step": 41225
    },
    {
      "entropy": 5.5770008087158205,
      "epoch": 4.630538552254731,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030530819680472926,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23030556738376617,
      "num_tokens": 92049578.0,
      "step": 41230
    },
    {
      "entropy": 5.7828113555908205,
      "epoch": 4.6311001291626885,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030526841351861717,
      "loss": 5.08,
      "mean_token_accuracy": 0.22209900617599487,
      "num_tokens": 92062868.0,
      "step": 41235
    },
    {
      "entropy": 5.778105545043945,
      "epoch": 4.631661706070647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030522862926874697,
      "loss": 4.9379,
      "mean_token_accuracy": 0.23563838601112366,
      "num_tokens": 92073819.0,
      "step": 41240
    },
    {
      "entropy": 5.739188528060913,
      "epoch": 4.632223282978604,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003051888440563855,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2182888478040695,
      "num_tokens": 92085212.0,
      "step": 41245
    },
    {
      "entropy": 5.718963670730591,
      "epoch": 4.632784859886561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003051490578827996,
      "loss": 4.9619,
      "mean_token_accuracy": 0.22849655747413636,
      "num_tokens": 92096807.0,
      "step": 41250
    },
    {
      "entropy": 5.714445781707764,
      "epoch": 4.633346436794519,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030510927074925594,
      "loss": 4.9042,
      "mean_token_accuracy": 0.2339370921254158,
      "num_tokens": 92108619.0,
      "step": 41255
    },
    {
      "entropy": 5.7203411102294925,
      "epoch": 4.633908013702476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030506948265702144,
      "loss": 5.0153,
      "mean_token_accuracy": 0.22963874638080597,
      "num_tokens": 92120815.0,
      "step": 41260
    },
    {
      "entropy": 5.63869194984436,
      "epoch": 4.6344695906104345,
      "grad_norm": 1.203125,
      "learning_rate": 0.000305029693607363,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2341860517859459,
      "num_tokens": 92131957.0,
      "step": 41265
    },
    {
      "entropy": 5.650761127471924,
      "epoch": 4.635031167518392,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030498990360154746,
      "loss": 4.8884,
      "mean_token_accuracy": 0.23543034195899964,
      "num_tokens": 92143936.0,
      "step": 41270
    },
    {
      "entropy": 5.677337694168091,
      "epoch": 4.635592744426349,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030495011264084165,
      "loss": 4.9331,
      "mean_token_accuracy": 0.23378072530031205,
      "num_tokens": 92155320.0,
      "step": 41275
    },
    {
      "entropy": 5.712032794952393,
      "epoch": 4.636154321334307,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030491032072651266,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2347971811890602,
      "num_tokens": 92166831.0,
      "step": 41280
    },
    {
      "entropy": 5.721610021591187,
      "epoch": 4.636715898242264,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030487052785982757,
      "loss": 4.9421,
      "mean_token_accuracy": 0.22581842094659804,
      "num_tokens": 92177927.0,
      "step": 41285
    },
    {
      "entropy": 5.714066886901856,
      "epoch": 4.6372774751502215,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003048307340420532,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22286898046731948,
      "num_tokens": 92188755.0,
      "step": 41290
    },
    {
      "entropy": 5.685167503356934,
      "epoch": 4.63783905205818,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003047909392744566,
      "loss": 4.8818,
      "mean_token_accuracy": 0.23188982009887696,
      "num_tokens": 92199156.0,
      "step": 41295
    },
    {
      "entropy": 5.670714426040649,
      "epoch": 4.638400628966137,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030475114355830513,
      "loss": 4.8784,
      "mean_token_accuracy": 0.23589251041412354,
      "num_tokens": 92211080.0,
      "step": 41300
    },
    {
      "entropy": 5.715793180465698,
      "epoch": 4.638962205874094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030471134689486563,
      "loss": 4.9925,
      "mean_token_accuracy": 0.22419875115156174,
      "num_tokens": 92220734.0,
      "step": 41305
    },
    {
      "entropy": 5.7334390640258786,
      "epoch": 4.639523782782052,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003046715492854053,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21828629076480865,
      "num_tokens": 92231166.0,
      "step": 41310
    },
    {
      "entropy": 5.733653926849366,
      "epoch": 4.640085359690009,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030463175073119133,
      "loss": 4.9484,
      "mean_token_accuracy": 0.2277468129992485,
      "num_tokens": 92241920.0,
      "step": 41315
    },
    {
      "entropy": 5.673638343811035,
      "epoch": 4.6406469365979675,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030459195123349095,
      "loss": 4.9308,
      "mean_token_accuracy": 0.228682079911232,
      "num_tokens": 92253745.0,
      "step": 41320
    },
    {
      "entropy": 5.65122356414795,
      "epoch": 4.641208513505925,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003045521507935714,
      "loss": 4.939,
      "mean_token_accuracy": 0.23084062933921815,
      "num_tokens": 92263672.0,
      "step": 41325
    },
    {
      "entropy": 5.7329517841339115,
      "epoch": 4.641770090413882,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003045123494126999,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22719368785619737,
      "num_tokens": 92275124.0,
      "step": 41330
    },
    {
      "entropy": 5.7913518905639645,
      "epoch": 4.64233166732184,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003044725470921437,
      "loss": 5.0029,
      "mean_token_accuracy": 0.22609869241714478,
      "num_tokens": 92287077.0,
      "step": 41335
    },
    {
      "entropy": 5.739186000823975,
      "epoch": 4.642893244229797,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003044327438331702,
      "loss": 5.0082,
      "mean_token_accuracy": 0.228166663646698,
      "num_tokens": 92298649.0,
      "step": 41340
    },
    {
      "entropy": 5.634911251068115,
      "epoch": 4.643454821137754,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003043929396370467,
      "loss": 4.8844,
      "mean_token_accuracy": 0.22878722697496415,
      "num_tokens": 92309589.0,
      "step": 41345
    },
    {
      "entropy": 5.579046678543091,
      "epoch": 4.6440163980457125,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00030435313450504056,
      "loss": 4.8623,
      "mean_token_accuracy": 0.23134008347988128,
      "num_tokens": 92319927.0,
      "step": 41350
    },
    {
      "entropy": 5.689573144912719,
      "epoch": 4.64457797495367,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003043133284384193,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22742380946874619,
      "num_tokens": 92333156.0,
      "step": 41355
    },
    {
      "entropy": 5.6078057289123535,
      "epoch": 4.645139551861627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003042735214384503,
      "loss": 4.7974,
      "mean_token_accuracy": 0.23666640967130662,
      "num_tokens": 92345228.0,
      "step": 41360
    },
    {
      "entropy": 5.725588941574097,
      "epoch": 4.645701128769585,
      "grad_norm": 1.265625,
      "learning_rate": 0.000304233713506401,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2282459869980812,
      "num_tokens": 92355922.0,
      "step": 41365
    },
    {
      "entropy": 5.681649160385132,
      "epoch": 4.646262705677542,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003041939046435389,
      "loss": 4.9538,
      "mean_token_accuracy": 0.23175796270370483,
      "num_tokens": 92367255.0,
      "step": 41370
    },
    {
      "entropy": 5.6193342208862305,
      "epoch": 4.6468242825855,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030415409485113156,
      "loss": 4.8859,
      "mean_token_accuracy": 0.23491671681404114,
      "num_tokens": 92378577.0,
      "step": 41375
    },
    {
      "entropy": 5.704701614379883,
      "epoch": 4.647385859493458,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030411428413044657,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22846240550279617,
      "num_tokens": 92389822.0,
      "step": 41380
    },
    {
      "entropy": 5.783034706115723,
      "epoch": 4.647947436401415,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030407447248275137,
      "loss": 5.0406,
      "mean_token_accuracy": 0.2223149910569191,
      "num_tokens": 92400682.0,
      "step": 41385
    },
    {
      "entropy": 5.756892538070678,
      "epoch": 4.648509013309373,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003040346599093138,
      "loss": 4.92,
      "mean_token_accuracy": 0.2363565132021904,
      "num_tokens": 92412194.0,
      "step": 41390
    },
    {
      "entropy": 5.7249046802520756,
      "epoch": 4.64907059021733,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003039948464114014,
      "loss": 4.913,
      "mean_token_accuracy": 0.23253780454397202,
      "num_tokens": 92423311.0,
      "step": 41395
    },
    {
      "entropy": 5.66365385055542,
      "epoch": 4.649632167125288,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030395503199028186,
      "loss": 5.0268,
      "mean_token_accuracy": 0.22669171988964082,
      "num_tokens": 92434008.0,
      "step": 41400
    },
    {
      "entropy": 5.633940649032593,
      "epoch": 4.6501937440332455,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003039152166472228,
      "loss": 4.8859,
      "mean_token_accuracy": 0.2374495342373848,
      "num_tokens": 92445887.0,
      "step": 41405
    },
    {
      "entropy": 5.75937328338623,
      "epoch": 4.650755320941203,
      "grad_norm": 1.125,
      "learning_rate": 0.0003038754003834921,
      "loss": 4.9947,
      "mean_token_accuracy": 0.23641966581344603,
      "num_tokens": 92457856.0,
      "step": 41410
    },
    {
      "entropy": 5.765560626983643,
      "epoch": 4.651316897849161,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030383558320035744,
      "loss": 4.9836,
      "mean_token_accuracy": 0.23398985713720322,
      "num_tokens": 92469976.0,
      "step": 41415
    },
    {
      "entropy": 5.646824979782105,
      "epoch": 4.651878474757118,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030379576509908656,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22619887739419936,
      "num_tokens": 92481035.0,
      "step": 41420
    },
    {
      "entropy": 5.637125062942505,
      "epoch": 4.652440051665075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030375594608094737,
      "loss": 4.959,
      "mean_token_accuracy": 0.22585269063711166,
      "num_tokens": 92492145.0,
      "step": 41425
    },
    {
      "entropy": 5.683604621887207,
      "epoch": 4.653001628573033,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003037161261472078,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2267582267522812,
      "num_tokens": 92504541.0,
      "step": 41430
    },
    {
      "entropy": 5.745898771286011,
      "epoch": 4.653563205480991,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030367630529913554,
      "loss": 5.0019,
      "mean_token_accuracy": 0.22934795171022415,
      "num_tokens": 92516911.0,
      "step": 41435
    },
    {
      "entropy": 5.7383111953735355,
      "epoch": 4.654124782388948,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030363648353799857,
      "loss": 4.9705,
      "mean_token_accuracy": 0.22830431610345842,
      "num_tokens": 92526619.0,
      "step": 41440
    },
    {
      "entropy": 5.766182994842529,
      "epoch": 4.654686359296906,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030359666086506493,
      "loss": 4.98,
      "mean_token_accuracy": 0.22945407181978225,
      "num_tokens": 92536579.0,
      "step": 41445
    },
    {
      "entropy": 5.696295738220215,
      "epoch": 4.655247936204863,
      "grad_norm": 1.25,
      "learning_rate": 0.00030355683728160247,
      "loss": 4.8772,
      "mean_token_accuracy": 0.23719919174909593,
      "num_tokens": 92548147.0,
      "step": 41450
    },
    {
      "entropy": 5.707804203033447,
      "epoch": 4.655809513112821,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003035170127888793,
      "loss": 4.9523,
      "mean_token_accuracy": 0.23802599161863328,
      "num_tokens": 92559335.0,
      "step": 41455
    },
    {
      "entropy": 5.683331537246704,
      "epoch": 4.656371090020778,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00030347718738816333,
      "loss": 4.8659,
      "mean_token_accuracy": 0.229104620218277,
      "num_tokens": 92571527.0,
      "step": 41460
    },
    {
      "entropy": 5.72288031578064,
      "epoch": 4.656932666928736,
      "grad_norm": 1.296875,
      "learning_rate": 0.00030343736108072266,
      "loss": 4.9611,
      "mean_token_accuracy": 0.23121015578508378,
      "num_tokens": 92581996.0,
      "step": 41465
    },
    {
      "entropy": 5.68002724647522,
      "epoch": 4.657494243836694,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003033975338678254,
      "loss": 4.8968,
      "mean_token_accuracy": 0.23395519852638244,
      "num_tokens": 92593134.0,
      "step": 41470
    },
    {
      "entropy": 5.655831289291382,
      "epoch": 4.658055820744651,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030335770575073967,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22901405096054078,
      "num_tokens": 92605196.0,
      "step": 41475
    },
    {
      "entropy": 5.663869476318359,
      "epoch": 4.658617397652608,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003033178767307336,
      "loss": 4.9373,
      "mean_token_accuracy": 0.2289673462510109,
      "num_tokens": 92615151.0,
      "step": 41480
    },
    {
      "entropy": 5.72773175239563,
      "epoch": 4.659178974560566,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003032780468090753,
      "loss": 4.972,
      "mean_token_accuracy": 0.22852029651403427,
      "num_tokens": 92625723.0,
      "step": 41485
    },
    {
      "entropy": 5.731389284133911,
      "epoch": 4.6597405514685235,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003032382159870331,
      "loss": 5.0147,
      "mean_token_accuracy": 0.22779725939035417,
      "num_tokens": 92636699.0,
      "step": 41490
    },
    {
      "entropy": 5.686628007888794,
      "epoch": 4.660302128376481,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030319838426587505,
      "loss": 4.964,
      "mean_token_accuracy": 0.22429894655942917,
      "num_tokens": 92646977.0,
      "step": 41495
    },
    {
      "entropy": 5.63306999206543,
      "epoch": 4.660863705284439,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030315855164686956,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2367559403181076,
      "num_tokens": 92658021.0,
      "step": 41500
    },
    {
      "entropy": 5.646017789840698,
      "epoch": 4.661425282192396,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003031187181312848,
      "loss": 4.8388,
      "mean_token_accuracy": 0.23884375989437104,
      "num_tokens": 92668573.0,
      "step": 41505
    },
    {
      "entropy": 5.716219425201416,
      "epoch": 4.661986859100354,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00030307888372038927,
      "loss": 4.9779,
      "mean_token_accuracy": 0.23334106355905532,
      "num_tokens": 92680286.0,
      "step": 41510
    },
    {
      "entropy": 5.6676353931427,
      "epoch": 4.662548436008311,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030303904841545113,
      "loss": 4.9778,
      "mean_token_accuracy": 0.2250085785984993,
      "num_tokens": 92692309.0,
      "step": 41515
    },
    {
      "entropy": 5.622205638885498,
      "epoch": 4.663110012916269,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003029992122177388,
      "loss": 4.8921,
      "mean_token_accuracy": 0.23518387973308563,
      "num_tokens": 92703194.0,
      "step": 41520
    },
    {
      "entropy": 5.813261222839356,
      "epoch": 4.663671589824227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030295937512852075,
      "loss": 5.0516,
      "mean_token_accuracy": 0.22615775018930434,
      "num_tokens": 92714218.0,
      "step": 41525
    },
    {
      "entropy": 5.748193740844727,
      "epoch": 4.664233166732184,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003029195371490652,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2381975919008255,
      "num_tokens": 92726422.0,
      "step": 41530
    },
    {
      "entropy": 5.696509265899659,
      "epoch": 4.664794743640142,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030287969828064087,
      "loss": 4.95,
      "mean_token_accuracy": 0.22665638625621795,
      "num_tokens": 92736835.0,
      "step": 41535
    },
    {
      "entropy": 5.722731494903565,
      "epoch": 4.665356320548099,
      "grad_norm": 1.328125,
      "learning_rate": 0.00030283985852451624,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2370985582470894,
      "num_tokens": 92746665.0,
      "step": 41540
    },
    {
      "entropy": 5.716018342971802,
      "epoch": 4.6659178974560565,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003028000178819596,
      "loss": 4.9374,
      "mean_token_accuracy": 0.22916506677865983,
      "num_tokens": 92758059.0,
      "step": 41545
    },
    {
      "entropy": 5.60120325088501,
      "epoch": 4.666479474364014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030276017635423965,
      "loss": 4.8801,
      "mean_token_accuracy": 0.23701571077108383,
      "num_tokens": 92768982.0,
      "step": 41550
    },
    {
      "entropy": 5.616196393966675,
      "epoch": 4.667041051271972,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00030272033394262483,
      "loss": 4.8218,
      "mean_token_accuracy": 0.2371814429759979,
      "num_tokens": 92779117.0,
      "step": 41555
    },
    {
      "entropy": 5.656359767913818,
      "epoch": 4.667602628179929,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030268049064838395,
      "loss": 4.9354,
      "mean_token_accuracy": 0.2249496027827263,
      "num_tokens": 92790595.0,
      "step": 41560
    },
    {
      "entropy": 5.737684059143066,
      "epoch": 4.668164205087887,
      "grad_norm": 1.28125,
      "learning_rate": 0.00030264064647278537,
      "loss": 4.9627,
      "mean_token_accuracy": 0.2354002222418785,
      "num_tokens": 92801661.0,
      "step": 41565
    },
    {
      "entropy": 5.617568397521973,
      "epoch": 4.668725781995844,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000302600801417098,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23919143974781037,
      "num_tokens": 92813223.0,
      "step": 41570
    },
    {
      "entropy": 5.623089790344238,
      "epoch": 4.6692873589038015,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030256095548259035,
      "loss": 4.9567,
      "mean_token_accuracy": 0.23071215003728868,
      "num_tokens": 92823839.0,
      "step": 41575
    },
    {
      "entropy": 5.5807780742645265,
      "epoch": 4.66984893581176,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003025211086705312,
      "loss": 4.8117,
      "mean_token_accuracy": 0.2336247444152832,
      "num_tokens": 92834739.0,
      "step": 41580
    },
    {
      "entropy": 5.710099411010742,
      "epoch": 4.670410512719717,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003024812609821893,
      "loss": 4.9537,
      "mean_token_accuracy": 0.234686741232872,
      "num_tokens": 92846369.0,
      "step": 41585
    },
    {
      "entropy": 5.632830286026001,
      "epoch": 4.670972089627675,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003024414124188333,
      "loss": 4.9102,
      "mean_token_accuracy": 0.23415900021791458,
      "num_tokens": 92857963.0,
      "step": 41590
    },
    {
      "entropy": 5.676542854309082,
      "epoch": 4.671533666535632,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003024015629817322,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2313639774918556,
      "num_tokens": 92868922.0,
      "step": 41595
    },
    {
      "entropy": 5.674742555618286,
      "epoch": 4.672095243443589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003023617126721547,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22879253774881364,
      "num_tokens": 92881107.0,
      "step": 41600
    },
    {
      "entropy": 5.732442378997803,
      "epoch": 4.6726568203515475,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003023218614913696,
      "loss": 4.9397,
      "mean_token_accuracy": 0.2258987694978714,
      "num_tokens": 92891538.0,
      "step": 41605
    },
    {
      "entropy": 5.683339595794678,
      "epoch": 4.673218397259505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003022820094406458,
      "loss": 4.956,
      "mean_token_accuracy": 0.22382611632347107,
      "num_tokens": 92903015.0,
      "step": 41610
    },
    {
      "entropy": 5.643917608261108,
      "epoch": 4.673779974167462,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003022421565212524,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23225258886814118,
      "num_tokens": 92914237.0,
      "step": 41615
    },
    {
      "entropy": 5.685141658782959,
      "epoch": 4.67434155107542,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030220230273445806,
      "loss": 4.8775,
      "mean_token_accuracy": 0.23368179947137832,
      "num_tokens": 92925351.0,
      "step": 41620
    },
    {
      "entropy": 5.685832357406616,
      "epoch": 4.674903127983377,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030216244808153186,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2367968514561653,
      "num_tokens": 92935472.0,
      "step": 41625
    },
    {
      "entropy": 5.726482820510864,
      "epoch": 4.6754647048913345,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003021225925637428,
      "loss": 5.0312,
      "mean_token_accuracy": 0.218711556494236,
      "num_tokens": 92948328.0,
      "step": 41630
    },
    {
      "entropy": 5.6694262504577635,
      "epoch": 4.676026281799293,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003020827361823598,
      "loss": 4.8979,
      "mean_token_accuracy": 0.2335902377963066,
      "num_tokens": 92958458.0,
      "step": 41635
    },
    {
      "entropy": 5.697475862503052,
      "epoch": 4.67658785870725,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030204287893865204,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2263140857219696,
      "num_tokens": 92969870.0,
      "step": 41640
    },
    {
      "entropy": 5.62992115020752,
      "epoch": 4.677149435615208,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030200302083388856,
      "loss": 4.9146,
      "mean_token_accuracy": 0.23025126308202742,
      "num_tokens": 92981597.0,
      "step": 41645
    },
    {
      "entropy": 5.728918027877808,
      "epoch": 4.677711012523165,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030196316186933834,
      "loss": 4.9725,
      "mean_token_accuracy": 0.23631372302770615,
      "num_tokens": 92992525.0,
      "step": 41650
    },
    {
      "entropy": 5.690793848037719,
      "epoch": 4.678272589431122,
      "grad_norm": 1.203125,
      "learning_rate": 0.00030192330204627054,
      "loss": 4.9422,
      "mean_token_accuracy": 0.230189548432827,
      "num_tokens": 93003652.0,
      "step": 41655
    },
    {
      "entropy": 5.786211156845093,
      "epoch": 4.6788341663390804,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003018834413659545,
      "loss": 5.0251,
      "mean_token_accuracy": 0.22931241244077682,
      "num_tokens": 93015794.0,
      "step": 41660
    },
    {
      "entropy": 5.633232498168946,
      "epoch": 4.679395743247038,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003018435798296592,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2323737010359764,
      "num_tokens": 93026454.0,
      "step": 41665
    },
    {
      "entropy": 5.642833042144775,
      "epoch": 4.679957320154995,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003018037174386539,
      "loss": 4.8922,
      "mean_token_accuracy": 0.22752099633216857,
      "num_tokens": 93039263.0,
      "step": 41670
    },
    {
      "entropy": 5.744030427932739,
      "epoch": 4.680518897062953,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003017638541942078,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2325679451227188,
      "num_tokens": 93050437.0,
      "step": 41675
    },
    {
      "entropy": 5.735906028747559,
      "epoch": 4.68108047397091,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00030172399009759034,
      "loss": 4.9784,
      "mean_token_accuracy": 0.2287391185760498,
      "num_tokens": 93060602.0,
      "step": 41680
    },
    {
      "entropy": 5.719601154327393,
      "epoch": 4.681642050878867,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003016841251500705,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22885636985301971,
      "num_tokens": 93073062.0,
      "step": 41685
    },
    {
      "entropy": 5.699177837371826,
      "epoch": 4.6822036277868255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00030164425935291794,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22614961862564087,
      "num_tokens": 93083150.0,
      "step": 41690
    },
    {
      "entropy": 5.7296223640441895,
      "epoch": 4.682765204694783,
      "grad_norm": 1.125,
      "learning_rate": 0.0003016043927074018,
      "loss": 4.9914,
      "mean_token_accuracy": 0.22581008523702623,
      "num_tokens": 93094506.0,
      "step": 41695
    },
    {
      "entropy": 5.6020440578460695,
      "epoch": 4.683326781602741,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003015645252147914,
      "loss": 4.8158,
      "mean_token_accuracy": 0.23173703998327255,
      "num_tokens": 93105148.0,
      "step": 41700
    },
    {
      "entropy": 5.641522789001465,
      "epoch": 4.683888358510698,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003015246568763562,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22822455912828446,
      "num_tokens": 93116470.0,
      "step": 41705
    },
    {
      "entropy": 5.618566179275513,
      "epoch": 4.684449935418655,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030148478769336564,
      "loss": 4.919,
      "mean_token_accuracy": 0.22771757394075393,
      "num_tokens": 93126662.0,
      "step": 41710
    },
    {
      "entropy": 5.705268812179566,
      "epoch": 4.685011512326613,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003014449176670892,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22824546843767166,
      "num_tokens": 93137690.0,
      "step": 41715
    },
    {
      "entropy": 5.626635408401489,
      "epoch": 4.685573089234571,
      "grad_norm": 1.15625,
      "learning_rate": 0.00030140504679879637,
      "loss": 4.8938,
      "mean_token_accuracy": 0.23273188024759292,
      "num_tokens": 93149760.0,
      "step": 41720
    },
    {
      "entropy": 5.707535696029663,
      "epoch": 4.686134666142529,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003013651750897566,
      "loss": 4.959,
      "mean_token_accuracy": 0.24084142446517945,
      "num_tokens": 93162046.0,
      "step": 41725
    },
    {
      "entropy": 5.721870040893554,
      "epoch": 4.686696243050486,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003013253025412393,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23407733738422393,
      "num_tokens": 93172912.0,
      "step": 41730
    },
    {
      "entropy": 5.711162328720093,
      "epoch": 4.687257819958443,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030128542915451427,
      "loss": 4.9813,
      "mean_token_accuracy": 0.22578435391187668,
      "num_tokens": 93184595.0,
      "step": 41735
    },
    {
      "entropy": 5.660102367401123,
      "epoch": 4.6878193968664,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030124555493085095,
      "loss": 4.8787,
      "mean_token_accuracy": 0.23310376703739166,
      "num_tokens": 93194753.0,
      "step": 41740
    },
    {
      "entropy": 5.695162773132324,
      "epoch": 4.6883809737743585,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003012056798715189,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2334428310394287,
      "num_tokens": 93204259.0,
      "step": 41745
    },
    {
      "entropy": 5.6730297088623045,
      "epoch": 4.688942550682316,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000301165803977788,
      "loss": 4.8687,
      "mean_token_accuracy": 0.24357122629880906,
      "num_tokens": 93214984.0,
      "step": 41750
    },
    {
      "entropy": 5.655963277816772,
      "epoch": 4.689504127590274,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003011259272509277,
      "loss": 4.9849,
      "mean_token_accuracy": 0.22743973433971404,
      "num_tokens": 93227004.0,
      "step": 41755
    },
    {
      "entropy": 5.79460072517395,
      "epoch": 4.690065704498231,
      "grad_norm": 1.125,
      "learning_rate": 0.0003010860496922077,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21910538375377656,
      "num_tokens": 93238485.0,
      "step": 41760
    },
    {
      "entropy": 5.7235125541687015,
      "epoch": 4.690627281406188,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00030104617130289783,
      "loss": 4.9681,
      "mean_token_accuracy": 0.22726578265428543,
      "num_tokens": 93249318.0,
      "step": 41765
    },
    {
      "entropy": 5.731232118606568,
      "epoch": 4.691188858314146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003010062920842677,
      "loss": 4.9653,
      "mean_token_accuracy": 0.23463724553585052,
      "num_tokens": 93262040.0,
      "step": 41770
    },
    {
      "entropy": 5.681745386123657,
      "epoch": 4.6917504352221036,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030096641203758716,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22771921902894973,
      "num_tokens": 93272766.0,
      "step": 41775
    },
    {
      "entropy": 5.793645334243775,
      "epoch": 4.692312012130062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00030092653116412605,
      "loss": 4.9759,
      "mean_token_accuracy": 0.23594192117452623,
      "num_tokens": 93283179.0,
      "step": 41780
    },
    {
      "entropy": 5.7111388683319095,
      "epoch": 4.692873589038019,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000300886649465154,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22692153304815293,
      "num_tokens": 93293484.0,
      "step": 41785
    },
    {
      "entropy": 5.692033433914185,
      "epoch": 4.693435165945976,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003008467669419411,
      "loss": 4.904,
      "mean_token_accuracy": 0.22740947753190993,
      "num_tokens": 93305611.0,
      "step": 41790
    },
    {
      "entropy": 5.707994079589843,
      "epoch": 4.693996742853934,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003008068835957571,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22744762003421784,
      "num_tokens": 93316495.0,
      "step": 41795
    },
    {
      "entropy": 5.655802583694458,
      "epoch": 4.694558319761891,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003007669994278719,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23423291593790055,
      "num_tokens": 93328484.0,
      "step": 41800
    },
    {
      "entropy": 5.705582904815674,
      "epoch": 4.695119896669849,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003007271144395554,
      "loss": 4.9867,
      "mean_token_accuracy": 0.22573504447937012,
      "num_tokens": 93339826.0,
      "step": 41805
    },
    {
      "entropy": 5.68152379989624,
      "epoch": 4.695681473577807,
      "grad_norm": 1.125,
      "learning_rate": 0.0003006872286320777,
      "loss": 4.956,
      "mean_token_accuracy": 0.2375759080052376,
      "num_tokens": 93351923.0,
      "step": 41810
    },
    {
      "entropy": 5.662656688690186,
      "epoch": 4.696243050485764,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003006473420067086,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2325012817978859,
      "num_tokens": 93363813.0,
      "step": 41815
    },
    {
      "entropy": 5.706428289413452,
      "epoch": 4.696804627393721,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003006074545647182,
      "loss": 5.0075,
      "mean_token_accuracy": 0.23173199743032455,
      "num_tokens": 93374905.0,
      "step": 41820
    },
    {
      "entropy": 5.778895616531372,
      "epoch": 4.697366204301679,
      "grad_norm": 1.140625,
      "learning_rate": 0.00030056756630737655,
      "loss": 4.9978,
      "mean_token_accuracy": 0.22848965972661972,
      "num_tokens": 93386473.0,
      "step": 41825
    },
    {
      "entropy": 5.712508964538574,
      "epoch": 4.6979277812096365,
      "grad_norm": 1.25,
      "learning_rate": 0.00030052767723595365,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22939593642950057,
      "num_tokens": 93398010.0,
      "step": 41830
    },
    {
      "entropy": 5.741260194778443,
      "epoch": 4.698489358117595,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003004877873517196,
      "loss": 5.0022,
      "mean_token_accuracy": 0.22213850170373917,
      "num_tokens": 93410692.0,
      "step": 41835
    },
    {
      "entropy": 5.701325273513794,
      "epoch": 4.699050935025552,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00030044789665594444,
      "loss": 4.894,
      "mean_token_accuracy": 0.22530041635036469,
      "num_tokens": 93421830.0,
      "step": 41840
    },
    {
      "entropy": 5.8018042087554935,
      "epoch": 4.699612511933509,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003004080051498985,
      "loss": 5.0813,
      "mean_token_accuracy": 0.2258393481373787,
      "num_tokens": 93434136.0,
      "step": 41845
    },
    {
      "entropy": 5.666636323928833,
      "epoch": 4.700174088841467,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00030036811283485184,
      "loss": 4.8425,
      "mean_token_accuracy": 0.24114398211240767,
      "num_tokens": 93444559.0,
      "step": 41850
    },
    {
      "entropy": 5.728766250610351,
      "epoch": 4.700735665749424,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030032821971207453,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22464921772480012,
      "num_tokens": 93456258.0,
      "step": 41855
    },
    {
      "entropy": 5.722061634063721,
      "epoch": 4.701297242657382,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003002883257828369,
      "loss": 4.9099,
      "mean_token_accuracy": 0.23706090599298477,
      "num_tokens": 93467402.0,
      "step": 41860
    },
    {
      "entropy": 5.690036201477051,
      "epoch": 4.70185881956534,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030024843104840924,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2236340671777725,
      "num_tokens": 93477515.0,
      "step": 41865
    },
    {
      "entropy": 5.720642518997193,
      "epoch": 4.702420396473297,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00030020853551006164,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2263110339641571,
      "num_tokens": 93488173.0,
      "step": 41870
    },
    {
      "entropy": 5.66466178894043,
      "epoch": 4.702981973381254,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003001686391690645,
      "loss": 4.81,
      "mean_token_accuracy": 0.2414587751030922,
      "num_tokens": 93498236.0,
      "step": 41875
    },
    {
      "entropy": 5.653865480422974,
      "epoch": 4.703543550289212,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00030012874202668823,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23566197752952575,
      "num_tokens": 93510162.0,
      "step": 41880
    },
    {
      "entropy": 5.721675968170166,
      "epoch": 4.7041051271971694,
      "grad_norm": 1.1875,
      "learning_rate": 0.00030008884408420293,
      "loss": 4.9664,
      "mean_token_accuracy": 0.23398030400276185,
      "num_tokens": 93521344.0,
      "step": 41885
    },
    {
      "entropy": 5.723297357559204,
      "epoch": 4.7046667041051276,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003000489453428792,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2263870507478714,
      "num_tokens": 93533116.0,
      "step": 41890
    },
    {
      "entropy": 5.7034684181213375,
      "epoch": 4.705228281013085,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003000090458039873,
      "loss": 4.9813,
      "mean_token_accuracy": 0.23012051582336426,
      "num_tokens": 93544612.0,
      "step": 41895
    },
    {
      "entropy": 5.736394691467285,
      "epoch": 4.705789857921042,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029996914546879773,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2284672513604164,
      "num_tokens": 93556033.0,
      "step": 41900
    },
    {
      "entropy": 5.769410276412964,
      "epoch": 4.706351434829,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002999292443385809,
      "loss": 5.062,
      "mean_token_accuracy": 0.22700261622667312,
      "num_tokens": 93567250.0,
      "step": 41905
    },
    {
      "entropy": 5.7501613140106205,
      "epoch": 4.706913011736957,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002998893424146072,
      "loss": 4.9781,
      "mean_token_accuracy": 0.22330669909715653,
      "num_tokens": 93578521.0,
      "step": 41910
    },
    {
      "entropy": 5.742960739135742,
      "epoch": 4.707474588644915,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002998494396981472,
      "loss": 4.9639,
      "mean_token_accuracy": 0.2363390952348709,
      "num_tokens": 93589809.0,
      "step": 41915
    },
    {
      "entropy": 5.626487970352173,
      "epoch": 4.708036165552873,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002998095361904714,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23874028027057648,
      "num_tokens": 93599688.0,
      "step": 41920
    },
    {
      "entropy": 5.624829864501953,
      "epoch": 4.70859774246083,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002997696318928503,
      "loss": 4.9164,
      "mean_token_accuracy": 0.23628917187452317,
      "num_tokens": 93610575.0,
      "step": 41925
    },
    {
      "entropy": 5.650420188903809,
      "epoch": 4.709159319368787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002997297268065546,
      "loss": 4.8837,
      "mean_token_accuracy": 0.2431356281042099,
      "num_tokens": 93622329.0,
      "step": 41930
    },
    {
      "entropy": 5.681970643997192,
      "epoch": 4.709720896276745,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029968982093285475,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23012347221374513,
      "num_tokens": 93632686.0,
      "step": 41935
    },
    {
      "entropy": 5.763223552703858,
      "epoch": 4.710282473184702,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029964991427302147,
      "loss": 4.9582,
      "mean_token_accuracy": 0.23447455912828447,
      "num_tokens": 93643086.0,
      "step": 41940
    },
    {
      "entropy": 5.770403242111206,
      "epoch": 4.7108440500926605,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002996100068283253,
      "loss": 4.9716,
      "mean_token_accuracy": 0.22351535856723787,
      "num_tokens": 93654457.0,
      "step": 41945
    },
    {
      "entropy": 5.74160532951355,
      "epoch": 4.711405627000618,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029957009860003697,
      "loss": 4.9901,
      "mean_token_accuracy": 0.22382545471191406,
      "num_tokens": 93665806.0,
      "step": 41950
    },
    {
      "entropy": 5.653541707992554,
      "epoch": 4.711967203908575,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029953018958942717,
      "loss": 4.9531,
      "mean_token_accuracy": 0.225405552983284,
      "num_tokens": 93676685.0,
      "step": 41955
    },
    {
      "entropy": 5.763412570953369,
      "epoch": 4.712528780816533,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002994902797977667,
      "loss": 5.0535,
      "mean_token_accuracy": 0.22188393026590347,
      "num_tokens": 93688111.0,
      "step": 41960
    },
    {
      "entropy": 5.702380514144897,
      "epoch": 4.71309035772449,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029945036922632615,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22725460082292556,
      "num_tokens": 93698919.0,
      "step": 41965
    },
    {
      "entropy": 5.721646976470947,
      "epoch": 4.713651934632448,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029941045787637636,
      "loss": 4.9584,
      "mean_token_accuracy": 0.2326871484518051,
      "num_tokens": 93711367.0,
      "step": 41970
    },
    {
      "entropy": 5.6945396900177006,
      "epoch": 4.714213511540406,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002993705457491882,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23632178604602813,
      "num_tokens": 93722649.0,
      "step": 41975
    },
    {
      "entropy": 5.71936445236206,
      "epoch": 4.714775088448363,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029933063284603235,
      "loss": 5.0074,
      "mean_token_accuracy": 0.2285478562116623,
      "num_tokens": 93733873.0,
      "step": 41980
    },
    {
      "entropy": 5.649118661880493,
      "epoch": 4.715336665356321,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002992907191681797,
      "loss": 4.7843,
      "mean_token_accuracy": 0.24470624327659607,
      "num_tokens": 93744954.0,
      "step": 41985
    },
    {
      "entropy": 5.690971851348877,
      "epoch": 4.715898242264278,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029925080471690117,
      "loss": 4.908,
      "mean_token_accuracy": 0.2313394069671631,
      "num_tokens": 93756607.0,
      "step": 41990
    },
    {
      "entropy": 5.6451356410980225,
      "epoch": 4.716459819172235,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029921088949346765,
      "loss": 4.8771,
      "mean_token_accuracy": 0.23141492307186126,
      "num_tokens": 93767496.0,
      "step": 41995
    },
    {
      "entropy": 5.628691482543945,
      "epoch": 4.717021396080193,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002991709734991499,
      "loss": 4.8575,
      "mean_token_accuracy": 0.23047405779361724,
      "num_tokens": 93778432.0,
      "step": 42000
    },
    {
      "epoch": 4.717021396080193,
      "eval_entropy": 5.42746700501362,
      "eval_loss": 5.053643703460693,
      "eval_mean_token_accuracy": 0.2319347684569263,
      "eval_num_tokens": 93778432.0,
      "eval_runtime": 18.4067,
      "eval_samples_per_second": 1684.879,
      "eval_steps_per_second": 210.63,
      "step": 42000
    },
    {
      "entropy": 5.657371950149536,
      "epoch": 4.717582972988151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029913105673521907,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2292170360684395,
      "num_tokens": 93789652.0,
      "step": 42005
    },
    {
      "entropy": 5.689206981658936,
      "epoch": 4.718144549896108,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000299091139202946,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22900517880916596,
      "num_tokens": 93800386.0,
      "step": 42010
    },
    {
      "entropy": 5.7428343296051025,
      "epoch": 4.718706126804066,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029905122090360167,
      "loss": 4.9408,
      "mean_token_accuracy": 0.22979079335927963,
      "num_tokens": 93810184.0,
      "step": 42015
    },
    {
      "entropy": 5.690641164779663,
      "epoch": 4.719267703712023,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002990113018384572,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23476769626140595,
      "num_tokens": 93820139.0,
      "step": 42020
    },
    {
      "entropy": 5.699264240264893,
      "epoch": 4.719829280619981,
      "grad_norm": 1.25,
      "learning_rate": 0.0002989713820087836,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23704392462968826,
      "num_tokens": 93831323.0,
      "step": 42025
    },
    {
      "entropy": 5.7300057888031,
      "epoch": 4.7203908575279385,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029893146141585187,
      "loss": 4.9328,
      "mean_token_accuracy": 0.23309300243854522,
      "num_tokens": 93841160.0,
      "step": 42030
    },
    {
      "entropy": 5.749588108062744,
      "epoch": 4.720952434435896,
      "grad_norm": 1.25,
      "learning_rate": 0.00029889154006093306,
      "loss": 5.0522,
      "mean_token_accuracy": 0.22126954644918442,
      "num_tokens": 93851530.0,
      "step": 42035
    },
    {
      "entropy": 5.756308126449585,
      "epoch": 4.721514011343854,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002988516179452984,
      "loss": 4.9711,
      "mean_token_accuracy": 0.235040183365345,
      "num_tokens": 93862466.0,
      "step": 42040
    },
    {
      "entropy": 5.75855131149292,
      "epoch": 4.722075588251811,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029881169507021895,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2444574311375618,
      "num_tokens": 93874399.0,
      "step": 42045
    },
    {
      "entropy": 5.702462387084961,
      "epoch": 4.722637165159768,
      "grad_norm": 1.171875,
      "learning_rate": 0.000298771771436966,
      "loss": 4.9054,
      "mean_token_accuracy": 0.23898811638355255,
      "num_tokens": 93885054.0,
      "step": 42050
    },
    {
      "entropy": 5.697149038314819,
      "epoch": 4.723198742067726,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002987318470468105,
      "loss": 4.8705,
      "mean_token_accuracy": 0.24465490728616715,
      "num_tokens": 93894813.0,
      "step": 42055
    },
    {
      "entropy": 5.61518087387085,
      "epoch": 4.723760318975684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029869192190102385,
      "loss": 4.9145,
      "mean_token_accuracy": 0.23609619438648224,
      "num_tokens": 93905625.0,
      "step": 42060
    },
    {
      "entropy": 5.67480878829956,
      "epoch": 4.724321895883641,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029865199600087724,
      "loss": 4.9124,
      "mean_token_accuracy": 0.23358847498893737,
      "num_tokens": 93917539.0,
      "step": 42065
    },
    {
      "entropy": 5.708007383346557,
      "epoch": 4.724883472791599,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002986120693476418,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22992242574691774,
      "num_tokens": 93927307.0,
      "step": 42070
    },
    {
      "entropy": 5.705095911026001,
      "epoch": 4.725445049699556,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029857214194258896,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23572915494441987,
      "num_tokens": 93938252.0,
      "step": 42075
    },
    {
      "entropy": 5.719238758087158,
      "epoch": 4.726006626607514,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029853221378699,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23509208709001542,
      "num_tokens": 93947762.0,
      "step": 42080
    },
    {
      "entropy": 5.7294165134429935,
      "epoch": 4.7265682035154715,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002984922848821162,
      "loss": 5.0126,
      "mean_token_accuracy": 0.22589306831359862,
      "num_tokens": 93958410.0,
      "step": 42085
    },
    {
      "entropy": 5.697194671630859,
      "epoch": 4.727129780423429,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029845235522923903,
      "loss": 4.9367,
      "mean_token_accuracy": 0.23385144025087357,
      "num_tokens": 93969610.0,
      "step": 42090
    },
    {
      "entropy": 5.71330304145813,
      "epoch": 4.727691357331387,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00029841242482962966,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22920062243938447,
      "num_tokens": 93978404.0,
      "step": 42095
    },
    {
      "entropy": 5.638808107376098,
      "epoch": 4.728252934239344,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00029837249368455963,
      "loss": 4.891,
      "mean_token_accuracy": 0.2285655230283737,
      "num_tokens": 93989578.0,
      "step": 42100
    },
    {
      "entropy": 5.732800006866455,
      "epoch": 4.728814511147302,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002983325617953004,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22585612833499907,
      "num_tokens": 94000010.0,
      "step": 42105
    },
    {
      "entropy": 5.691270351409912,
      "epoch": 4.729376088055259,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029829262916312326,
      "loss": 4.8867,
      "mean_token_accuracy": 0.23963342756032943,
      "num_tokens": 94009386.0,
      "step": 42110
    },
    {
      "entropy": 5.605353689193725,
      "epoch": 4.7299376649632165,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029825269578929986,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2393480896949768,
      "num_tokens": 94021594.0,
      "step": 42115
    },
    {
      "entropy": 5.710771656036377,
      "epoch": 4.730499241871174,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029821276167510153,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22247459441423417,
      "num_tokens": 94033518.0,
      "step": 42120
    },
    {
      "entropy": 5.6567771434783936,
      "epoch": 4.731060818779132,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002981728268217999,
      "loss": 4.931,
      "mean_token_accuracy": 0.22766750752925874,
      "num_tokens": 94044174.0,
      "step": 42125
    },
    {
      "entropy": 5.672130060195923,
      "epoch": 4.731622395687089,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002981328912306665,
      "loss": 4.8803,
      "mean_token_accuracy": 0.23431216478347777,
      "num_tokens": 94056000.0,
      "step": 42130
    },
    {
      "entropy": 5.782611942291259,
      "epoch": 4.732183972595047,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029809295490297285,
      "loss": 5.0244,
      "mean_token_accuracy": 0.22982101440429686,
      "num_tokens": 94067274.0,
      "step": 42135
    },
    {
      "entropy": 5.686967134475708,
      "epoch": 4.732745549503004,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002980530178399906,
      "loss": 4.9006,
      "mean_token_accuracy": 0.22734563648700715,
      "num_tokens": 94079113.0,
      "step": 42140
    },
    {
      "entropy": 5.645139980316162,
      "epoch": 4.733307126410962,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002980130800429912,
      "loss": 4.9286,
      "mean_token_accuracy": 0.2287539318203926,
      "num_tokens": 94090438.0,
      "step": 42145
    },
    {
      "entropy": 5.683363056182861,
      "epoch": 4.73386870331892,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029797314151324653,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22956735491752625,
      "num_tokens": 94101153.0,
      "step": 42150
    },
    {
      "entropy": 5.821297836303711,
      "epoch": 4.734430280226877,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029793320225202807,
      "loss": 5.0728,
      "mean_token_accuracy": 0.22370893210172654,
      "num_tokens": 94112366.0,
      "step": 42155
    },
    {
      "entropy": 5.668909692764283,
      "epoch": 4.734991857134835,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002978932622606075,
      "loss": 4.9378,
      "mean_token_accuracy": 0.23038185983896256,
      "num_tokens": 94122919.0,
      "step": 42160
    },
    {
      "entropy": 5.707004022598267,
      "epoch": 4.735553434042792,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029785332154025676,
      "loss": 5.0181,
      "mean_token_accuracy": 0.23145560622215272,
      "num_tokens": 94135158.0,
      "step": 42165
    },
    {
      "entropy": 5.720366811752319,
      "epoch": 4.7361150109507495,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002978133800922473,
      "loss": 4.9689,
      "mean_token_accuracy": 0.2209291025996208,
      "num_tokens": 94145773.0,
      "step": 42170
    },
    {
      "entropy": 5.626021528244019,
      "epoch": 4.736676587858708,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002977734379178509,
      "loss": 4.8846,
      "mean_token_accuracy": 0.23247865289449693,
      "num_tokens": 94157260.0,
      "step": 42175
    },
    {
      "entropy": 5.6760382652282715,
      "epoch": 4.737238164766665,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002977334950183394,
      "loss": 4.9081,
      "mean_token_accuracy": 0.23003436923027037,
      "num_tokens": 94166513.0,
      "step": 42180
    },
    {
      "entropy": 5.788450574874878,
      "epoch": 4.737799741674622,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029769355139498465,
      "loss": 5.0396,
      "mean_token_accuracy": 0.21188512891530992,
      "num_tokens": 94178565.0,
      "step": 42185
    },
    {
      "entropy": 5.682226181030273,
      "epoch": 4.73836131858258,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002976536070490585,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23241359293460845,
      "num_tokens": 94189319.0,
      "step": 42190
    },
    {
      "entropy": 5.701906013488769,
      "epoch": 4.738922895490537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029761366198183264,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22489549666643144,
      "num_tokens": 94203030.0,
      "step": 42195
    },
    {
      "entropy": 5.726356124877929,
      "epoch": 4.739484472398495,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029757371619457904,
      "loss": 4.9803,
      "mean_token_accuracy": 0.22860509753227234,
      "num_tokens": 94214901.0,
      "step": 42200
    },
    {
      "entropy": 5.691463804244995,
      "epoch": 4.740046049306453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029753376968856955,
      "loss": 4.897,
      "mean_token_accuracy": 0.23456557989120483,
      "num_tokens": 94226833.0,
      "step": 42205
    },
    {
      "entropy": 5.7547272682189945,
      "epoch": 4.74060762621441,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029749382246507604,
      "loss": 4.934,
      "mean_token_accuracy": 0.23283011317253113,
      "num_tokens": 94239464.0,
      "step": 42210
    },
    {
      "entropy": 5.70140814781189,
      "epoch": 4.741169203122368,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029745387452537056,
      "loss": 4.8618,
      "mean_token_accuracy": 0.24274754375219346,
      "num_tokens": 94250692.0,
      "step": 42215
    },
    {
      "entropy": 5.647552871704102,
      "epoch": 4.741730780030325,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029741392587072504,
      "loss": 4.9503,
      "mean_token_accuracy": 0.22650099098682402,
      "num_tokens": 94261130.0,
      "step": 42220
    },
    {
      "entropy": 5.67020058631897,
      "epoch": 4.742292356938282,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029737397650241135,
      "loss": 4.9614,
      "mean_token_accuracy": 0.22675612568855286,
      "num_tokens": 94271581.0,
      "step": 42225
    },
    {
      "entropy": 5.696014022827148,
      "epoch": 4.7428539338462405,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002973340264217016,
      "loss": 4.9521,
      "mean_token_accuracy": 0.23013346195220946,
      "num_tokens": 94283353.0,
      "step": 42230
    },
    {
      "entropy": 5.688822937011719,
      "epoch": 4.743415510754198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029729407562986774,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22591713368892669,
      "num_tokens": 94294669.0,
      "step": 42235
    },
    {
      "entropy": 5.710670852661133,
      "epoch": 4.743977087662155,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029725412412818196,
      "loss": 4.9122,
      "mean_token_accuracy": 0.23530512005090715,
      "num_tokens": 94306127.0,
      "step": 42240
    },
    {
      "entropy": 5.662252807617188,
      "epoch": 4.744538664570113,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00029721417191791617,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23501886278390885,
      "num_tokens": 94316621.0,
      "step": 42245
    },
    {
      "entropy": 5.6915788650512695,
      "epoch": 4.74510024147807,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029717421900034253,
      "loss": 4.9344,
      "mean_token_accuracy": 0.23095409274101258,
      "num_tokens": 94327715.0,
      "step": 42250
    },
    {
      "entropy": 5.704024696350098,
      "epoch": 4.7456618183860275,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002971342653767332,
      "loss": 4.8831,
      "mean_token_accuracy": 0.23326753526926042,
      "num_tokens": 94339452.0,
      "step": 42255
    },
    {
      "entropy": 5.67284460067749,
      "epoch": 4.746223395293986,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029709431104836027,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23793748021125793,
      "num_tokens": 94350309.0,
      "step": 42260
    },
    {
      "entropy": 5.695765256881714,
      "epoch": 4.746784972201943,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029705435601649584,
      "loss": 5.0275,
      "mean_token_accuracy": 0.21884616762399672,
      "num_tokens": 94361665.0,
      "step": 42265
    },
    {
      "entropy": 5.831287431716919,
      "epoch": 4.747346549109901,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002970144002824122,
      "loss": 5.0273,
      "mean_token_accuracy": 0.221736441552639,
      "num_tokens": 94374141.0,
      "step": 42270
    },
    {
      "entropy": 5.667177820205689,
      "epoch": 4.747908126017858,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002969744438473815,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22173160016536714,
      "num_tokens": 94384379.0,
      "step": 42275
    },
    {
      "entropy": 5.668751049041748,
      "epoch": 4.748469702925815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002969344867126761,
      "loss": 4.8504,
      "mean_token_accuracy": 0.239642234146595,
      "num_tokens": 94394747.0,
      "step": 42280
    },
    {
      "entropy": 5.682827520370483,
      "epoch": 4.7490312798337735,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029689452887956804,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2252325803041458,
      "num_tokens": 94407530.0,
      "step": 42285
    },
    {
      "entropy": 5.730138492584229,
      "epoch": 4.749592856741731,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029685457034932966,
      "loss": 4.9821,
      "mean_token_accuracy": 0.2270084798336029,
      "num_tokens": 94420031.0,
      "step": 42290
    },
    {
      "entropy": 5.716902351379394,
      "epoch": 4.750154433649689,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002968146111232333,
      "loss": 4.8832,
      "mean_token_accuracy": 0.2300173446536064,
      "num_tokens": 94429634.0,
      "step": 42295
    },
    {
      "entropy": 5.698299407958984,
      "epoch": 4.750716010557646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029677465120255133,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22600042074918747,
      "num_tokens": 94440310.0,
      "step": 42300
    },
    {
      "entropy": 5.701813793182373,
      "epoch": 4.751277587465603,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002967346905885559,
      "loss": 4.9988,
      "mean_token_accuracy": 0.223689304292202,
      "num_tokens": 94451050.0,
      "step": 42305
    },
    {
      "entropy": 5.725679969787597,
      "epoch": 4.7518391643735605,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029669472928251954,
      "loss": 4.9949,
      "mean_token_accuracy": 0.23378086537122728,
      "num_tokens": 94462210.0,
      "step": 42310
    },
    {
      "entropy": 5.622905540466308,
      "epoch": 4.752400741281519,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029665476728571466,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23181022554636002,
      "num_tokens": 94472716.0,
      "step": 42315
    },
    {
      "entropy": 5.659111547470093,
      "epoch": 4.752962318189476,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002966148045994135,
      "loss": 4.8361,
      "mean_token_accuracy": 0.2344396561384201,
      "num_tokens": 94482446.0,
      "step": 42320
    },
    {
      "entropy": 5.6839683055877686,
      "epoch": 4.753523895097434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002965748412248886,
      "loss": 4.915,
      "mean_token_accuracy": 0.2309856414794922,
      "num_tokens": 94492476.0,
      "step": 42325
    },
    {
      "entropy": 5.645195198059082,
      "epoch": 4.754085472005391,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002965348771634124,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23752198964357377,
      "num_tokens": 94502607.0,
      "step": 42330
    },
    {
      "entropy": 5.632456588745117,
      "epoch": 4.754647048913348,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029649491241625736,
      "loss": 4.8043,
      "mean_token_accuracy": 0.2317063257098198,
      "num_tokens": 94512528.0,
      "step": 42335
    },
    {
      "entropy": 5.876932001113891,
      "epoch": 4.755208625821306,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029645494698469604,
      "loss": 5.1246,
      "mean_token_accuracy": 0.22187190055847167,
      "num_tokens": 94524890.0,
      "step": 42340
    },
    {
      "entropy": 5.6587334156036375,
      "epoch": 4.755770202729264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029641498087000083,
      "loss": 4.8208,
      "mean_token_accuracy": 0.2419935494661331,
      "num_tokens": 94536483.0,
      "step": 42345
    },
    {
      "entropy": 5.735397863388061,
      "epoch": 4.756331779637222,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002963750140734443,
      "loss": 5.0738,
      "mean_token_accuracy": 0.22014571130275726,
      "num_tokens": 94547235.0,
      "step": 42350
    },
    {
      "entropy": 5.764744138717651,
      "epoch": 4.756893356545179,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002963350465962991,
      "loss": 4.9648,
      "mean_token_accuracy": 0.23140071779489518,
      "num_tokens": 94558511.0,
      "step": 42355
    },
    {
      "entropy": 5.666898632049561,
      "epoch": 4.757454933453136,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002962950784398376,
      "loss": 4.9479,
      "mean_token_accuracy": 0.2245735391974449,
      "num_tokens": 94569949.0,
      "step": 42360
    },
    {
      "entropy": 5.700965738296508,
      "epoch": 4.758016510361094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002962551096053327,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2312767446041107,
      "num_tokens": 94581353.0,
      "step": 42365
    },
    {
      "entropy": 5.659055900573731,
      "epoch": 4.7585780872690515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029621514009405675,
      "loss": 4.8287,
      "mean_token_accuracy": 0.24038074761629105,
      "num_tokens": 94591653.0,
      "step": 42370
    },
    {
      "entropy": 5.675115442276001,
      "epoch": 4.759139664177009,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002961751699072826,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22847562283277512,
      "num_tokens": 94603615.0,
      "step": 42375
    },
    {
      "entropy": 5.709891939163208,
      "epoch": 4.759701241084967,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002961351990462827,
      "loss": 5.0188,
      "mean_token_accuracy": 0.2292404890060425,
      "num_tokens": 94614813.0,
      "step": 42380
    },
    {
      "entropy": 5.678616762161255,
      "epoch": 4.760262817992924,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029609522751232996,
      "loss": 4.8309,
      "mean_token_accuracy": 0.23607788383960723,
      "num_tokens": 94624483.0,
      "step": 42385
    },
    {
      "entropy": 5.734954023361206,
      "epoch": 4.760824394900881,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029605525530669705,
      "loss": 4.9478,
      "mean_token_accuracy": 0.2312121257185936,
      "num_tokens": 94634332.0,
      "step": 42390
    },
    {
      "entropy": 5.6505334854125975,
      "epoch": 4.761385971808839,
      "grad_norm": 1.1875,
      "learning_rate": 0.00029601528243065656,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2354821667075157,
      "num_tokens": 94645019.0,
      "step": 42395
    },
    {
      "entropy": 5.735039758682251,
      "epoch": 4.761947548716797,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029597530888548135,
      "loss": 5.0258,
      "mean_token_accuracy": 0.2298056110739708,
      "num_tokens": 94656393.0,
      "step": 42400
    },
    {
      "entropy": 5.692217206954956,
      "epoch": 4.762509125624755,
      "grad_norm": 1.328125,
      "learning_rate": 0.00029593533467244415,
      "loss": 4.889,
      "mean_token_accuracy": 0.236580428481102,
      "num_tokens": 94666730.0,
      "step": 42405
    },
    {
      "entropy": 5.651279497146606,
      "epoch": 4.763070702532712,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002958953597928178,
      "loss": 4.9749,
      "mean_token_accuracy": 0.22850618064403533,
      "num_tokens": 94677937.0,
      "step": 42410
    },
    {
      "entropy": 5.700097131729126,
      "epoch": 4.763632279440669,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002958553842478751,
      "loss": 4.9376,
      "mean_token_accuracy": 0.23015437126159669,
      "num_tokens": 94688141.0,
      "step": 42415
    },
    {
      "entropy": 5.708169317245483,
      "epoch": 4.764193856348627,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029581540803888886,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2315129444003105,
      "num_tokens": 94700055.0,
      "step": 42420
    },
    {
      "entropy": 5.747883653640747,
      "epoch": 4.7647554332565845,
      "grad_norm": 1.125,
      "learning_rate": 0.00029577543116713197,
      "loss": 5.0017,
      "mean_token_accuracy": 0.22509052306413652,
      "num_tokens": 94711858.0,
      "step": 42425
    },
    {
      "entropy": 5.693014097213745,
      "epoch": 4.765317010164542,
      "grad_norm": 1.359375,
      "learning_rate": 0.00029573545363387725,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22698540836572648,
      "num_tokens": 94722691.0,
      "step": 42430
    },
    {
      "entropy": 5.700670480728149,
      "epoch": 4.7658785870725,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002956954754403977,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22859255820512772,
      "num_tokens": 94733570.0,
      "step": 42435
    },
    {
      "entropy": 5.698260259628296,
      "epoch": 4.766440163980457,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002956554965879662,
      "loss": 4.904,
      "mean_token_accuracy": 0.2343561604619026,
      "num_tokens": 94744719.0,
      "step": 42440
    },
    {
      "entropy": 5.7716938018798825,
      "epoch": 4.767001740888414,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002956155170778556,
      "loss": 5.0197,
      "mean_token_accuracy": 0.22426697462797165,
      "num_tokens": 94755272.0,
      "step": 42445
    },
    {
      "entropy": 5.80545129776001,
      "epoch": 4.767563317796372,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002955755369113389,
      "loss": 5.0638,
      "mean_token_accuracy": 0.22555208802223206,
      "num_tokens": 94767178.0,
      "step": 42450
    },
    {
      "entropy": 5.675721883773804,
      "epoch": 4.7681248947043295,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029553555608968925,
      "loss": 4.9211,
      "mean_token_accuracy": 0.2296203151345253,
      "num_tokens": 94777088.0,
      "step": 42455
    },
    {
      "entropy": 5.759205770492554,
      "epoch": 4.768686471612288,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002954955746141795,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22794855535030364,
      "num_tokens": 94787995.0,
      "step": 42460
    },
    {
      "entropy": 5.746955394744873,
      "epoch": 4.769248048520245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002954555924860827,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22805265337228775,
      "num_tokens": 94797810.0,
      "step": 42465
    },
    {
      "entropy": 5.7364420890808105,
      "epoch": 4.769809625428202,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002954156097066718,
      "loss": 4.9355,
      "mean_token_accuracy": 0.23451170921325684,
      "num_tokens": 94809725.0,
      "step": 42470
    },
    {
      "entropy": 5.738846921920777,
      "epoch": 4.77037120233616,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002953756262772201,
      "loss": 4.9279,
      "mean_token_accuracy": 0.2358589306473732,
      "num_tokens": 94819740.0,
      "step": 42475
    },
    {
      "entropy": 5.770650911331177,
      "epoch": 4.770932779244117,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002953356421990005,
      "loss": 5.0201,
      "mean_token_accuracy": 0.22086895555257796,
      "num_tokens": 94830419.0,
      "step": 42480
    },
    {
      "entropy": 5.754526853561401,
      "epoch": 4.7714943561520755,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0002952956574732862,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21826611012220382,
      "num_tokens": 94842796.0,
      "step": 42485
    },
    {
      "entropy": 5.75910587310791,
      "epoch": 4.772055933060033,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002952556721013503,
      "loss": 5.022,
      "mean_token_accuracy": 0.229005129635334,
      "num_tokens": 94854065.0,
      "step": 42490
    },
    {
      "entropy": 5.741798162460327,
      "epoch": 4.77261750996799,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002952156860844659,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2399565264582634,
      "num_tokens": 94866264.0,
      "step": 42495
    },
    {
      "entropy": 5.728284740447998,
      "epoch": 4.773179086875947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002951756994239063,
      "loss": 4.9708,
      "mean_token_accuracy": 0.22495795786380768,
      "num_tokens": 94878549.0,
      "step": 42500
    },
    {
      "entropy": 5.736676931381226,
      "epoch": 4.773740663783905,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029513571212094443,
      "loss": 4.9474,
      "mean_token_accuracy": 0.23008476197719574,
      "num_tokens": 94890527.0,
      "step": 42505
    },
    {
      "entropy": 5.658100175857544,
      "epoch": 4.7743022406918625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002950957241768539,
      "loss": 4.8944,
      "mean_token_accuracy": 0.2258795291185379,
      "num_tokens": 94901424.0,
      "step": 42510
    },
    {
      "entropy": 5.67502269744873,
      "epoch": 4.774863817599821,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029505573559290756,
      "loss": 4.8872,
      "mean_token_accuracy": 0.23349632322788239,
      "num_tokens": 94912153.0,
      "step": 42515
    },
    {
      "entropy": 5.714575242996216,
      "epoch": 4.775425394507778,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029501574637037884,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2321494460105896,
      "num_tokens": 94923912.0,
      "step": 42520
    },
    {
      "entropy": 5.665852499008179,
      "epoch": 4.775986971415735,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000294975756510541,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23633304983377457,
      "num_tokens": 94934768.0,
      "step": 42525
    },
    {
      "entropy": 5.699869632720947,
      "epoch": 4.776548548323693,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002949357660146673,
      "loss": 5.0194,
      "mean_token_accuracy": 0.22003922164440154,
      "num_tokens": 94946617.0,
      "step": 42530
    },
    {
      "entropy": 5.7773560047149655,
      "epoch": 4.77711012523165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002948957748840311,
      "loss": 4.9899,
      "mean_token_accuracy": 0.2174219712615013,
      "num_tokens": 94958863.0,
      "step": 42535
    },
    {
      "entropy": 5.7405397415161135,
      "epoch": 4.7776717021396085,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029485578311990574,
      "loss": 4.9449,
      "mean_token_accuracy": 0.23288127034902573,
      "num_tokens": 94969930.0,
      "step": 42540
    },
    {
      "entropy": 5.713601875305176,
      "epoch": 4.778233279047566,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029481579072356456,
      "loss": 4.9768,
      "mean_token_accuracy": 0.2287659153342247,
      "num_tokens": 94980858.0,
      "step": 42545
    },
    {
      "entropy": 5.834850835800171,
      "epoch": 4.778794855955523,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002947757976962809,
      "loss": 5.0416,
      "mean_token_accuracy": 0.22784137725830078,
      "num_tokens": 94993075.0,
      "step": 42550
    },
    {
      "entropy": 5.676159620285034,
      "epoch": 4.779356432863481,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029473580403932814,
      "loss": 4.9003,
      "mean_token_accuracy": 0.2287295550107956,
      "num_tokens": 95003849.0,
      "step": 42555
    },
    {
      "entropy": 5.703995180130005,
      "epoch": 4.779918009771438,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002946958097539798,
      "loss": 4.9468,
      "mean_token_accuracy": 0.23142216503620147,
      "num_tokens": 95015766.0,
      "step": 42560
    },
    {
      "entropy": 5.673462915420532,
      "epoch": 4.780479586679395,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002946558148415092,
      "loss": 4.9915,
      "mean_token_accuracy": 0.22659453004598618,
      "num_tokens": 95026690.0,
      "step": 42565
    },
    {
      "entropy": 5.660297870635986,
      "epoch": 4.7810411635873535,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002946158193031898,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23907938301563264,
      "num_tokens": 95038591.0,
      "step": 42570
    },
    {
      "entropy": 5.80186357498169,
      "epoch": 4.781602740495311,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029457582314029507,
      "loss": 5.0431,
      "mean_token_accuracy": 0.22071599960327148,
      "num_tokens": 95050268.0,
      "step": 42575
    },
    {
      "entropy": 5.783039188385009,
      "epoch": 4.782164317403268,
      "grad_norm": 1.125,
      "learning_rate": 0.00029453582635409866,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2308189481496811,
      "num_tokens": 95061448.0,
      "step": 42580
    },
    {
      "entropy": 5.734941530227661,
      "epoch": 4.782725894311226,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002944958289458739,
      "loss": 4.9278,
      "mean_token_accuracy": 0.226494599878788,
      "num_tokens": 95072331.0,
      "step": 42585
    },
    {
      "entropy": 5.710793924331665,
      "epoch": 4.783287471219183,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029445583091689437,
      "loss": 4.9487,
      "mean_token_accuracy": 0.23026504814624787,
      "num_tokens": 95082536.0,
      "step": 42590
    },
    {
      "entropy": 5.669659614562988,
      "epoch": 4.783849048127141,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002944158322684336,
      "loss": 4.8752,
      "mean_token_accuracy": 0.24058228731155396,
      "num_tokens": 95093877.0,
      "step": 42595
    },
    {
      "entropy": 5.73660683631897,
      "epoch": 4.784410625035099,
      "grad_norm": 1.25,
      "learning_rate": 0.00029437583300176526,
      "loss": 4.9435,
      "mean_token_accuracy": 0.22847702950239182,
      "num_tokens": 95104672.0,
      "step": 42600
    },
    {
      "entropy": 5.681925964355469,
      "epoch": 4.784972201943056,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029433583311816284,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22294702380895615,
      "num_tokens": 95115206.0,
      "step": 42605
    },
    {
      "entropy": 5.674098873138428,
      "epoch": 4.785533778851014,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002942958326189,
      "loss": 4.8845,
      "mean_token_accuracy": 0.23295631259679794,
      "num_tokens": 95125142.0,
      "step": 42610
    },
    {
      "entropy": 5.706883668899536,
      "epoch": 4.786095355758971,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029425583150525044,
      "loss": 4.9424,
      "mean_token_accuracy": 0.23073548823595047,
      "num_tokens": 95136108.0,
      "step": 42615
    },
    {
      "entropy": 5.663631582260132,
      "epoch": 4.786656932666928,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029421582977848763,
      "loss": 4.9358,
      "mean_token_accuracy": 0.23187037706375122,
      "num_tokens": 95146452.0,
      "step": 42620
    },
    {
      "entropy": 5.577750396728516,
      "epoch": 4.7872185095748865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029417582743988547,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23774691075086593,
      "num_tokens": 95156665.0,
      "step": 42625
    },
    {
      "entropy": 5.707576656341553,
      "epoch": 4.787780086482844,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029413582449071745,
      "loss": 4.9116,
      "mean_token_accuracy": 0.23298300355672835,
      "num_tokens": 95167092.0,
      "step": 42630
    },
    {
      "entropy": 5.70326600074768,
      "epoch": 4.788341663390801,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002940958209322574,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23785788267850877,
      "num_tokens": 95178158.0,
      "step": 42635
    },
    {
      "entropy": 5.661779832839966,
      "epoch": 4.788903240298759,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029405581676577897,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22953829914331436,
      "num_tokens": 95189432.0,
      "step": 42640
    },
    {
      "entropy": 5.6666590690612795,
      "epoch": 4.789464817206716,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000294015811992556,
      "loss": 4.9193,
      "mean_token_accuracy": 0.2320726692676544,
      "num_tokens": 95200682.0,
      "step": 42645
    },
    {
      "entropy": 5.6537463665008545,
      "epoch": 4.790026394114674,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002939758066138621,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23505038917064666,
      "num_tokens": 95212643.0,
      "step": 42650
    },
    {
      "entropy": 5.6894202709198,
      "epoch": 4.790587971022632,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002939358006309712,
      "loss": 4.9375,
      "mean_token_accuracy": 0.234685255587101,
      "num_tokens": 95223098.0,
      "step": 42655
    },
    {
      "entropy": 5.660877418518067,
      "epoch": 4.791149547930589,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002938957940451571,
      "loss": 4.8909,
      "mean_token_accuracy": 0.2314804121851921,
      "num_tokens": 95232946.0,
      "step": 42660
    },
    {
      "entropy": 5.7347719192504885,
      "epoch": 4.791711124838547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029385578685769346,
      "loss": 4.9649,
      "mean_token_accuracy": 0.2311735302209854,
      "num_tokens": 95243966.0,
      "step": 42665
    },
    {
      "entropy": 5.682469749450684,
      "epoch": 4.792272701746504,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002938157790698544,
      "loss": 4.9454,
      "mean_token_accuracy": 0.23627436012029648,
      "num_tokens": 95255251.0,
      "step": 42670
    },
    {
      "entropy": 5.631655550003051,
      "epoch": 4.792834278654462,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002937757706829136,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2399556264281273,
      "num_tokens": 95267517.0,
      "step": 42675
    },
    {
      "entropy": 5.689837503433227,
      "epoch": 4.793395855562419,
      "grad_norm": 1.34375,
      "learning_rate": 0.000293735761698145,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22862776964902878,
      "num_tokens": 95277508.0,
      "step": 42680
    },
    {
      "entropy": 5.6433343410491945,
      "epoch": 4.793957432470377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002936957521168224,
      "loss": 4.9258,
      "mean_token_accuracy": 0.23293453603982925,
      "num_tokens": 95289757.0,
      "step": 42685
    },
    {
      "entropy": 5.746298980712891,
      "epoch": 4.794519009378335,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002936557419402198,
      "loss": 4.9727,
      "mean_token_accuracy": 0.22796741425991057,
      "num_tokens": 95301541.0,
      "step": 42690
    },
    {
      "entropy": 5.652620553970337,
      "epoch": 4.795080586286292,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002936157311696112,
      "loss": 4.8966,
      "mean_token_accuracy": 0.23256882578134536,
      "num_tokens": 95312840.0,
      "step": 42695
    },
    {
      "entropy": 5.663419485092163,
      "epoch": 4.795642163194249,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002935757198062705,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2385842517018318,
      "num_tokens": 95322909.0,
      "step": 42700
    },
    {
      "entropy": 5.553342819213867,
      "epoch": 4.796203740102207,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002935357078514717,
      "loss": 4.781,
      "mean_token_accuracy": 0.23644957989454268,
      "num_tokens": 95333437.0,
      "step": 42705
    },
    {
      "entropy": 5.631757307052612,
      "epoch": 4.7967653170101645,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029349569530648877,
      "loss": 4.9021,
      "mean_token_accuracy": 0.2332710549235344,
      "num_tokens": 95344477.0,
      "step": 42710
    },
    {
      "entropy": 5.735032033920288,
      "epoch": 4.797326893918122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002934556821725957,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2236006498336792,
      "num_tokens": 95354354.0,
      "step": 42715
    },
    {
      "entropy": 5.672476530075073,
      "epoch": 4.79788847082608,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029341566845106647,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22548894137144088,
      "num_tokens": 95365026.0,
      "step": 42720
    },
    {
      "entropy": 5.764074897766113,
      "epoch": 4.798450047734037,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002933756541431753,
      "loss": 4.9947,
      "mean_token_accuracy": 0.23315987288951873,
      "num_tokens": 95376625.0,
      "step": 42725
    },
    {
      "entropy": 5.691479349136353,
      "epoch": 4.799011624641995,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029333563925019616,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22322509735822677,
      "num_tokens": 95387066.0,
      "step": 42730
    },
    {
      "entropy": 5.659113550186158,
      "epoch": 4.799573201549952,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029329562377340327,
      "loss": 4.9161,
      "mean_token_accuracy": 0.22910490483045579,
      "num_tokens": 95398623.0,
      "step": 42735
    },
    {
      "entropy": 5.763869524002075,
      "epoch": 4.80013477845791,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002932556077140704,
      "loss": 5.0113,
      "mean_token_accuracy": 0.22628776729106903,
      "num_tokens": 95412514.0,
      "step": 42740
    },
    {
      "entropy": 5.751785802841186,
      "epoch": 4.800696355365868,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029321559107347206,
      "loss": 4.9667,
      "mean_token_accuracy": 0.22442093044519423,
      "num_tokens": 95422885.0,
      "step": 42745
    },
    {
      "entropy": 5.735594320297241,
      "epoch": 4.801257932273825,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029317557385288216,
      "loss": 4.9527,
      "mean_token_accuracy": 0.22829098701477052,
      "num_tokens": 95433659.0,
      "step": 42750
    },
    {
      "entropy": 5.75530424118042,
      "epoch": 4.801819509181782,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000293135556053575,
      "loss": 4.9685,
      "mean_token_accuracy": 0.22599520683288574,
      "num_tokens": 95444161.0,
      "step": 42755
    },
    {
      "entropy": 5.6854925632476805,
      "epoch": 4.80238108608974,
      "grad_norm": 1.15625,
      "learning_rate": 0.00029309553767682467,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2298245519399643,
      "num_tokens": 95454840.0,
      "step": 42760
    },
    {
      "entropy": 5.731783676147461,
      "epoch": 4.8029426629976975,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002930555187239054,
      "loss": 4.9546,
      "mean_token_accuracy": 0.23266539424657823,
      "num_tokens": 95466766.0,
      "step": 42765
    },
    {
      "entropy": 5.692830610275268,
      "epoch": 4.803504239905655,
      "grad_norm": 1.125,
      "learning_rate": 0.00029301549919609146,
      "loss": 4.9407,
      "mean_token_accuracy": 0.23150017261505126,
      "num_tokens": 95477767.0,
      "step": 42770
    },
    {
      "entropy": 5.676959896087647,
      "epoch": 4.804065816813613,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029297547909465697,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22568629682064056,
      "num_tokens": 95489428.0,
      "step": 42775
    },
    {
      "entropy": 5.753490829467774,
      "epoch": 4.80462739372157,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002929354584208763,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2257692039012909,
      "num_tokens": 95500844.0,
      "step": 42780
    },
    {
      "entropy": 5.685174608230591,
      "epoch": 4.805188970629528,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002928954371760237,
      "loss": 4.8532,
      "mean_token_accuracy": 0.23933292627334596,
      "num_tokens": 95511829.0,
      "step": 42785
    },
    {
      "entropy": 5.751764726638794,
      "epoch": 4.805750547537485,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002928554153613734,
      "loss": 5.0133,
      "mean_token_accuracy": 0.22806515842676162,
      "num_tokens": 95523366.0,
      "step": 42790
    },
    {
      "entropy": 5.607981395721436,
      "epoch": 4.8063121244454425,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002928153929781997,
      "loss": 4.8049,
      "mean_token_accuracy": 0.2386943906545639,
      "num_tokens": 95533495.0,
      "step": 42795
    },
    {
      "entropy": 5.6668284893035885,
      "epoch": 4.806873701353401,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029277537002777703,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23212879300117492,
      "num_tokens": 95545375.0,
      "step": 42800
    },
    {
      "entropy": 5.5910964012146,
      "epoch": 4.807435278261358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029273534651137967,
      "loss": 4.8483,
      "mean_token_accuracy": 0.23249890506267548,
      "num_tokens": 95557212.0,
      "step": 42805
    },
    {
      "entropy": 5.811478853225708,
      "epoch": 4.807996855169315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029269532243028195,
      "loss": 5.0864,
      "mean_token_accuracy": 0.22687375992536546,
      "num_tokens": 95568608.0,
      "step": 42810
    },
    {
      "entropy": 5.607064723968506,
      "epoch": 4.808558432077273,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002926552977857584,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23257657289505004,
      "num_tokens": 95579069.0,
      "step": 42815
    },
    {
      "entropy": 5.7288867950439455,
      "epoch": 4.80912000898523,
      "grad_norm": 1.25,
      "learning_rate": 0.00029261527257908325,
      "loss": 4.9671,
      "mean_token_accuracy": 0.23182038366794586,
      "num_tokens": 95589244.0,
      "step": 42820
    },
    {
      "entropy": 5.66544246673584,
      "epoch": 4.809681585893188,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029257524681153106,
      "loss": 4.9595,
      "mean_token_accuracy": 0.23702029287815093,
      "num_tokens": 95600277.0,
      "step": 42825
    },
    {
      "entropy": 5.743697261810302,
      "epoch": 4.810243162801146,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002925352204843761,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2316950336098671,
      "num_tokens": 95609913.0,
      "step": 42830
    },
    {
      "entropy": 5.875318574905395,
      "epoch": 4.810804739709103,
      "grad_norm": 1.15625,
      "learning_rate": 0.000292495193598893,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2273500829935074,
      "num_tokens": 95623142.0,
      "step": 42835
    },
    {
      "entropy": 5.67499098777771,
      "epoch": 4.811366316617061,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00029245516615635615,
      "loss": 4.9807,
      "mean_token_accuracy": 0.22531724721193314,
      "num_tokens": 95633711.0,
      "step": 42840
    },
    {
      "entropy": 5.67758207321167,
      "epoch": 4.811927893525018,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029241513815804,
      "loss": 4.9106,
      "mean_token_accuracy": 0.2310421958565712,
      "num_tokens": 95646956.0,
      "step": 42845
    },
    {
      "entropy": 5.802404499053955,
      "epoch": 4.8124894704329755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00029237510960521904,
      "loss": 4.9885,
      "mean_token_accuracy": 0.2247886523604393,
      "num_tokens": 95659260.0,
      "step": 42850
    },
    {
      "entropy": 5.80886697769165,
      "epoch": 4.813051047340934,
      "grad_norm": 1.125,
      "learning_rate": 0.000292335080499168,
      "loss": 5.031,
      "mean_token_accuracy": 0.22482990622520446,
      "num_tokens": 95671122.0,
      "step": 42855
    },
    {
      "entropy": 5.668935251235962,
      "epoch": 4.813612624248891,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00029229505084116123,
      "loss": 4.9202,
      "mean_token_accuracy": 0.22708136290311814,
      "num_tokens": 95681513.0,
      "step": 42860
    },
    {
      "entropy": 5.679649305343628,
      "epoch": 4.814174201156849,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002922550206324732,
      "loss": 4.8814,
      "mean_token_accuracy": 0.2378910854458809,
      "num_tokens": 95691670.0,
      "step": 42865
    },
    {
      "entropy": 5.675590467453003,
      "epoch": 4.814735778064806,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002922149898743787,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2318507045507431,
      "num_tokens": 95702359.0,
      "step": 42870
    },
    {
      "entropy": 5.750441455841065,
      "epoch": 4.815297354972763,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002921749585681523,
      "loss": 4.9527,
      "mean_token_accuracy": 0.23730728924274444,
      "num_tokens": 95714069.0,
      "step": 42875
    },
    {
      "entropy": 5.708536100387573,
      "epoch": 4.8158589318807214,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029213492671506855,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22534482628107072,
      "num_tokens": 95724420.0,
      "step": 42880
    },
    {
      "entropy": 5.6870368957519535,
      "epoch": 4.816420508788679,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00029209489431640203,
      "loss": 4.9099,
      "mean_token_accuracy": 0.234396293759346,
      "num_tokens": 95735229.0,
      "step": 42885
    },
    {
      "entropy": 5.613301753997803,
      "epoch": 4.816982085696636,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002920548613734275,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23928287327289582,
      "num_tokens": 95745246.0,
      "step": 42890
    },
    {
      "entropy": 5.654902648925781,
      "epoch": 4.817543662604594,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029201482788741953,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22724445313215255,
      "num_tokens": 95755752.0,
      "step": 42895
    },
    {
      "entropy": 5.732648468017578,
      "epoch": 4.818105239512551,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002919747938596528,
      "loss": 5.0052,
      "mean_token_accuracy": 0.22289814203977584,
      "num_tokens": 95766476.0,
      "step": 42900
    },
    {
      "entropy": 5.78139295578003,
      "epoch": 4.818666816420508,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002919347592914021,
      "loss": 5.0256,
      "mean_token_accuracy": 0.21655148416757583,
      "num_tokens": 95776862.0,
      "step": 42905
    },
    {
      "entropy": 5.713555145263672,
      "epoch": 4.8192283933284665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029189472418394215,
      "loss": 4.9164,
      "mean_token_accuracy": 0.23298409730195999,
      "num_tokens": 95788216.0,
      "step": 42910
    },
    {
      "entropy": 5.710208606719971,
      "epoch": 4.819789970236424,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029185468853854755,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2313855767250061,
      "num_tokens": 95799050.0,
      "step": 42915
    },
    {
      "entropy": 5.740206480026245,
      "epoch": 4.820351547144382,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002918146523564932,
      "loss": 4.9495,
      "mean_token_accuracy": 0.22867397516965865,
      "num_tokens": 95810763.0,
      "step": 42920
    },
    {
      "entropy": 5.736631679534912,
      "epoch": 4.820913124052339,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029177461563905375,
      "loss": 4.9485,
      "mean_token_accuracy": 0.23188409060239792,
      "num_tokens": 95821730.0,
      "step": 42925
    },
    {
      "entropy": 5.7869774341583256,
      "epoch": 4.821474700960296,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029173457838750405,
      "loss": 4.9958,
      "mean_token_accuracy": 0.22502914369106292,
      "num_tokens": 95834037.0,
      "step": 42930
    },
    {
      "entropy": 5.653584909439087,
      "epoch": 4.822036277868254,
      "grad_norm": 1.203125,
      "learning_rate": 0.00029169454060311886,
      "loss": 4.8929,
      "mean_token_accuracy": 0.22954643070697783,
      "num_tokens": 95845074.0,
      "step": 42935
    },
    {
      "entropy": 5.763040781021118,
      "epoch": 4.822597854776212,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000291654502287173,
      "loss": 5.0211,
      "mean_token_accuracy": 0.23096326887607574,
      "num_tokens": 95855439.0,
      "step": 42940
    },
    {
      "entropy": 5.821300363540649,
      "epoch": 4.823159431684169,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002916144634409414,
      "loss": 5.0145,
      "mean_token_accuracy": 0.22902335971593857,
      "num_tokens": 95866511.0,
      "step": 42945
    },
    {
      "entropy": 5.7466119766235355,
      "epoch": 4.823721008592127,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029157442406569885,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22798286825418473,
      "num_tokens": 95877795.0,
      "step": 42950
    },
    {
      "entropy": 5.662178564071655,
      "epoch": 4.824282585500084,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002915343841627201,
      "loss": 4.9451,
      "mean_token_accuracy": 0.2312462881207466,
      "num_tokens": 95889039.0,
      "step": 42955
    },
    {
      "entropy": 5.672784280776978,
      "epoch": 4.824844162408041,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002914943437332802,
      "loss": 4.8671,
      "mean_token_accuracy": 0.23310169130563735,
      "num_tokens": 95900101.0,
      "step": 42960
    },
    {
      "entropy": 5.6786257266998295,
      "epoch": 4.8254057393159995,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029145430277865406,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23166511356830596,
      "num_tokens": 95910023.0,
      "step": 42965
    },
    {
      "entropy": 5.676826333999633,
      "epoch": 4.825967316223957,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002914142613001165,
      "loss": 4.8412,
      "mean_token_accuracy": 0.24017102867364884,
      "num_tokens": 95920610.0,
      "step": 42970
    },
    {
      "entropy": 5.648611545562744,
      "epoch": 4.826528893131915,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029137421929894235,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23345322906970978,
      "num_tokens": 95931521.0,
      "step": 42975
    },
    {
      "entropy": 5.81859941482544,
      "epoch": 4.827090470039872,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002913341767764069,
      "loss": 5.022,
      "mean_token_accuracy": 0.2340457484126091,
      "num_tokens": 95943179.0,
      "step": 42980
    },
    {
      "entropy": 5.781518936157227,
      "epoch": 4.827652046947829,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00029129413373378484,
      "loss": 4.9322,
      "mean_token_accuracy": 0.23529603481292724,
      "num_tokens": 95955282.0,
      "step": 42985
    },
    {
      "entropy": 5.681556653976441,
      "epoch": 4.828213623855787,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029125409017235126,
      "loss": 4.851,
      "mean_token_accuracy": 0.23317316323518752,
      "num_tokens": 95966296.0,
      "step": 42990
    },
    {
      "entropy": 5.743850040435791,
      "epoch": 4.828775200763745,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002912140460933812,
      "loss": 4.9419,
      "mean_token_accuracy": 0.23210230171680452,
      "num_tokens": 95977039.0,
      "step": 42995
    },
    {
      "entropy": 5.765706777572632,
      "epoch": 4.829336777671702,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029117400149814953,
      "loss": 5.0681,
      "mean_token_accuracy": 0.22206531167030336,
      "num_tokens": 95988041.0,
      "step": 43000
    },
    {
      "entropy": 5.674776268005371,
      "epoch": 4.82989835457966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002911339563879315,
      "loss": 4.8589,
      "mean_token_accuracy": 0.24017106145620346,
      "num_tokens": 95998950.0,
      "step": 43005
    },
    {
      "entropy": 5.662245273590088,
      "epoch": 4.830459931487617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002910939107640019,
      "loss": 4.9133,
      "mean_token_accuracy": 0.22803997248411179,
      "num_tokens": 96010380.0,
      "step": 43010
    },
    {
      "entropy": 5.668388605117798,
      "epoch": 4.831021508395574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000291053864627636,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22879705131053923,
      "num_tokens": 96021860.0,
      "step": 43015
    },
    {
      "entropy": 5.779226350784302,
      "epoch": 4.831583085303532,
      "grad_norm": 1.296875,
      "learning_rate": 0.00029101381798010885,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2354763612151146,
      "num_tokens": 96034127.0,
      "step": 43020
    },
    {
      "entropy": 5.711602115631104,
      "epoch": 4.83214466221149,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002909737708226954,
      "loss": 4.9505,
      "mean_token_accuracy": 0.22977754324674607,
      "num_tokens": 96045029.0,
      "step": 43025
    },
    {
      "entropy": 5.705180311203003,
      "epoch": 4.832706239119448,
      "grad_norm": 1.25,
      "learning_rate": 0.00029093372315667105,
      "loss": 4.901,
      "mean_token_accuracy": 0.2336968019604683,
      "num_tokens": 96055409.0,
      "step": 43030
    },
    {
      "entropy": 5.724100160598755,
      "epoch": 4.833267816027405,
      "grad_norm": 1.265625,
      "learning_rate": 0.00029089367498331075,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2297528326511383,
      "num_tokens": 96064928.0,
      "step": 43035
    },
    {
      "entropy": 5.7474298000335695,
      "epoch": 4.833829392935362,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029085362630388965,
      "loss": 5.0164,
      "mean_token_accuracy": 0.23319049328565597,
      "num_tokens": 96077111.0,
      "step": 43040
    },
    {
      "entropy": 5.686125564575195,
      "epoch": 4.83439096984332,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002908135771196829,
      "loss": 4.8452,
      "mean_token_accuracy": 0.24342912584543228,
      "num_tokens": 96088850.0,
      "step": 43045
    },
    {
      "entropy": 5.728006505966187,
      "epoch": 4.8349525467512775,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002907735274319657,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22858661115169526,
      "num_tokens": 96099273.0,
      "step": 43050
    },
    {
      "entropy": 5.720585632324219,
      "epoch": 4.835514123659236,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002907334772420134,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21959613561630248,
      "num_tokens": 96111332.0,
      "step": 43055
    },
    {
      "entropy": 5.745026540756226,
      "epoch": 4.836075700567193,
      "grad_norm": 1.21875,
      "learning_rate": 0.000290693426551101,
      "loss": 4.9836,
      "mean_token_accuracy": 0.2281836286187172,
      "num_tokens": 96122225.0,
      "step": 43060
    },
    {
      "entropy": 5.71902322769165,
      "epoch": 4.83663727747515,
      "grad_norm": 1.28125,
      "learning_rate": 0.00029065337536050384,
      "loss": 4.9218,
      "mean_token_accuracy": 0.2398616537451744,
      "num_tokens": 96133465.0,
      "step": 43065
    },
    {
      "entropy": 5.630079650878907,
      "epoch": 4.837198854383108,
      "grad_norm": 1.234375,
      "learning_rate": 0.00029061332367149705,
      "loss": 4.8597,
      "mean_token_accuracy": 0.23815555423498153,
      "num_tokens": 96145036.0,
      "step": 43070
    },
    {
      "entropy": 5.586356115341187,
      "epoch": 4.837760431291065,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029057327148535614,
      "loss": 4.7888,
      "mean_token_accuracy": 0.24019581079483032,
      "num_tokens": 96154986.0,
      "step": 43075
    },
    {
      "entropy": 5.729847717285156,
      "epoch": 4.838322008199023,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000290533218803356,
      "loss": 5.0197,
      "mean_token_accuracy": 0.23087216168642044,
      "num_tokens": 96166271.0,
      "step": 43080
    },
    {
      "entropy": 5.734456634521484,
      "epoch": 4.838883585106981,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029049316562677235,
      "loss": 5.0031,
      "mean_token_accuracy": 0.2305922195315361,
      "num_tokens": 96177203.0,
      "step": 43085
    },
    {
      "entropy": 5.760515737533569,
      "epoch": 4.839445162014938,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029045311195688016,
      "loss": 4.9759,
      "mean_token_accuracy": 0.22395726293325424,
      "num_tokens": 96188749.0,
      "step": 43090
    },
    {
      "entropy": 5.732279443740845,
      "epoch": 4.840006738922895,
      "grad_norm": 1.265625,
      "learning_rate": 0.000290413057794955,
      "loss": 4.9513,
      "mean_token_accuracy": 0.23526628762483598,
      "num_tokens": 96198701.0,
      "step": 43095
    },
    {
      "entropy": 5.838872337341309,
      "epoch": 4.840568315830853,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00029037300314227205,
      "loss": 5.0921,
      "mean_token_accuracy": 0.22604451179504395,
      "num_tokens": 96211016.0,
      "step": 43100
    },
    {
      "entropy": 5.792295217514038,
      "epoch": 4.8411298927388104,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029033294800010666,
      "loss": 4.9759,
      "mean_token_accuracy": 0.22392407655715943,
      "num_tokens": 96222449.0,
      "step": 43105
    },
    {
      "entropy": 5.64677734375,
      "epoch": 4.8416914696467686,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00029029289236973436,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23113283812999724,
      "num_tokens": 96234434.0,
      "step": 43110
    },
    {
      "entropy": 5.6549108028411865,
      "epoch": 4.842253046554726,
      "grad_norm": 1.171875,
      "learning_rate": 0.00029025283625243043,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22890274077653885,
      "num_tokens": 96245612.0,
      "step": 43115
    },
    {
      "entropy": 5.715095663070679,
      "epoch": 4.842814623462683,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029021277964947023,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23625880628824233,
      "num_tokens": 96256790.0,
      "step": 43120
    },
    {
      "entropy": 5.7976405143737795,
      "epoch": 4.843376200370641,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029017272256212927,
      "loss": 4.9907,
      "mean_token_accuracy": 0.224477419257164,
      "num_tokens": 96267562.0,
      "step": 43125
    },
    {
      "entropy": 5.715603160858154,
      "epoch": 4.843937777278598,
      "grad_norm": 1.25,
      "learning_rate": 0.0002901326649916829,
      "loss": 4.8851,
      "mean_token_accuracy": 0.2342640519142151,
      "num_tokens": 96278469.0,
      "step": 43130
    },
    {
      "entropy": 5.741550064086914,
      "epoch": 4.8444993541865555,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029009260693940665,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2282655879855156,
      "num_tokens": 96289460.0,
      "step": 43135
    },
    {
      "entropy": 5.732986211776733,
      "epoch": 4.845060931094514,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002900525484065759,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22545095533132553,
      "num_tokens": 96300963.0,
      "step": 43140
    },
    {
      "entropy": 5.672202634811401,
      "epoch": 4.845622508002471,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00029001248939446617,
      "loss": 4.8411,
      "mean_token_accuracy": 0.2338864654302597,
      "num_tokens": 96311521.0,
      "step": 43145
    },
    {
      "entropy": 5.709089231491089,
      "epoch": 4.846184084910428,
      "grad_norm": 1.1875,
      "learning_rate": 0.000289972429904353,
      "loss": 4.991,
      "mean_token_accuracy": 0.22604741901159286,
      "num_tokens": 96321959.0,
      "step": 43150
    },
    {
      "entropy": 5.834938049316406,
      "epoch": 4.846745661818386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028993236993751174,
      "loss": 5.1374,
      "mean_token_accuracy": 0.22231039255857468,
      "num_tokens": 96333916.0,
      "step": 43155
    },
    {
      "entropy": 5.775455141067505,
      "epoch": 4.847307238726343,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028989230949521817,
      "loss": 4.9338,
      "mean_token_accuracy": 0.2327294796705246,
      "num_tokens": 96345081.0,
      "step": 43160
    },
    {
      "entropy": 5.666913604736328,
      "epoch": 4.8478688156343015,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028985224857874765,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22426653057336807,
      "num_tokens": 96356394.0,
      "step": 43165
    },
    {
      "entropy": 5.601595258712768,
      "epoch": 4.848430392542259,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028981218718937573,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22954446077346802,
      "num_tokens": 96366952.0,
      "step": 43170
    },
    {
      "entropy": 5.725664567947388,
      "epoch": 4.848991969450216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028977212532837804,
      "loss": 4.9969,
      "mean_token_accuracy": 0.22127334922552108,
      "num_tokens": 96378165.0,
      "step": 43175
    },
    {
      "entropy": 5.84023494720459,
      "epoch": 4.849553546358174,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002897320629970302,
      "loss": 4.9753,
      "mean_token_accuracy": 0.2384838044643402,
      "num_tokens": 96390606.0,
      "step": 43180
    },
    {
      "entropy": 5.671832704544068,
      "epoch": 4.850115123266131,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028969200019660774,
      "loss": 4.8511,
      "mean_token_accuracy": 0.23445437848567963,
      "num_tokens": 96401089.0,
      "step": 43185
    },
    {
      "entropy": 5.689079332351684,
      "epoch": 4.8506767001740885,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002896519369283862,
      "loss": 5.0157,
      "mean_token_accuracy": 0.23061469793319703,
      "num_tokens": 96411908.0,
      "step": 43190
    },
    {
      "entropy": 5.673396110534668,
      "epoch": 4.851238277082047,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002896118731936414,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23842365443706512,
      "num_tokens": 96422859.0,
      "step": 43195
    },
    {
      "entropy": 5.7229790687561035,
      "epoch": 4.851799853990004,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002895718089936489,
      "loss": 4.9358,
      "mean_token_accuracy": 0.23587630242109298,
      "num_tokens": 96435122.0,
      "step": 43200
    },
    {
      "entropy": 5.731399297714233,
      "epoch": 4.852361430897961,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002895317443296843,
      "loss": 4.9821,
      "mean_token_accuracy": 0.22401996105909347,
      "num_tokens": 96446916.0,
      "step": 43205
    },
    {
      "entropy": 5.707595920562744,
      "epoch": 4.852923007805919,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028949167920302337,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22869594246149064,
      "num_tokens": 96458241.0,
      "step": 43210
    },
    {
      "entropy": 5.742794322967529,
      "epoch": 4.853484584713876,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028945161361494177,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23025307506322862,
      "num_tokens": 96468581.0,
      "step": 43215
    },
    {
      "entropy": 5.705128908157349,
      "epoch": 4.854046161621834,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002894115475667151,
      "loss": 4.8991,
      "mean_token_accuracy": 0.23497075736522674,
      "num_tokens": 96479119.0,
      "step": 43220
    },
    {
      "entropy": 5.735240745544433,
      "epoch": 4.854607738529792,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028937148105961924,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22319111675024034,
      "num_tokens": 96490306.0,
      "step": 43225
    },
    {
      "entropy": 5.629713821411133,
      "epoch": 4.855169315437749,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002893314140949298,
      "loss": 4.8949,
      "mean_token_accuracy": 0.2309869721531868,
      "num_tokens": 96501222.0,
      "step": 43230
    },
    {
      "entropy": 5.703954458236694,
      "epoch": 4.855730892345707,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028929134667392265,
      "loss": 5.0276,
      "mean_token_accuracy": 0.23054590076208115,
      "num_tokens": 96513283.0,
      "step": 43235
    },
    {
      "entropy": 5.70406551361084,
      "epoch": 4.856292469253664,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028925127879787343,
      "loss": 4.953,
      "mean_token_accuracy": 0.2285987287759781,
      "num_tokens": 96523546.0,
      "step": 43240
    },
    {
      "entropy": 5.681722164154053,
      "epoch": 4.856854046161622,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002892112104680579,
      "loss": 4.9545,
      "mean_token_accuracy": 0.23286230415105819,
      "num_tokens": 96534906.0,
      "step": 43245
    },
    {
      "entropy": 5.638595771789551,
      "epoch": 4.8574156230695795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028917114168575196,
      "loss": 4.8822,
      "mean_token_accuracy": 0.23154242485761642,
      "num_tokens": 96547029.0,
      "step": 43250
    },
    {
      "entropy": 5.7314423561096195,
      "epoch": 4.857977199977537,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002891310724522314,
      "loss": 4.8558,
      "mean_token_accuracy": 0.24617898166179658,
      "num_tokens": 96558632.0,
      "step": 43255
    },
    {
      "entropy": 5.753760766983032,
      "epoch": 4.858538776885495,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028909100276877203,
      "loss": 4.9519,
      "mean_token_accuracy": 0.236458483338356,
      "num_tokens": 96568644.0,
      "step": 43260
    },
    {
      "entropy": 5.664576864242553,
      "epoch": 4.859100353793452,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028905093263664966,
      "loss": 4.8862,
      "mean_token_accuracy": 0.23467094898223878,
      "num_tokens": 96578844.0,
      "step": 43265
    },
    {
      "entropy": 5.684169769287109,
      "epoch": 4.859661930701409,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028901086205714016,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22539834529161454,
      "num_tokens": 96590715.0,
      "step": 43270
    },
    {
      "entropy": 5.694973373413086,
      "epoch": 4.860223507609367,
      "grad_norm": 1.3125,
      "learning_rate": 0.00028897079103151936,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23168628811836242,
      "num_tokens": 96601001.0,
      "step": 43275
    },
    {
      "entropy": 5.664028120040894,
      "epoch": 4.860785084517325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028893071956106315,
      "loss": 4.8671,
      "mean_token_accuracy": 0.23803477883338928,
      "num_tokens": 96612591.0,
      "step": 43280
    },
    {
      "entropy": 5.63280611038208,
      "epoch": 4.861346661425282,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028889064764704735,
      "loss": 4.928,
      "mean_token_accuracy": 0.229250967502594,
      "num_tokens": 96624353.0,
      "step": 43285
    },
    {
      "entropy": 5.729393863677979,
      "epoch": 4.86190823833324,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028885057529074804,
      "loss": 4.9028,
      "mean_token_accuracy": 0.233727391064167,
      "num_tokens": 96634293.0,
      "step": 43290
    },
    {
      "entropy": 5.7683305740356445,
      "epoch": 4.862469815241197,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028881050249344103,
      "loss": 4.9992,
      "mean_token_accuracy": 0.2251561090350151,
      "num_tokens": 96645308.0,
      "step": 43295
    },
    {
      "entropy": 5.747433996200561,
      "epoch": 4.863031392149155,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002887704292564022,
      "loss": 4.9823,
      "mean_token_accuracy": 0.22717714458703994,
      "num_tokens": 96655198.0,
      "step": 43300
    },
    {
      "entropy": 5.667249298095703,
      "epoch": 4.8635929690571125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028873035558090766,
      "loss": 4.9295,
      "mean_token_accuracy": 0.23074736595153808,
      "num_tokens": 96665753.0,
      "step": 43305
    },
    {
      "entropy": 5.693967723846436,
      "epoch": 4.86415454596507,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002886902814682332,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2302071213722229,
      "num_tokens": 96676038.0,
      "step": 43310
    },
    {
      "entropy": 5.777924823760986,
      "epoch": 4.864716122873028,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028865020691965485,
      "loss": 4.8892,
      "mean_token_accuracy": 0.23009323924779893,
      "num_tokens": 96687893.0,
      "step": 43315
    },
    {
      "entropy": 5.815437126159668,
      "epoch": 4.865277699780985,
      "grad_norm": 1.25,
      "learning_rate": 0.00028861013193644865,
      "loss": 5.0233,
      "mean_token_accuracy": 0.22479260861873626,
      "num_tokens": 96698937.0,
      "step": 43320
    },
    {
      "entropy": 5.69466347694397,
      "epoch": 4.865839276688942,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002885700565198906,
      "loss": 4.9066,
      "mean_token_accuracy": 0.23147106915712357,
      "num_tokens": 96710117.0,
      "step": 43325
    },
    {
      "entropy": 5.635403823852539,
      "epoch": 4.8664008535969,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028852998067125675,
      "loss": 4.869,
      "mean_token_accuracy": 0.23158470988273622,
      "num_tokens": 96719831.0,
      "step": 43330
    },
    {
      "entropy": 5.676929330825805,
      "epoch": 4.8669624305048576,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002884899043918229,
      "loss": 5.0478,
      "mean_token_accuracy": 0.22485671043395997,
      "num_tokens": 96732127.0,
      "step": 43335
    },
    {
      "entropy": 5.753324508666992,
      "epoch": 4.867524007412815,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002884498276828653,
      "loss": 5.0423,
      "mean_token_accuracy": 0.22031626850366592,
      "num_tokens": 96744406.0,
      "step": 43340
    },
    {
      "entropy": 5.743700122833252,
      "epoch": 4.868085584320773,
      "grad_norm": 1.25,
      "learning_rate": 0.00028840975054566007,
      "loss": 4.9275,
      "mean_token_accuracy": 0.23807866275310516,
      "num_tokens": 96754810.0,
      "step": 43345
    },
    {
      "entropy": 5.659544897079468,
      "epoch": 4.86864716122873,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028836967298148317,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23095679730176927,
      "num_tokens": 96765520.0,
      "step": 43350
    },
    {
      "entropy": 5.7645988941192625,
      "epoch": 4.869208738136688,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028832959499161065,
      "loss": 5.0683,
      "mean_token_accuracy": 0.2293346792459488,
      "num_tokens": 96777262.0,
      "step": 43355
    },
    {
      "entropy": 5.718339681625366,
      "epoch": 4.869770315044645,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002882895165773187,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22867709547281265,
      "num_tokens": 96788452.0,
      "step": 43360
    },
    {
      "entropy": 5.691296672821045,
      "epoch": 4.870331891952603,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002882494377398833,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23122954070568086,
      "num_tokens": 96799402.0,
      "step": 43365
    },
    {
      "entropy": 5.722491598129272,
      "epoch": 4.870893468860561,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028820935848058066,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2324262946844101,
      "num_tokens": 96810467.0,
      "step": 43370
    },
    {
      "entropy": 5.681900310516357,
      "epoch": 4.871455045768518,
      "grad_norm": 1.25,
      "learning_rate": 0.000288169278800687,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23601808696985244,
      "num_tokens": 96821480.0,
      "step": 43375
    },
    {
      "entropy": 5.6598801612854,
      "epoch": 4.872016622676475,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028812919870147837,
      "loss": 4.858,
      "mean_token_accuracy": 0.24245477914810182,
      "num_tokens": 96832473.0,
      "step": 43380
    },
    {
      "entropy": 5.719105577468872,
      "epoch": 4.872578199584433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028808911818423096,
      "loss": 4.9252,
      "mean_token_accuracy": 0.22866150885820388,
      "num_tokens": 96844019.0,
      "step": 43385
    },
    {
      "entropy": 5.627691078186035,
      "epoch": 4.8731397764923905,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028804903725022094,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2292925611138344,
      "num_tokens": 96853902.0,
      "step": 43390
    },
    {
      "entropy": 5.702834224700927,
      "epoch": 4.873701353400348,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002880089559007245,
      "loss": 4.956,
      "mean_token_accuracy": 0.23434995710849763,
      "num_tokens": 96865066.0,
      "step": 43395
    },
    {
      "entropy": 5.704900598526001,
      "epoch": 4.874262930308306,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002879688741370179,
      "loss": 5.0212,
      "mean_token_accuracy": 0.22641372978687285,
      "num_tokens": 96876329.0,
      "step": 43400
    },
    {
      "entropy": 5.688631868362426,
      "epoch": 4.874824507216263,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002879287919603773,
      "loss": 4.9004,
      "mean_token_accuracy": 0.2276931270956993,
      "num_tokens": 96887108.0,
      "step": 43405
    },
    {
      "entropy": 5.702048921585083,
      "epoch": 4.875386084124221,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028788870937207896,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2263719081878662,
      "num_tokens": 96898956.0,
      "step": 43410
    },
    {
      "entropy": 5.690140199661255,
      "epoch": 4.875947661032178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002878486263733991,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2282978266477585,
      "num_tokens": 96910889.0,
      "step": 43415
    },
    {
      "entropy": 5.6720153331756595,
      "epoch": 4.876509237940136,
      "grad_norm": 1.125,
      "learning_rate": 0.000287808542965614,
      "loss": 4.86,
      "mean_token_accuracy": 0.23277241140604019,
      "num_tokens": 96921682.0,
      "step": 43420
    },
    {
      "entropy": 5.711611270904541,
      "epoch": 4.877070814848094,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028776845914999996,
      "loss": 4.9077,
      "mean_token_accuracy": 0.23479963690042496,
      "num_tokens": 96933502.0,
      "step": 43425
    },
    {
      "entropy": 5.61089129447937,
      "epoch": 4.877632391756051,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002877283749278332,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2371505856513977,
      "num_tokens": 96945621.0,
      "step": 43430
    },
    {
      "entropy": 5.777002239227295,
      "epoch": 4.878193968664009,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028768829030039,
      "loss": 5.0555,
      "mean_token_accuracy": 0.2255280837416649,
      "num_tokens": 96957164.0,
      "step": 43435
    },
    {
      "entropy": 5.776324129104614,
      "epoch": 4.878755545571966,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028764820526894675,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22915752232074738,
      "num_tokens": 96968646.0,
      "step": 43440
    },
    {
      "entropy": 5.749939918518066,
      "epoch": 4.879317122479923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028760811983477976,
      "loss": 4.901,
      "mean_token_accuracy": 0.2378918081521988,
      "num_tokens": 96979855.0,
      "step": 43445
    },
    {
      "entropy": 5.727704429626465,
      "epoch": 4.8798786993878815,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028756803399916534,
      "loss": 4.9011,
      "mean_token_accuracy": 0.240702848136425,
      "num_tokens": 96989603.0,
      "step": 43450
    },
    {
      "entropy": 5.719157791137695,
      "epoch": 4.880440276295839,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002875279477633798,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23628336191177368,
      "num_tokens": 96999511.0,
      "step": 43455
    },
    {
      "entropy": 5.7408562183380125,
      "epoch": 4.881001853203796,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002874878611286995,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22948821187019347,
      "num_tokens": 97011210.0,
      "step": 43460
    },
    {
      "entropy": 5.7081352233886715,
      "epoch": 4.881563430111754,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000287447774096401,
      "loss": 4.9405,
      "mean_token_accuracy": 0.24139778316020966,
      "num_tokens": 97022351.0,
      "step": 43465
    },
    {
      "entropy": 5.735789346694946,
      "epoch": 4.882125007019711,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028740768666776046,
      "loss": 4.9165,
      "mean_token_accuracy": 0.2331501752138138,
      "num_tokens": 97033269.0,
      "step": 43470
    },
    {
      "entropy": 5.732113409042358,
      "epoch": 4.8826865839276685,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028736759884405433,
      "loss": 4.9607,
      "mean_token_accuracy": 0.23514956831932068,
      "num_tokens": 97045278.0,
      "step": 43475
    },
    {
      "entropy": 5.677454662322998,
      "epoch": 4.883248160835627,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002873275106265591,
      "loss": 4.864,
      "mean_token_accuracy": 0.2430917888879776,
      "num_tokens": 97055617.0,
      "step": 43480
    },
    {
      "entropy": 5.762910079956055,
      "epoch": 4.883809737743584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028728742201655116,
      "loss": 5.0516,
      "mean_token_accuracy": 0.22778767496347427,
      "num_tokens": 97068815.0,
      "step": 43485
    },
    {
      "entropy": 5.722346830368042,
      "epoch": 4.884371314651542,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000287247333015307,
      "loss": 5.0348,
      "mean_token_accuracy": 0.22161055356264114,
      "num_tokens": 97078812.0,
      "step": 43490
    },
    {
      "entropy": 5.715624380111694,
      "epoch": 4.884932891559499,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002872072436241029,
      "loss": 4.8559,
      "mean_token_accuracy": 0.24796293973922728,
      "num_tokens": 97090423.0,
      "step": 43495
    },
    {
      "entropy": 5.666943740844727,
      "epoch": 4.885494468467456,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002871671538442155,
      "loss": 4.7874,
      "mean_token_accuracy": 0.2383962631225586,
      "num_tokens": 97100756.0,
      "step": 43500
    },
    {
      "entropy": 5.622339963912964,
      "epoch": 4.8860560453754145,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002871270636769212,
      "loss": 4.8971,
      "mean_token_accuracy": 0.2369857758283615,
      "num_tokens": 97111016.0,
      "step": 43505
    },
    {
      "entropy": 5.752533721923828,
      "epoch": 4.886617622283372,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002870869731234965,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22778636515140532,
      "num_tokens": 97122148.0,
      "step": 43510
    },
    {
      "entropy": 5.705188322067261,
      "epoch": 4.887179199191329,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002870468821852178,
      "loss": 4.9906,
      "mean_token_accuracy": 0.22654651403427123,
      "num_tokens": 97133301.0,
      "step": 43515
    },
    {
      "entropy": 5.69499397277832,
      "epoch": 4.887740776099287,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028700679086336175,
      "loss": 4.924,
      "mean_token_accuracy": 0.22638589888811111,
      "num_tokens": 97144637.0,
      "step": 43520
    },
    {
      "entropy": 5.678155088424683,
      "epoch": 4.888302353007244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002869666991592048,
      "loss": 4.7651,
      "mean_token_accuracy": 0.24404677450656892,
      "num_tokens": 97156047.0,
      "step": 43525
    },
    {
      "entropy": 5.686189651489258,
      "epoch": 4.8888639299152015,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002869266070740235,
      "loss": 4.9144,
      "mean_token_accuracy": 0.22880120873451232,
      "num_tokens": 97167767.0,
      "step": 43530
    },
    {
      "entropy": 5.5989948272705075,
      "epoch": 4.88942550682316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002868865146090944,
      "loss": 4.8787,
      "mean_token_accuracy": 0.2324773460626602,
      "num_tokens": 97179326.0,
      "step": 43535
    },
    {
      "entropy": 5.698182106018066,
      "epoch": 4.889987083731117,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00028684642176569414,
      "loss": 4.9718,
      "mean_token_accuracy": 0.23113081306219102,
      "num_tokens": 97190437.0,
      "step": 43540
    },
    {
      "entropy": 5.6554662704467775,
      "epoch": 4.890548660639075,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002868063285450992,
      "loss": 4.8353,
      "mean_token_accuracy": 0.23695864230394365,
      "num_tokens": 97200913.0,
      "step": 43545
    },
    {
      "entropy": 5.645918655395508,
      "epoch": 4.891110237547032,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002867662349485861,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2344474658370018,
      "num_tokens": 97210935.0,
      "step": 43550
    },
    {
      "entropy": 5.682251262664795,
      "epoch": 4.891671814454989,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028672614097743156,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23548792749643327,
      "num_tokens": 97221568.0,
      "step": 43555
    },
    {
      "entropy": 5.6821506977081295,
      "epoch": 4.892233391362947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028668604663291213,
      "loss": 4.9122,
      "mean_token_accuracy": 0.23160963356494904,
      "num_tokens": 97232149.0,
      "step": 43560
    },
    {
      "entropy": 5.700012063980102,
      "epoch": 4.892794968270905,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028664595191630455,
      "loss": 4.9731,
      "mean_token_accuracy": 0.23128597140312196,
      "num_tokens": 97242800.0,
      "step": 43565
    },
    {
      "entropy": 5.73804612159729,
      "epoch": 4.893356545178862,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028660585682888524,
      "loss": 4.9573,
      "mean_token_accuracy": 0.23319227993488312,
      "num_tokens": 97253681.0,
      "step": 43570
    },
    {
      "entropy": 5.778824043273926,
      "epoch": 4.89391812208682,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028656576137193096,
      "loss": 4.9635,
      "mean_token_accuracy": 0.22932882457971573,
      "num_tokens": 97264949.0,
      "step": 43575
    },
    {
      "entropy": 5.740573215484619,
      "epoch": 4.894479698994777,
      "grad_norm": 1.25,
      "learning_rate": 0.0002865256655467184,
      "loss": 4.9481,
      "mean_token_accuracy": 0.2388340264558792,
      "num_tokens": 97276670.0,
      "step": 43580
    },
    {
      "entropy": 5.630705642700195,
      "epoch": 4.895041275902734,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002864855693545241,
      "loss": 4.798,
      "mean_token_accuracy": 0.23916922509670258,
      "num_tokens": 97287134.0,
      "step": 43585
    },
    {
      "entropy": 5.678757429122925,
      "epoch": 4.8956028528106925,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028644547279662485,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2332915782928467,
      "num_tokens": 97298250.0,
      "step": 43590
    },
    {
      "entropy": 5.6386558532714846,
      "epoch": 4.89616442971865,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028640537587429734,
      "loss": 4.8642,
      "mean_token_accuracy": 0.22949228733778,
      "num_tokens": 97309144.0,
      "step": 43595
    },
    {
      "entropy": 5.7423101425170895,
      "epoch": 4.896726006626608,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002863652785888182,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22891953140497207,
      "num_tokens": 97320586.0,
      "step": 43600
    },
    {
      "entropy": 5.631533288955689,
      "epoch": 4.897287583534565,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002863251809414642,
      "loss": 4.8478,
      "mean_token_accuracy": 0.2378303200006485,
      "num_tokens": 97330924.0,
      "step": 43605
    },
    {
      "entropy": 5.758518981933594,
      "epoch": 4.897849160442522,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000286285082933512,
      "loss": 5.0008,
      "mean_token_accuracy": 0.22896233350038528,
      "num_tokens": 97341926.0,
      "step": 43610
    },
    {
      "entropy": 5.732301139831543,
      "epoch": 4.89841073735048,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002862449845662385,
      "loss": 4.9943,
      "mean_token_accuracy": 0.23089877218008042,
      "num_tokens": 97353809.0,
      "step": 43615
    },
    {
      "entropy": 5.740403032302856,
      "epoch": 4.898972314258438,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002862048858409201,
      "loss": 4.9721,
      "mean_token_accuracy": 0.23180684447288513,
      "num_tokens": 97365317.0,
      "step": 43620
    },
    {
      "entropy": 5.747049570083618,
      "epoch": 4.899533891166396,
      "grad_norm": 1.390625,
      "learning_rate": 0.000286164786758834,
      "loss": 4.9107,
      "mean_token_accuracy": 0.230870221555233,
      "num_tokens": 97375581.0,
      "step": 43625
    },
    {
      "entropy": 5.856049633026123,
      "epoch": 4.900095468074353,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028612468732125667,
      "loss": 5.0681,
      "mean_token_accuracy": 0.22877597361803054,
      "num_tokens": 97387371.0,
      "step": 43630
    },
    {
      "entropy": 5.704463195800781,
      "epoch": 4.90065704498231,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002860845875294651,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23940492123365403,
      "num_tokens": 97398929.0,
      "step": 43635
    },
    {
      "entropy": 5.828391027450562,
      "epoch": 4.901218621890268,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002860444873847358,
      "loss": 5.0975,
      "mean_token_accuracy": 0.2209021955728531,
      "num_tokens": 97410469.0,
      "step": 43640
    },
    {
      "entropy": 5.692647647857666,
      "epoch": 4.9017801987982255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028600438688834587,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2353803515434265,
      "num_tokens": 97422253.0,
      "step": 43645
    },
    {
      "entropy": 5.713660001754761,
      "epoch": 4.902341775706183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000285964286041572,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22143887728452682,
      "num_tokens": 97433384.0,
      "step": 43650
    },
    {
      "entropy": 5.7177252769470215,
      "epoch": 4.902903352614141,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028592418484569095,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2278228223323822,
      "num_tokens": 97444755.0,
      "step": 43655
    },
    {
      "entropy": 5.61105170249939,
      "epoch": 4.903464929522098,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002858840833019795,
      "loss": 4.7785,
      "mean_token_accuracy": 0.23807331919670105,
      "num_tokens": 97456156.0,
      "step": 43660
    },
    {
      "entropy": 5.680418252944946,
      "epoch": 4.904026506430055,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002858439814117148,
      "loss": 4.8911,
      "mean_token_accuracy": 0.2388056129217148,
      "num_tokens": 97467494.0,
      "step": 43665
    },
    {
      "entropy": 5.644849586486816,
      "epoch": 4.904588083338013,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028580387917617343,
      "loss": 4.8561,
      "mean_token_accuracy": 0.23700400739908217,
      "num_tokens": 97478317.0,
      "step": 43670
    },
    {
      "entropy": 5.71134729385376,
      "epoch": 4.9051496602459705,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002857637765966324,
      "loss": 4.9194,
      "mean_token_accuracy": 0.23580996245145797,
      "num_tokens": 97488530.0,
      "step": 43675
    },
    {
      "entropy": 5.673905849456787,
      "epoch": 4.905711237153929,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002857236736743685,
      "loss": 4.8704,
      "mean_token_accuracy": 0.23843507915735246,
      "num_tokens": 97501148.0,
      "step": 43680
    },
    {
      "entropy": 5.699034357070923,
      "epoch": 4.906272814061886,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002856835704106587,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2326090469956398,
      "num_tokens": 97512136.0,
      "step": 43685
    },
    {
      "entropy": 5.712855672836303,
      "epoch": 4.906834390969843,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00028564346680677995,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23722852468490602,
      "num_tokens": 97523787.0,
      "step": 43690
    },
    {
      "entropy": 5.755889129638672,
      "epoch": 4.907395967877801,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002856033628640089,
      "loss": 4.9458,
      "mean_token_accuracy": 0.229066164791584,
      "num_tokens": 97534851.0,
      "step": 43695
    },
    {
      "entropy": 5.702289724349976,
      "epoch": 4.907957544785758,
      "grad_norm": 1.25,
      "learning_rate": 0.0002855632585836228,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22808694541454316,
      "num_tokens": 97545160.0,
      "step": 43700
    },
    {
      "entropy": 5.708875608444214,
      "epoch": 4.908519121693716,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028552315396689846,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2271127834916115,
      "num_tokens": 97556321.0,
      "step": 43705
    },
    {
      "entropy": 5.705869674682617,
      "epoch": 4.909080698601674,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002854830490151128,
      "loss": 4.9826,
      "mean_token_accuracy": 0.2293954908847809,
      "num_tokens": 97567406.0,
      "step": 43710
    },
    {
      "entropy": 5.700047588348388,
      "epoch": 4.909642275509631,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028544294372954276,
      "loss": 4.8684,
      "mean_token_accuracy": 0.23451010435819625,
      "num_tokens": 97577998.0,
      "step": 43715
    },
    {
      "entropy": 5.667426681518554,
      "epoch": 4.910203852417588,
      "grad_norm": 1.375,
      "learning_rate": 0.0002854028381114654,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2311697334051132,
      "num_tokens": 97588262.0,
      "step": 43720
    },
    {
      "entropy": 5.6741069793701175,
      "epoch": 4.910765429325546,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028536273216215756,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23777104318141937,
      "num_tokens": 97600011.0,
      "step": 43725
    },
    {
      "entropy": 5.803030776977539,
      "epoch": 4.9113270062335035,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028532262588289636,
      "loss": 5.0901,
      "mean_token_accuracy": 0.22933548986911773,
      "num_tokens": 97612037.0,
      "step": 43730
    },
    {
      "entropy": 5.696525764465332,
      "epoch": 4.911888583141462,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028528251927495874,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23910184800624848,
      "num_tokens": 97622643.0,
      "step": 43735
    },
    {
      "entropy": 5.728034591674804,
      "epoch": 4.912450160049419,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028524241233962176,
      "loss": 4.9805,
      "mean_token_accuracy": 0.23034869581460954,
      "num_tokens": 97634720.0,
      "step": 43740
    },
    {
      "entropy": 5.815951871871948,
      "epoch": 4.913011736957376,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028520230507816236,
      "loss": 5.0993,
      "mean_token_accuracy": 0.2203242614865303,
      "num_tokens": 97646314.0,
      "step": 43745
    },
    {
      "entropy": 5.750430727005005,
      "epoch": 4.913573313865334,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028516219749185755,
      "loss": 4.9396,
      "mean_token_accuracy": 0.23027554750442505,
      "num_tokens": 97657478.0,
      "step": 43750
    },
    {
      "entropy": 5.746048927307129,
      "epoch": 4.914134890773291,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002851220895819845,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23202553987503052,
      "num_tokens": 97668084.0,
      "step": 43755
    },
    {
      "entropy": 5.722646331787109,
      "epoch": 4.914696467681249,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028508198134982016,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2279120936989784,
      "num_tokens": 97678769.0,
      "step": 43760
    },
    {
      "entropy": 5.712761163711548,
      "epoch": 4.915258044589207,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028504187279664164,
      "loss": 4.9481,
      "mean_token_accuracy": 0.22585882395505905,
      "num_tokens": 97689587.0,
      "step": 43765
    },
    {
      "entropy": 5.693034934997558,
      "epoch": 4.915819621497164,
      "grad_norm": 1.171875,
      "learning_rate": 0.000285001763923726,
      "loss": 4.9682,
      "mean_token_accuracy": 0.2348466396331787,
      "num_tokens": 97700334.0,
      "step": 43770
    },
    {
      "entropy": 5.664456176757812,
      "epoch": 4.916381198405121,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002849616547323503,
      "loss": 4.867,
      "mean_token_accuracy": 0.2383845940232277,
      "num_tokens": 97713038.0,
      "step": 43775
    },
    {
      "entropy": 5.710511064529419,
      "epoch": 4.916942775313079,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028492154522379167,
      "loss": 4.9542,
      "mean_token_accuracy": 0.2289426952600479,
      "num_tokens": 97724651.0,
      "step": 43780
    },
    {
      "entropy": 5.738123989105224,
      "epoch": 4.917504352221036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002848814353993273,
      "loss": 4.987,
      "mean_token_accuracy": 0.22583379596471786,
      "num_tokens": 97735862.0,
      "step": 43785
    },
    {
      "entropy": 5.73235855102539,
      "epoch": 4.9180659291289945,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028484132526023404,
      "loss": 4.9227,
      "mean_token_accuracy": 0.23144608736038208,
      "num_tokens": 97747356.0,
      "step": 43790
    },
    {
      "entropy": 5.702679634094238,
      "epoch": 4.918627506036952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002848012148077892,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22471805810928344,
      "num_tokens": 97758670.0,
      "step": 43795
    },
    {
      "entropy": 5.751239538192749,
      "epoch": 4.919189082944909,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028476110404326984,
      "loss": 4.9939,
      "mean_token_accuracy": 0.23473086804151536,
      "num_tokens": 97770291.0,
      "step": 43800
    },
    {
      "entropy": 5.693882703781128,
      "epoch": 4.919750659852867,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002847209929679532,
      "loss": 4.9157,
      "mean_token_accuracy": 0.23100600987672806,
      "num_tokens": 97781567.0,
      "step": 43805
    },
    {
      "entropy": 5.635890913009644,
      "epoch": 4.920312236760824,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002846808815831164,
      "loss": 4.882,
      "mean_token_accuracy": 0.2340260237455368,
      "num_tokens": 97792281.0,
      "step": 43810
    },
    {
      "entropy": 5.566014051437378,
      "epoch": 4.920873813668782,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002846407698900365,
      "loss": 4.8008,
      "mean_token_accuracy": 0.23533001095056533,
      "num_tokens": 97803606.0,
      "step": 43815
    },
    {
      "entropy": 5.667036151885986,
      "epoch": 4.92143539057674,
      "grad_norm": 1.125,
      "learning_rate": 0.00028460065788999073,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2322281450033188,
      "num_tokens": 97816024.0,
      "step": 43820
    },
    {
      "entropy": 5.7425848007202145,
      "epoch": 4.921996967484697,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002845605455842564,
      "loss": 4.9027,
      "mean_token_accuracy": 0.23442244082689284,
      "num_tokens": 97827121.0,
      "step": 43825
    },
    {
      "entropy": 5.776212215423584,
      "epoch": 4.922558544392655,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002845204329741106,
      "loss": 4.9276,
      "mean_token_accuracy": 0.23352518826723098,
      "num_tokens": 97838758.0,
      "step": 43830
    },
    {
      "entropy": 5.606924200057984,
      "epoch": 4.923120121300612,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002844803200608304,
      "loss": 4.7777,
      "mean_token_accuracy": 0.24783345460891723,
      "num_tokens": 97850208.0,
      "step": 43835
    },
    {
      "entropy": 5.597445487976074,
      "epoch": 4.923681698208569,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002844402068456931,
      "loss": 4.7857,
      "mean_token_accuracy": 0.23384030014276505,
      "num_tokens": 97860210.0,
      "step": 43840
    },
    {
      "entropy": 5.679818344116211,
      "epoch": 4.9242432751165275,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002844000933299761,
      "loss": 4.8934,
      "mean_token_accuracy": 0.23512939363718033,
      "num_tokens": 97871348.0,
      "step": 43845
    },
    {
      "entropy": 5.735741329193115,
      "epoch": 4.924804852024485,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028435997951495643,
      "loss": 4.926,
      "mean_token_accuracy": 0.23648772835731507,
      "num_tokens": 97882767.0,
      "step": 43850
    },
    {
      "entropy": 5.761954355239868,
      "epoch": 4.925366428932442,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002843198654019114,
      "loss": 4.9766,
      "mean_token_accuracy": 0.23346748799085618,
      "num_tokens": 97895690.0,
      "step": 43855
    },
    {
      "entropy": 5.717254543304444,
      "epoch": 4.9259280058404,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002842797509921182,
      "loss": 4.9386,
      "mean_token_accuracy": 0.23705699145793915,
      "num_tokens": 97906951.0,
      "step": 43860
    },
    {
      "entropy": 5.7366533279418945,
      "epoch": 4.926489582748357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028423963628685414,
      "loss": 4.969,
      "mean_token_accuracy": 0.22738430500030518,
      "num_tokens": 97918538.0,
      "step": 43865
    },
    {
      "entropy": 5.71606216430664,
      "epoch": 4.927051159656315,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002841995212873965,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23890914618968964,
      "num_tokens": 97930175.0,
      "step": 43870
    },
    {
      "entropy": 5.709343099594117,
      "epoch": 4.927612736564273,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002841594059950225,
      "loss": 4.9693,
      "mean_token_accuracy": 0.22927669882774354,
      "num_tokens": 97940870.0,
      "step": 43875
    },
    {
      "entropy": 5.660088205337525,
      "epoch": 4.92817431347223,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028411929041100947,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22977514564990997,
      "num_tokens": 97951492.0,
      "step": 43880
    },
    {
      "entropy": 5.632071018218994,
      "epoch": 4.928735890380188,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002840791745366347,
      "loss": 4.8839,
      "mean_token_accuracy": 0.23729989677667618,
      "num_tokens": 97962243.0,
      "step": 43885
    },
    {
      "entropy": 5.673801898956299,
      "epoch": 4.929297467288145,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028403905837317547,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23854342252016067,
      "num_tokens": 97973150.0,
      "step": 43890
    },
    {
      "entropy": 5.694925260543823,
      "epoch": 4.929859044196102,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028399894192190916,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22524674832820893,
      "num_tokens": 97985314.0,
      "step": 43895
    },
    {
      "entropy": 5.71712818145752,
      "epoch": 4.93042062110406,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028395882518411305,
      "loss": 4.9338,
      "mean_token_accuracy": 0.23483835458755492,
      "num_tokens": 97997065.0,
      "step": 43900
    },
    {
      "entropy": 5.684762763977051,
      "epoch": 4.930982198012018,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002839187081610645,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22837381362915038,
      "num_tokens": 98007576.0,
      "step": 43905
    },
    {
      "entropy": 5.650691890716553,
      "epoch": 4.931543774919975,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002838785908540407,
      "loss": 4.8924,
      "mean_token_accuracy": 0.23175847232341767,
      "num_tokens": 98017870.0,
      "step": 43910
    },
    {
      "entropy": 5.71232123374939,
      "epoch": 4.932105351827933,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00028383847326431924,
      "loss": 5.0033,
      "mean_token_accuracy": 0.2264047756791115,
      "num_tokens": 98029076.0,
      "step": 43915
    },
    {
      "entropy": 5.711356782913208,
      "epoch": 4.93266692873589,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028379835539317727,
      "loss": 4.8641,
      "mean_token_accuracy": 0.23647669702768326,
      "num_tokens": 98042099.0,
      "step": 43920
    },
    {
      "entropy": 5.678464651107788,
      "epoch": 4.933228505643848,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002837582372418923,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23421311676502227,
      "num_tokens": 98053335.0,
      "step": 43925
    },
    {
      "entropy": 5.741113185882568,
      "epoch": 4.9337900825518055,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002837181188117417,
      "loss": 4.9608,
      "mean_token_accuracy": 0.2275925576686859,
      "num_tokens": 98063723.0,
      "step": 43930
    },
    {
      "entropy": 5.682729864120484,
      "epoch": 4.934351659459763,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002836780001040028,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2331816777586937,
      "num_tokens": 98074756.0,
      "step": 43935
    },
    {
      "entropy": 5.690720081329346,
      "epoch": 4.934913236367721,
      "grad_norm": 1.203125,
      "learning_rate": 0.000283637881119953,
      "loss": 4.9203,
      "mean_token_accuracy": 0.23122934848070145,
      "num_tokens": 98086584.0,
      "step": 43940
    },
    {
      "entropy": 5.772793292999268,
      "epoch": 4.935474813275678,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028359776186086965,
      "loss": 5.1069,
      "mean_token_accuracy": 0.22628004848957062,
      "num_tokens": 98099996.0,
      "step": 43945
    },
    {
      "entropy": 5.706452703475952,
      "epoch": 4.936036390183635,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002835576423280303,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2316274508833885,
      "num_tokens": 98111740.0,
      "step": 43950
    },
    {
      "entropy": 5.757088136672974,
      "epoch": 4.936597967091593,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002835175225227123,
      "loss": 4.9003,
      "mean_token_accuracy": 0.23134036362171173,
      "num_tokens": 98123946.0,
      "step": 43955
    },
    {
      "entropy": 5.726846837997437,
      "epoch": 4.937159543999551,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002834774024461931,
      "loss": 4.8977,
      "mean_token_accuracy": 0.23299143016338347,
      "num_tokens": 98134980.0,
      "step": 43960
    },
    {
      "entropy": 5.704193925857544,
      "epoch": 4.937721120907508,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00028343728209975,
      "loss": 4.9716,
      "mean_token_accuracy": 0.23164784014225007,
      "num_tokens": 98145639.0,
      "step": 43965
    },
    {
      "entropy": 5.725542736053467,
      "epoch": 4.938282697815466,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028339716148466067,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2321861505508423,
      "num_tokens": 98156078.0,
      "step": 43970
    },
    {
      "entropy": 5.753729295730591,
      "epoch": 4.938844274723423,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028335704060220246,
      "loss": 5.0457,
      "mean_token_accuracy": 0.22829782515764235,
      "num_tokens": 98168486.0,
      "step": 43975
    },
    {
      "entropy": 5.805467557907105,
      "epoch": 4.939405851631381,
      "grad_norm": 1.28125,
      "learning_rate": 0.00028331691945365274,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22929691821336745,
      "num_tokens": 98178196.0,
      "step": 43980
    },
    {
      "entropy": 5.695907115936279,
      "epoch": 4.9399674285393385,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002832767980402892,
      "loss": 4.9222,
      "mean_token_accuracy": 0.23189273923635484,
      "num_tokens": 98188912.0,
      "step": 43985
    },
    {
      "entropy": 5.7232996940612795,
      "epoch": 4.940529005447296,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002832366763633892,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2252522423863411,
      "num_tokens": 98200207.0,
      "step": 43990
    },
    {
      "entropy": 5.609023857116699,
      "epoch": 4.941090582355254,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028319655442423015,
      "loss": 4.8275,
      "mean_token_accuracy": 0.24133458286523818,
      "num_tokens": 98210846.0,
      "step": 43995
    },
    {
      "entropy": 5.79401593208313,
      "epoch": 4.941652159263211,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028315643222408965,
      "loss": 5.03,
      "mean_token_accuracy": 0.22152153998613358,
      "num_tokens": 98222303.0,
      "step": 44000
    },
    {
      "entropy": 5.661882019042968,
      "epoch": 4.942213736171169,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028311630976424526,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23226557075977325,
      "num_tokens": 98233374.0,
      "step": 44005
    },
    {
      "entropy": 5.7677679538726805,
      "epoch": 4.942775313079126,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002830761870459743,
      "loss": 5.068,
      "mean_token_accuracy": 0.22778857946395875,
      "num_tokens": 98244840.0,
      "step": 44010
    },
    {
      "entropy": 5.644395208358764,
      "epoch": 4.9433368899870835,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002830360640705545,
      "loss": 4.908,
      "mean_token_accuracy": 0.22989758998155593,
      "num_tokens": 98254829.0,
      "step": 44015
    },
    {
      "entropy": 5.7214147567749025,
      "epoch": 4.943898466895042,
      "grad_norm": 1.25,
      "learning_rate": 0.00028299594083926324,
      "loss": 4.9527,
      "mean_token_accuracy": 0.2363619476556778,
      "num_tokens": 98266296.0,
      "step": 44020
    },
    {
      "entropy": 5.691424465179443,
      "epoch": 4.944460043802999,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00028295581735337817,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2407580643892288,
      "num_tokens": 98277723.0,
      "step": 44025
    },
    {
      "entropy": 5.762285947799683,
      "epoch": 4.945021620710956,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028291569361417674,
      "loss": 4.944,
      "mean_token_accuracy": 0.23791351467370986,
      "num_tokens": 98288799.0,
      "step": 44030
    },
    {
      "entropy": 5.6957769870758055,
      "epoch": 4.945583197618914,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002828755696229366,
      "loss": 4.9601,
      "mean_token_accuracy": 0.22775065153837204,
      "num_tokens": 98299731.0,
      "step": 44035
    },
    {
      "entropy": 5.706949090957641,
      "epoch": 4.946144774526871,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002828354453809352,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23647764325141907,
      "num_tokens": 98310607.0,
      "step": 44040
    },
    {
      "entropy": 5.67028431892395,
      "epoch": 4.946706351434829,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002827953208894503,
      "loss": 4.875,
      "mean_token_accuracy": 0.2310766190290451,
      "num_tokens": 98320149.0,
      "step": 44045
    },
    {
      "entropy": 5.708218050003052,
      "epoch": 4.947267928342787,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002827551961497593,
      "loss": 4.8878,
      "mean_token_accuracy": 0.23029569387435914,
      "num_tokens": 98330816.0,
      "step": 44050
    },
    {
      "entropy": 5.636600732803345,
      "epoch": 4.947829505250744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002827150711631398,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23664275407791138,
      "num_tokens": 98342041.0,
      "step": 44055
    },
    {
      "entropy": 5.681701612472534,
      "epoch": 4.948391082158702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028267494593086945,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2239858999848366,
      "num_tokens": 98353891.0,
      "step": 44060
    },
    {
      "entropy": 5.755402898788452,
      "epoch": 4.948952659066659,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028263482045422597,
      "loss": 5.0099,
      "mean_token_accuracy": 0.22630140632390977,
      "num_tokens": 98364931.0,
      "step": 44065
    },
    {
      "entropy": 5.814941024780273,
      "epoch": 4.9495142359746165,
      "grad_norm": 1.125,
      "learning_rate": 0.0002825946947344867,
      "loss": 4.9643,
      "mean_token_accuracy": 0.23390149921178818,
      "num_tokens": 98376787.0,
      "step": 44070
    },
    {
      "entropy": 5.7669768810272215,
      "epoch": 4.950075812882575,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002825545687729295,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22914035320281984,
      "num_tokens": 98387750.0,
      "step": 44075
    },
    {
      "entropy": 5.687379932403564,
      "epoch": 4.950637389790532,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028251444257083186,
      "loss": 4.9066,
      "mean_token_accuracy": 0.23649332970380782,
      "num_tokens": 98399309.0,
      "step": 44080
    },
    {
      "entropy": 5.645137071609497,
      "epoch": 4.951198966698489,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028247431612947146,
      "loss": 4.8409,
      "mean_token_accuracy": 0.230777370929718,
      "num_tokens": 98410850.0,
      "step": 44085
    },
    {
      "entropy": 5.786360168457032,
      "epoch": 4.951760543606447,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002824341894501259,
      "loss": 4.9859,
      "mean_token_accuracy": 0.22663630694150924,
      "num_tokens": 98422166.0,
      "step": 44090
    },
    {
      "entropy": 5.718017721176148,
      "epoch": 4.952322120514404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028239406253407286,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2313748002052307,
      "num_tokens": 98432885.0,
      "step": 44095
    },
    {
      "entropy": 5.741445779800415,
      "epoch": 4.952883697422362,
      "grad_norm": 1.25,
      "learning_rate": 0.00028235393538259007,
      "loss": 4.9439,
      "mean_token_accuracy": 0.22961859703063964,
      "num_tokens": 98444513.0,
      "step": 44100
    },
    {
      "entropy": 5.728229236602783,
      "epoch": 4.95344527433032,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002823138079969551,
      "loss": 4.9406,
      "mean_token_accuracy": 0.23012309968471528,
      "num_tokens": 98456009.0,
      "step": 44105
    },
    {
      "entropy": 5.811790943145752,
      "epoch": 4.954006851238277,
      "grad_norm": 1.203125,
      "learning_rate": 0.00028227368037844557,
      "loss": 5.02,
      "mean_token_accuracy": 0.22515436708927156,
      "num_tokens": 98466282.0,
      "step": 44110
    },
    {
      "entropy": 5.741364765167236,
      "epoch": 4.954568428146235,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028223355252833933,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22883377820253373,
      "num_tokens": 98478334.0,
      "step": 44115
    },
    {
      "entropy": 5.735665988922119,
      "epoch": 4.955130005054192,
      "grad_norm": 1.125,
      "learning_rate": 0.00028219342444791396,
      "loss": 4.9577,
      "mean_token_accuracy": 0.2280513510107994,
      "num_tokens": 98488446.0,
      "step": 44120
    },
    {
      "entropy": 5.6711766719818115,
      "epoch": 4.955691581962149,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028215329613844715,
      "loss": 4.8013,
      "mean_token_accuracy": 0.2403314858675003,
      "num_tokens": 98499698.0,
      "step": 44125
    },
    {
      "entropy": 5.72365984916687,
      "epoch": 4.9562531588701075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028211316760121656,
      "loss": 4.9516,
      "mean_token_accuracy": 0.23373856395483017,
      "num_tokens": 98509980.0,
      "step": 44130
    },
    {
      "entropy": 5.800006771087647,
      "epoch": 4.956814735778065,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002820730388375001,
      "loss": 4.9987,
      "mean_token_accuracy": 0.23562338650226594,
      "num_tokens": 98521406.0,
      "step": 44135
    },
    {
      "entropy": 5.829752588272095,
      "epoch": 4.957376312686022,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002820329098485752,
      "loss": 5.098,
      "mean_token_accuracy": 0.22463662773370743,
      "num_tokens": 98532845.0,
      "step": 44140
    },
    {
      "entropy": 5.661608409881592,
      "epoch": 4.95793788959398,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028199278063571976,
      "loss": 4.8128,
      "mean_token_accuracy": 0.24056552350521088,
      "num_tokens": 98543639.0,
      "step": 44145
    },
    {
      "entropy": 5.74378547668457,
      "epoch": 4.958499466501937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028195265120021157,
      "loss": 4.9256,
      "mean_token_accuracy": 0.23014688938856126,
      "num_tokens": 98554604.0,
      "step": 44150
    },
    {
      "entropy": 5.728053283691406,
      "epoch": 4.9590610434098945,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028191252154332816,
      "loss": 4.9497,
      "mean_token_accuracy": 0.2247973069548607,
      "num_tokens": 98566388.0,
      "step": 44155
    },
    {
      "entropy": 5.755152606964112,
      "epoch": 4.959622620317853,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028187239166634737,
      "loss": 4.997,
      "mean_token_accuracy": 0.23163741677999497,
      "num_tokens": 98577215.0,
      "step": 44160
    },
    {
      "entropy": 5.817703580856323,
      "epoch": 4.96018419722581,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028183226157054705,
      "loss": 5.1199,
      "mean_token_accuracy": 0.2218694806098938,
      "num_tokens": 98588640.0,
      "step": 44165
    },
    {
      "entropy": 5.743334007263184,
      "epoch": 4.960745774133768,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002817921312572048,
      "loss": 4.9088,
      "mean_token_accuracy": 0.23451436012983323,
      "num_tokens": 98599710.0,
      "step": 44170
    },
    {
      "entropy": 5.674370002746582,
      "epoch": 4.961307351041725,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002817520007275985,
      "loss": 4.999,
      "mean_token_accuracy": 0.2284787729382515,
      "num_tokens": 98611510.0,
      "step": 44175
    },
    {
      "entropy": 5.682364845275879,
      "epoch": 4.961868927949682,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002817118699830058,
      "loss": 4.902,
      "mean_token_accuracy": 0.2316669911146164,
      "num_tokens": 98622982.0,
      "step": 44180
    },
    {
      "entropy": 5.827095794677734,
      "epoch": 4.9624305048576405,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002816717390247046,
      "loss": 5.0721,
      "mean_token_accuracy": 0.22875319570302963,
      "num_tokens": 98634216.0,
      "step": 44185
    },
    {
      "entropy": 5.814006567001343,
      "epoch": 4.962992081765598,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002816316078539727,
      "loss": 5.0533,
      "mean_token_accuracy": 0.22657645046710967,
      "num_tokens": 98646469.0,
      "step": 44190
    },
    {
      "entropy": 5.781308174133301,
      "epoch": 4.963553658673556,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028159147647208784,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2212061896920204,
      "num_tokens": 98659249.0,
      "step": 44195
    },
    {
      "entropy": 5.697348785400391,
      "epoch": 4.964115235581513,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002815513448803277,
      "loss": 4.8568,
      "mean_token_accuracy": 0.23880195170640944,
      "num_tokens": 98671322.0,
      "step": 44200
    },
    {
      "entropy": 5.75070972442627,
      "epoch": 4.96467681248947,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002815112130799703,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2356574520468712,
      "num_tokens": 98684236.0,
      "step": 44205
    },
    {
      "entropy": 5.668997192382813,
      "epoch": 4.965238389397428,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00028147108107229324,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22971357852220536,
      "num_tokens": 98694581.0,
      "step": 44210
    },
    {
      "entropy": 5.74655556678772,
      "epoch": 4.965799966305386,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028143094885857455,
      "loss": 4.9674,
      "mean_token_accuracy": 0.22348420023918153,
      "num_tokens": 98705002.0,
      "step": 44215
    },
    {
      "entropy": 5.734104061126709,
      "epoch": 4.966361543213343,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000281390816440092,
      "loss": 4.9327,
      "mean_token_accuracy": 0.23105983138084413,
      "num_tokens": 98715131.0,
      "step": 44220
    },
    {
      "entropy": 5.633748769760132,
      "epoch": 4.966923120121301,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002813506838181233,
      "loss": 4.8425,
      "mean_token_accuracy": 0.23502167463302612,
      "num_tokens": 98724992.0,
      "step": 44225
    },
    {
      "entropy": 5.679039382934571,
      "epoch": 4.967484697029258,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002813105509939464,
      "loss": 4.9163,
      "mean_token_accuracy": 0.2308916985988617,
      "num_tokens": 98736651.0,
      "step": 44230
    },
    {
      "entropy": 5.7590007305145265,
      "epoch": 4.968046273937215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000281270417968839,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23259158432483673,
      "num_tokens": 98747365.0,
      "step": 44235
    },
    {
      "entropy": 5.763219976425171,
      "epoch": 4.968607850845173,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002812302847440791,
      "loss": 5.0241,
      "mean_token_accuracy": 0.2222556620836258,
      "num_tokens": 98759610.0,
      "step": 44240
    },
    {
      "entropy": 5.724096965789795,
      "epoch": 4.969169427753131,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028119015132094455,
      "loss": 4.9419,
      "mean_token_accuracy": 0.23520705550909043,
      "num_tokens": 98770166.0,
      "step": 44245
    },
    {
      "entropy": 5.70335202217102,
      "epoch": 4.969731004661089,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002811500177007132,
      "loss": 4.954,
      "mean_token_accuracy": 0.2337472468614578,
      "num_tokens": 98782416.0,
      "step": 44250
    },
    {
      "entropy": 5.751177215576172,
      "epoch": 4.970292581569046,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028110988388466287,
      "loss": 4.984,
      "mean_token_accuracy": 0.22825003564357757,
      "num_tokens": 98793341.0,
      "step": 44255
    },
    {
      "entropy": 5.652658605575562,
      "epoch": 4.970854158477003,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002810697498740715,
      "loss": 4.8615,
      "mean_token_accuracy": 0.23722876161336898,
      "num_tokens": 98803642.0,
      "step": 44260
    },
    {
      "entropy": 5.684006834030152,
      "epoch": 4.971415735384961,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002810296156702169,
      "loss": 4.8655,
      "mean_token_accuracy": 0.23540905863046646,
      "num_tokens": 98814650.0,
      "step": 44265
    },
    {
      "entropy": 5.737416172027588,
      "epoch": 4.9719773122929185,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028098948127437695,
      "loss": 4.9044,
      "mean_token_accuracy": 0.22629882842302323,
      "num_tokens": 98825928.0,
      "step": 44270
    },
    {
      "entropy": 5.747047424316406,
      "epoch": 4.972538889200876,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002809493466878297,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2273867130279541,
      "num_tokens": 98837447.0,
      "step": 44275
    },
    {
      "entropy": 5.682340431213379,
      "epoch": 4.973100466108834,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028090921191185295,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2290763661265373,
      "num_tokens": 98847720.0,
      "step": 44280
    },
    {
      "entropy": 5.627958536148071,
      "epoch": 4.973662043016791,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002808690769477245,
      "loss": 4.8544,
      "mean_token_accuracy": 0.24254194945096968,
      "num_tokens": 98859315.0,
      "step": 44285
    },
    {
      "entropy": 5.639615726470947,
      "epoch": 4.974223619924748,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002808289417967225,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22900413125753402,
      "num_tokens": 98869921.0,
      "step": 44290
    },
    {
      "entropy": 5.680972290039063,
      "epoch": 4.974785196832706,
      "grad_norm": 1.15625,
      "learning_rate": 0.00028078880646012467,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2411016568541527,
      "num_tokens": 98881473.0,
      "step": 44295
    },
    {
      "entropy": 5.692086839675904,
      "epoch": 4.975346773740664,
      "grad_norm": 1.234375,
      "learning_rate": 0.00028074867093920897,
      "loss": 4.8856,
      "mean_token_accuracy": 0.23472434878349305,
      "num_tokens": 98892601.0,
      "step": 44300
    },
    {
      "entropy": 5.738140630722046,
      "epoch": 4.975908350648622,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028070853523525337,
      "loss": 4.994,
      "mean_token_accuracy": 0.22656564563512802,
      "num_tokens": 98903823.0,
      "step": 44305
    },
    {
      "entropy": 5.684837388992309,
      "epoch": 4.976469927556579,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00028066839934953573,
      "loss": 4.9539,
      "mean_token_accuracy": 0.22779441773891448,
      "num_tokens": 98914266.0,
      "step": 44310
    },
    {
      "entropy": 5.703308629989624,
      "epoch": 4.977031504464536,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028062826328333415,
      "loss": 4.9037,
      "mean_token_accuracy": 0.2279985725879669,
      "num_tokens": 98924693.0,
      "step": 44315
    },
    {
      "entropy": 5.700107192993164,
      "epoch": 4.977593081372494,
      "grad_norm": 1.125,
      "learning_rate": 0.0002805881270379265,
      "loss": 4.9457,
      "mean_token_accuracy": 0.228223080933094,
      "num_tokens": 98936081.0,
      "step": 44320
    },
    {
      "entropy": 5.716934728622436,
      "epoch": 4.9781546582804515,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002805479906145906,
      "loss": 4.916,
      "mean_token_accuracy": 0.23119878619909287,
      "num_tokens": 98946495.0,
      "step": 44325
    },
    {
      "entropy": 5.695006370544434,
      "epoch": 4.978716235188409,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002805078540146046,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22794557362794876,
      "num_tokens": 98956873.0,
      "step": 44330
    },
    {
      "entropy": 5.634523868560791,
      "epoch": 4.979277812096367,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002804677172392464,
      "loss": 4.8963,
      "mean_token_accuracy": 0.23352068662643433,
      "num_tokens": 98968349.0,
      "step": 44335
    },
    {
      "entropy": 5.673640203475952,
      "epoch": 4.979839389004324,
      "grad_norm": 1.1875,
      "learning_rate": 0.000280427580289794,
      "loss": 4.9013,
      "mean_token_accuracy": 0.23908883035182954,
      "num_tokens": 98978558.0,
      "step": 44340
    },
    {
      "entropy": 5.784474229812622,
      "epoch": 4.980400965912281,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028038744316752524,
      "loss": 5.0524,
      "mean_token_accuracy": 0.2283596932888031,
      "num_tokens": 98990435.0,
      "step": 44345
    },
    {
      "entropy": 5.754787015914917,
      "epoch": 4.980962542820239,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028034730587371824,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2310713142156601,
      "num_tokens": 99001516.0,
      "step": 44350
    },
    {
      "entropy": 5.717418241500854,
      "epoch": 4.9815241197281965,
      "grad_norm": 1.265625,
      "learning_rate": 0.00028030716840965096,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2289329618215561,
      "num_tokens": 99012810.0,
      "step": 44355
    },
    {
      "entropy": 5.7677136898040775,
      "epoch": 4.982085696636155,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00028026703077660143,
      "loss": 5.0298,
      "mean_token_accuracy": 0.22215814292430877,
      "num_tokens": 99023342.0,
      "step": 44360
    },
    {
      "entropy": 5.65390419960022,
      "epoch": 4.982647273544112,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002802268929758475,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2382773905992508,
      "num_tokens": 99034047.0,
      "step": 44365
    },
    {
      "entropy": 5.769520950317383,
      "epoch": 4.983208850452069,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002801867550086674,
      "loss": 4.997,
      "mean_token_accuracy": 0.22633339762687682,
      "num_tokens": 99045333.0,
      "step": 44370
    },
    {
      "entropy": 5.736665487289429,
      "epoch": 4.983770427360027,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002801466168763389,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2332674890756607,
      "num_tokens": 99057582.0,
      "step": 44375
    },
    {
      "entropy": 5.726194953918457,
      "epoch": 4.984332004267984,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002801064785801401,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22925459891557692,
      "num_tokens": 99067940.0,
      "step": 44380
    },
    {
      "entropy": 5.742639827728271,
      "epoch": 4.9848935811759425,
      "grad_norm": 1.171875,
      "learning_rate": 0.00028006634012134915,
      "loss": 4.9932,
      "mean_token_accuracy": 0.2253882959485054,
      "num_tokens": 99079408.0,
      "step": 44385
    },
    {
      "entropy": 5.666633796691895,
      "epoch": 4.9854551580839,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00028002620150124386,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23073238432407378,
      "num_tokens": 99090320.0,
      "step": 44390
    },
    {
      "entropy": 5.722715377807617,
      "epoch": 4.986016734991857,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027998606272110237,
      "loss": 4.9867,
      "mean_token_accuracy": 0.22519256174564362,
      "num_tokens": 99101503.0,
      "step": 44395
    },
    {
      "entropy": 5.806135082244873,
      "epoch": 4.986578311899815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002799459237822028,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2279892936348915,
      "num_tokens": 99112768.0,
      "step": 44400
    },
    {
      "entropy": 5.723201131820678,
      "epoch": 4.987139888807772,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027990578468582304,
      "loss": 4.9195,
      "mean_token_accuracy": 0.23499235808849334,
      "num_tokens": 99123473.0,
      "step": 44405
    },
    {
      "entropy": 5.751761484146118,
      "epoch": 4.9877014657157295,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002798656454332412,
      "loss": 4.9769,
      "mean_token_accuracy": 0.23011718541383744,
      "num_tokens": 99136005.0,
      "step": 44410
    },
    {
      "entropy": 5.783205986022949,
      "epoch": 4.988263042623688,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027982550602573524,
      "loss": 5.0117,
      "mean_token_accuracy": 0.23601753860712052,
      "num_tokens": 99147494.0,
      "step": 44415
    },
    {
      "entropy": 5.789635229110718,
      "epoch": 4.988824619531645,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002797853664645834,
      "loss": 5.0069,
      "mean_token_accuracy": 0.22212600857019424,
      "num_tokens": 99159255.0,
      "step": 44420
    },
    {
      "entropy": 5.749743700027466,
      "epoch": 4.989386196439602,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002797452267510637,
      "loss": 4.9223,
      "mean_token_accuracy": 0.23519034534692765,
      "num_tokens": 99169406.0,
      "step": 44425
    },
    {
      "entropy": 5.6885637760162355,
      "epoch": 4.98994777334756,
      "grad_norm": 1.203125,
      "learning_rate": 0.000279705086886454,
      "loss": 4.9259,
      "mean_token_accuracy": 0.23922231495380403,
      "num_tokens": 99180392.0,
      "step": 44430
    },
    {
      "entropy": 5.661696290969848,
      "epoch": 4.990509350255517,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002796649468720326,
      "loss": 4.8557,
      "mean_token_accuracy": 0.2346661329269409,
      "num_tokens": 99190387.0,
      "step": 44435
    },
    {
      "entropy": 5.719713926315308,
      "epoch": 4.9910709271634754,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027962480670907743,
      "loss": 4.9535,
      "mean_token_accuracy": 0.23255554735660552,
      "num_tokens": 99202070.0,
      "step": 44440
    },
    {
      "entropy": 5.774237823486328,
      "epoch": 4.991632504071433,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027958466639886663,
      "loss": 4.9924,
      "mean_token_accuracy": 0.23233315646648406,
      "num_tokens": 99213323.0,
      "step": 44445
    },
    {
      "entropy": 5.723667621612549,
      "epoch": 4.99219408097939,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027954452594267825,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22969242036342621,
      "num_tokens": 99224178.0,
      "step": 44450
    },
    {
      "entropy": 5.808000946044922,
      "epoch": 4.992755657887348,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00027950438534179046,
      "loss": 5.0108,
      "mean_token_accuracy": 0.22954239845275878,
      "num_tokens": 99234538.0,
      "step": 44455
    },
    {
      "entropy": 5.774686431884765,
      "epoch": 4.993317234795305,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027946424459748126,
      "loss": 4.9396,
      "mean_token_accuracy": 0.2259129300713539,
      "num_tokens": 99245737.0,
      "step": 44460
    },
    {
      "entropy": 5.71666464805603,
      "epoch": 4.993878811703262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027942410371102877,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22915612310171127,
      "num_tokens": 99257047.0,
      "step": 44465
    },
    {
      "entropy": 5.7701808452606205,
      "epoch": 4.9944403886112205,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002793839626837111,
      "loss": 5.0405,
      "mean_token_accuracy": 0.22578528821468352,
      "num_tokens": 99267664.0,
      "step": 44470
    },
    {
      "entropy": 5.663380718231201,
      "epoch": 4.995001965519178,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027934382151680644,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2252355471253395,
      "num_tokens": 99279270.0,
      "step": 44475
    },
    {
      "entropy": 5.711811971664429,
      "epoch": 4.995563542427135,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027930368021159273,
      "loss": 4.8888,
      "mean_token_accuracy": 0.23252922892570496,
      "num_tokens": 99289724.0,
      "step": 44480
    },
    {
      "entropy": 5.796668910980225,
      "epoch": 4.996125119335093,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002792635387693481,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2304462492465973,
      "num_tokens": 99300326.0,
      "step": 44485
    },
    {
      "entropy": 5.721752023696899,
      "epoch": 4.99668669624305,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027922339719135093,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23608564734458923,
      "num_tokens": 99312073.0,
      "step": 44490
    },
    {
      "entropy": 5.758451795578003,
      "epoch": 4.997248273151008,
      "grad_norm": 1.203125,
      "learning_rate": 0.000279183255478879,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2285988673567772,
      "num_tokens": 99323938.0,
      "step": 44495
    },
    {
      "entropy": 5.699099016189575,
      "epoch": 4.997809850058966,
      "grad_norm": 1.34375,
      "learning_rate": 0.00027914311363321073,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22939572781324385,
      "num_tokens": 99334773.0,
      "step": 44500
    },
    {
      "entropy": 5.711266756057739,
      "epoch": 4.998371426966923,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027910297165562406,
      "loss": 4.895,
      "mean_token_accuracy": 0.2421182245016098,
      "num_tokens": 99345688.0,
      "step": 44505
    },
    {
      "entropy": 5.747595119476318,
      "epoch": 4.998933003874881,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027906282954739714,
      "loss": 4.951,
      "mean_token_accuracy": 0.23122645765542985,
      "num_tokens": 99356307.0,
      "step": 44510
    },
    {
      "entropy": 5.674584150314331,
      "epoch": 4.999494580782838,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027902268730980816,
      "loss": 4.8437,
      "mean_token_accuracy": 0.23866217583417892,
      "num_tokens": 99367348.0,
      "step": 44515
    },
    {
      "entropy": 5.722233825259739,
      "epoch": 5.0,
      "grad_norm": 6.40625,
      "learning_rate": 0.0002789825449441353,
      "loss": 4.8769,
      "mean_token_accuracy": 0.23736775086985695,
      "num_tokens": 99377525.0,
      "step": 44520
    },
    {
      "entropy": 5.703025531768799,
      "epoch": 5.000561576907957,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002789424024516566,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23351562172174453,
      "num_tokens": 99388342.0,
      "step": 44525
    },
    {
      "entropy": 5.740617084503174,
      "epoch": 5.001123153815915,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027890225983365034,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2339566484093666,
      "num_tokens": 99400387.0,
      "step": 44530
    },
    {
      "entropy": 5.624771976470948,
      "epoch": 5.0016847307238725,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002788621170913946,
      "loss": 4.8157,
      "mean_token_accuracy": 0.23535728305578232,
      "num_tokens": 99411232.0,
      "step": 44535
    },
    {
      "entropy": 5.6578531742095945,
      "epoch": 5.00224630763183,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027882197422616746,
      "loss": 4.8894,
      "mean_token_accuracy": 0.235831218957901,
      "num_tokens": 99422839.0,
      "step": 44540
    },
    {
      "entropy": 5.753374290466309,
      "epoch": 5.002807884539788,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002787818312392473,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2361813947558403,
      "num_tokens": 99434835.0,
      "step": 44545
    },
    {
      "entropy": 5.695776271820068,
      "epoch": 5.003369461447745,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027874168813191207,
      "loss": 4.8914,
      "mean_token_accuracy": 0.2337503120303154,
      "num_tokens": 99445705.0,
      "step": 44550
    },
    {
      "entropy": 5.668764877319336,
      "epoch": 5.003931038355703,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002787015449054401,
      "loss": 4.8773,
      "mean_token_accuracy": 0.23847148716449737,
      "num_tokens": 99457127.0,
      "step": 44555
    },
    {
      "entropy": 5.653587436676025,
      "epoch": 5.00449261526366,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002786614015611095,
      "loss": 4.845,
      "mean_token_accuracy": 0.23615225702524184,
      "num_tokens": 99468653.0,
      "step": 44560
    },
    {
      "entropy": 5.811617422103882,
      "epoch": 5.005054192171618,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027862125810019844,
      "loss": 5.0735,
      "mean_token_accuracy": 0.22080178558826447,
      "num_tokens": 99481238.0,
      "step": 44565
    },
    {
      "entropy": 5.727873945236206,
      "epoch": 5.005615769079576,
      "grad_norm": 1.25,
      "learning_rate": 0.00027858111452398515,
      "loss": 4.8523,
      "mean_token_accuracy": 0.24338110238313676,
      "num_tokens": 99493025.0,
      "step": 44570
    },
    {
      "entropy": 5.685417318344117,
      "epoch": 5.006177345987533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027854097083374773,
      "loss": 4.9223,
      "mean_token_accuracy": 0.23792690634727479,
      "num_tokens": 99504675.0,
      "step": 44575
    },
    {
      "entropy": 5.747564506530762,
      "epoch": 5.00673892289549,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002785008270307645,
      "loss": 4.8841,
      "mean_token_accuracy": 0.23506583273410797,
      "num_tokens": 99516005.0,
      "step": 44580
    },
    {
      "entropy": 5.80488657951355,
      "epoch": 5.007300499803448,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002784606831163136,
      "loss": 5.0356,
      "mean_token_accuracy": 0.22303812801837922,
      "num_tokens": 99526917.0,
      "step": 44585
    },
    {
      "entropy": 5.644045782089234,
      "epoch": 5.0078620767114055,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027842053909167314,
      "loss": 4.871,
      "mean_token_accuracy": 0.23139696717262268,
      "num_tokens": 99538799.0,
      "step": 44590
    },
    {
      "entropy": 5.719938850402832,
      "epoch": 5.008423653619364,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027838039495812133,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2390056237578392,
      "num_tokens": 99550431.0,
      "step": 44595
    },
    {
      "entropy": 5.686437606811523,
      "epoch": 5.008985230527321,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027834025071693655,
      "loss": 4.829,
      "mean_token_accuracy": 0.23713678270578384,
      "num_tokens": 99561562.0,
      "step": 44600
    },
    {
      "entropy": 5.668091344833374,
      "epoch": 5.009546807435278,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027830010636939684,
      "loss": 4.8192,
      "mean_token_accuracy": 0.2402091383934021,
      "num_tokens": 99572101.0,
      "step": 44605
    },
    {
      "entropy": 5.693507432937622,
      "epoch": 5.010108384343236,
      "grad_norm": 1.25,
      "learning_rate": 0.00027825996191678043,
      "loss": 4.9003,
      "mean_token_accuracy": 0.23402196913957596,
      "num_tokens": 99583186.0,
      "step": 44610
    },
    {
      "entropy": 5.6779289722442625,
      "epoch": 5.010669961251193,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027821981736036555,
      "loss": 4.8775,
      "mean_token_accuracy": 0.24008159935474396,
      "num_tokens": 99593716.0,
      "step": 44615
    },
    {
      "entropy": 5.782568168640137,
      "epoch": 5.011231538159151,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002781796727014305,
      "loss": 4.9494,
      "mean_token_accuracy": 0.2269889533519745,
      "num_tokens": 99605798.0,
      "step": 44620
    },
    {
      "entropy": 5.731050586700439,
      "epoch": 5.011793115067109,
      "grad_norm": 1.125,
      "learning_rate": 0.0002781395279412535,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23399524986743928,
      "num_tokens": 99617938.0,
      "step": 44625
    },
    {
      "entropy": 5.716056489944458,
      "epoch": 5.012354691975066,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027809938308111254,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2355725184082985,
      "num_tokens": 99628761.0,
      "step": 44630
    },
    {
      "entropy": 5.687629175186157,
      "epoch": 5.012916268883023,
      "grad_norm": 1.25,
      "learning_rate": 0.00027805923812228615,
      "loss": 4.9317,
      "mean_token_accuracy": 0.2307425245642662,
      "num_tokens": 99640072.0,
      "step": 44635
    },
    {
      "entropy": 5.710982322692871,
      "epoch": 5.013477845790981,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027801909306605243,
      "loss": 4.9644,
      "mean_token_accuracy": 0.2345776841044426,
      "num_tokens": 99651280.0,
      "step": 44640
    },
    {
      "entropy": 5.6979457378387455,
      "epoch": 5.014039422698938,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002779789479136896,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2346952259540558,
      "num_tokens": 99661574.0,
      "step": 44645
    },
    {
      "entropy": 5.69944167137146,
      "epoch": 5.0146009996068965,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002779388026664759,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2332244336605072,
      "num_tokens": 99671092.0,
      "step": 44650
    },
    {
      "entropy": 5.8422037124633786,
      "epoch": 5.015162576514854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002778986573256896,
      "loss": 5.0173,
      "mean_token_accuracy": 0.22878357470035554,
      "num_tokens": 99682104.0,
      "step": 44655
    },
    {
      "entropy": 5.79989070892334,
      "epoch": 5.015724153422811,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002778585118926089,
      "loss": 4.9564,
      "mean_token_accuracy": 0.23301382064819337,
      "num_tokens": 99692152.0,
      "step": 44660
    },
    {
      "entropy": 5.674339246749878,
      "epoch": 5.016285730330769,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027781836636851207,
      "loss": 4.8616,
      "mean_token_accuracy": 0.23397193402051925,
      "num_tokens": 99703807.0,
      "step": 44665
    },
    {
      "entropy": 5.781098937988281,
      "epoch": 5.016847307238726,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027777822075467735,
      "loss": 4.9507,
      "mean_token_accuracy": 0.23692330867052078,
      "num_tokens": 99714039.0,
      "step": 44670
    },
    {
      "entropy": 5.664341592788697,
      "epoch": 5.0174088841466835,
      "grad_norm": 1.1875,
      "learning_rate": 0.000277738075052383,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2338688090443611,
      "num_tokens": 99725215.0,
      "step": 44675
    },
    {
      "entropy": 5.704753875732422,
      "epoch": 5.017970461054642,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002776979292629074,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23994003236293793,
      "num_tokens": 99736281.0,
      "step": 44680
    },
    {
      "entropy": 5.752308559417725,
      "epoch": 5.018532037962599,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002776577833875285,
      "loss": 4.9097,
      "mean_token_accuracy": 0.2347152829170227,
      "num_tokens": 99747544.0,
      "step": 44685
    },
    {
      "entropy": 5.657304430007935,
      "epoch": 5.019093614870557,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027761763742752485,
      "loss": 4.8358,
      "mean_token_accuracy": 0.2363751158118248,
      "num_tokens": 99758806.0,
      "step": 44690
    },
    {
      "entropy": 5.7066686153411865,
      "epoch": 5.019655191778514,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002775774913841746,
      "loss": 4.8641,
      "mean_token_accuracy": 0.23606125861406327,
      "num_tokens": 99769477.0,
      "step": 44695
    },
    {
      "entropy": 5.671803092956543,
      "epoch": 5.020216768686471,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000277537345258756,
      "loss": 4.8826,
      "mean_token_accuracy": 0.23387900590896607,
      "num_tokens": 99780283.0,
      "step": 44700
    },
    {
      "entropy": 5.793545722961426,
      "epoch": 5.0207783455944295,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027749719905254737,
      "loss": 4.9598,
      "mean_token_accuracy": 0.23231204003095626,
      "num_tokens": 99790119.0,
      "step": 44705
    },
    {
      "entropy": 5.77143816947937,
      "epoch": 5.021339922502387,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000277457052766827,
      "loss": 4.9444,
      "mean_token_accuracy": 0.22482341527938843,
      "num_tokens": 99801619.0,
      "step": 44710
    },
    {
      "entropy": 5.710837459564209,
      "epoch": 5.021901499410344,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027741690640287306,
      "loss": 4.8678,
      "mean_token_accuracy": 0.24385392367839814,
      "num_tokens": 99812972.0,
      "step": 44715
    },
    {
      "entropy": 5.696233177185059,
      "epoch": 5.022463076318302,
      "grad_norm": 1.25,
      "learning_rate": 0.0002773767599619639,
      "loss": 4.8443,
      "mean_token_accuracy": 0.23920773565769196,
      "num_tokens": 99823667.0,
      "step": 44720
    },
    {
      "entropy": 5.653852367401123,
      "epoch": 5.023024653226259,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027733661344537776,
      "loss": 4.7702,
      "mean_token_accuracy": 0.24380649477243424,
      "num_tokens": 99834654.0,
      "step": 44725
    },
    {
      "entropy": 5.741099262237549,
      "epoch": 5.0235862301342165,
      "grad_norm": 1.234375,
      "learning_rate": 0.000277296466854393,
      "loss": 4.9684,
      "mean_token_accuracy": 0.24058625102043152,
      "num_tokens": 99846393.0,
      "step": 44730
    },
    {
      "entropy": 5.689640378952026,
      "epoch": 5.024147807042175,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002772563201902877,
      "loss": 4.8188,
      "mean_token_accuracy": 0.24533071517944335,
      "num_tokens": 99857206.0,
      "step": 44735
    },
    {
      "entropy": 5.810824155807495,
      "epoch": 5.024709383950132,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027721617345434035,
      "loss": 5.0285,
      "mean_token_accuracy": 0.23812175095081328,
      "num_tokens": 99871497.0,
      "step": 44740
    },
    {
      "entropy": 5.657645559310913,
      "epoch": 5.02527096085809,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027717602664782925,
      "loss": 4.9518,
      "mean_token_accuracy": 0.22398791015148162,
      "num_tokens": 99882337.0,
      "step": 44745
    },
    {
      "entropy": 5.657897663116455,
      "epoch": 5.025832537766047,
      "grad_norm": 1.25,
      "learning_rate": 0.0002771358797720326,
      "loss": 4.8246,
      "mean_token_accuracy": 0.24071040153503417,
      "num_tokens": 99892863.0,
      "step": 44750
    },
    {
      "entropy": 5.676695537567139,
      "epoch": 5.026394114674004,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027709573282822864,
      "loss": 4.8246,
      "mean_token_accuracy": 0.24075867980718613,
      "num_tokens": 99903566.0,
      "step": 44755
    },
    {
      "entropy": 5.738750267028808,
      "epoch": 5.026955691581962,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002770555858176958,
      "loss": 4.8778,
      "mean_token_accuracy": 0.24074816852807998,
      "num_tokens": 99913466.0,
      "step": 44760
    },
    {
      "entropy": 5.724284076690674,
      "epoch": 5.02751726848992,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002770154387417123,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2326905757188797,
      "num_tokens": 99923579.0,
      "step": 44765
    },
    {
      "entropy": 5.787167644500732,
      "epoch": 5.028078845397877,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002769752916015565,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2275581806898117,
      "num_tokens": 99935471.0,
      "step": 44770
    },
    {
      "entropy": 5.82284951210022,
      "epoch": 5.028640422305835,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027693514439850646,
      "loss": 4.957,
      "mean_token_accuracy": 0.23425556123256683,
      "num_tokens": 99947082.0,
      "step": 44775
    },
    {
      "entropy": 5.755362224578858,
      "epoch": 5.029201999213792,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027689499713384085,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22833055406808853,
      "num_tokens": 99958186.0,
      "step": 44780
    },
    {
      "entropy": 5.755910205841064,
      "epoch": 5.02976357612175,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002768548498088378,
      "loss": 4.955,
      "mean_token_accuracy": 0.23390132188796997,
      "num_tokens": 99970928.0,
      "step": 44785
    },
    {
      "entropy": 5.6446209907531735,
      "epoch": 5.0303251530297075,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002768147024247756,
      "loss": 4.8403,
      "mean_token_accuracy": 0.24143043011426926,
      "num_tokens": 99982549.0,
      "step": 44790
    },
    {
      "entropy": 5.645953178405762,
      "epoch": 5.030886729937665,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002767745549829325,
      "loss": 4.7755,
      "mean_token_accuracy": 0.23579451888799668,
      "num_tokens": 99993835.0,
      "step": 44795
    },
    {
      "entropy": 5.768232631683349,
      "epoch": 5.031448306845623,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027673440748458687,
      "loss": 4.9131,
      "mean_token_accuracy": 0.23462584614753723,
      "num_tokens": 100004336.0,
      "step": 44800
    },
    {
      "entropy": 5.683434629440308,
      "epoch": 5.03200988375358,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027669425993101716,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23688265085220336,
      "num_tokens": 100015334.0,
      "step": 44805
    },
    {
      "entropy": 5.704148864746093,
      "epoch": 5.032571460661537,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002766541123235014,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22440105080604553,
      "num_tokens": 100026665.0,
      "step": 44810
    },
    {
      "entropy": 5.657615423202515,
      "epoch": 5.033133037569495,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002766139646633182,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23853999078273774,
      "num_tokens": 100037263.0,
      "step": 44815
    },
    {
      "entropy": 5.690860033035278,
      "epoch": 5.033694614477453,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027657381695174563,
      "loss": 4.8553,
      "mean_token_accuracy": 0.2355170175433159,
      "num_tokens": 100048482.0,
      "step": 44820
    },
    {
      "entropy": 5.663169717788696,
      "epoch": 5.03425619138541,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027653366919006217,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23438366800546645,
      "num_tokens": 100059351.0,
      "step": 44825
    },
    {
      "entropy": 5.695366764068604,
      "epoch": 5.034817768293368,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002764935213795461,
      "loss": 4.7703,
      "mean_token_accuracy": 0.24222986698150634,
      "num_tokens": 100069924.0,
      "step": 44830
    },
    {
      "entropy": 5.66162896156311,
      "epoch": 5.035379345201325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002764533735214757,
      "loss": 4.8093,
      "mean_token_accuracy": 0.24737904965877533,
      "num_tokens": 100080733.0,
      "step": 44835
    },
    {
      "entropy": 5.710772895812989,
      "epoch": 5.035940922109283,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027641322561712935,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2320835843682289,
      "num_tokens": 100091834.0,
      "step": 44840
    },
    {
      "entropy": 5.833002281188965,
      "epoch": 5.0365024990172405,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027637307766778524,
      "loss": 5.0987,
      "mean_token_accuracy": 0.22320126742124557,
      "num_tokens": 100103872.0,
      "step": 44845
    },
    {
      "entropy": 5.793795442581176,
      "epoch": 5.037064075925198,
      "grad_norm": 1.265625,
      "learning_rate": 0.000276332929674722,
      "loss": 4.9548,
      "mean_token_accuracy": 0.2325244814157486,
      "num_tokens": 100114391.0,
      "step": 44850
    },
    {
      "entropy": 5.742008829116822,
      "epoch": 5.037625652833156,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027629278163921763,
      "loss": 4.8628,
      "mean_token_accuracy": 0.24332905560731888,
      "num_tokens": 100125000.0,
      "step": 44855
    },
    {
      "entropy": 5.743748140335083,
      "epoch": 5.038187229741113,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027625263356255067,
      "loss": 4.9089,
      "mean_token_accuracy": 0.23987042903900146,
      "num_tokens": 100135394.0,
      "step": 44860
    },
    {
      "entropy": 5.635203838348389,
      "epoch": 5.03874880664907,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002762124854459992,
      "loss": 4.8059,
      "mean_token_accuracy": 0.2438623860478401,
      "num_tokens": 100146610.0,
      "step": 44865
    },
    {
      "entropy": 5.703980350494385,
      "epoch": 5.039310383557028,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027617233729084194,
      "loss": 4.9286,
      "mean_token_accuracy": 0.23470219373703002,
      "num_tokens": 100156936.0,
      "step": 44870
    },
    {
      "entropy": 5.689563322067261,
      "epoch": 5.0398719604649855,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027613218909835693,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2396249920129776,
      "num_tokens": 100168406.0,
      "step": 44875
    },
    {
      "entropy": 5.6399774074554445,
      "epoch": 5.040433537372944,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002760920408698225,
      "loss": 4.8575,
      "mean_token_accuracy": 0.23274324238300323,
      "num_tokens": 100179317.0,
      "step": 44880
    },
    {
      "entropy": 5.632083082199097,
      "epoch": 5.040995114280901,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00027605189260651716,
      "loss": 4.7867,
      "mean_token_accuracy": 0.24021087884902953,
      "num_tokens": 100190916.0,
      "step": 44885
    },
    {
      "entropy": 5.675491142272949,
      "epoch": 5.041556691188858,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002760117443097191,
      "loss": 4.7687,
      "mean_token_accuracy": 0.24911413788795472,
      "num_tokens": 100201437.0,
      "step": 44890
    },
    {
      "entropy": 5.697387409210205,
      "epoch": 5.042118268096816,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027597159598070675,
      "loss": 4.8384,
      "mean_token_accuracy": 0.237562195956707,
      "num_tokens": 100212405.0,
      "step": 44895
    },
    {
      "entropy": 5.669723129272461,
      "epoch": 5.042679845004773,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027593144762075837,
      "loss": 4.8137,
      "mean_token_accuracy": 0.24232966750860213,
      "num_tokens": 100223428.0,
      "step": 44900
    },
    {
      "entropy": 5.604140949249268,
      "epoch": 5.043241421912731,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027589129923115234,
      "loss": 4.7854,
      "mean_token_accuracy": 0.24240604192018508,
      "num_tokens": 100234128.0,
      "step": 44905
    },
    {
      "entropy": 5.714807605743408,
      "epoch": 5.043802998820689,
      "grad_norm": 1.234375,
      "learning_rate": 0.000275851150813167,
      "loss": 4.9178,
      "mean_token_accuracy": 0.22955562472343444,
      "num_tokens": 100244920.0,
      "step": 44910
    },
    {
      "entropy": 5.661243677139282,
      "epoch": 5.044364575728646,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002758110023680807,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2375393897294998,
      "num_tokens": 100256518.0,
      "step": 44915
    },
    {
      "entropy": 5.736815452575684,
      "epoch": 5.044926152636603,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002757708538971717,
      "loss": 4.9423,
      "mean_token_accuracy": 0.23749881684780122,
      "num_tokens": 100268203.0,
      "step": 44920
    },
    {
      "entropy": 5.735644197463989,
      "epoch": 5.045487729544561,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002757307054017185,
      "loss": 4.825,
      "mean_token_accuracy": 0.2384873613715172,
      "num_tokens": 100279440.0,
      "step": 44925
    },
    {
      "entropy": 5.745801734924316,
      "epoch": 5.0460493064525185,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002756905568829994,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2376450777053833,
      "num_tokens": 100290616.0,
      "step": 44930
    },
    {
      "entropy": 5.719456005096435,
      "epoch": 5.046610883360477,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027565040834229266,
      "loss": 4.9568,
      "mean_token_accuracy": 0.23069950342178344,
      "num_tokens": 100301602.0,
      "step": 44935
    },
    {
      "entropy": 5.725000858306885,
      "epoch": 5.047172460268434,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027561025978087663,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23337246030569075,
      "num_tokens": 100313311.0,
      "step": 44940
    },
    {
      "entropy": 5.747299909591675,
      "epoch": 5.047734037176391,
      "grad_norm": 1.171875,
      "learning_rate": 0.00027557011120002975,
      "loss": 4.9429,
      "mean_token_accuracy": 0.23230600506067275,
      "num_tokens": 100325394.0,
      "step": 44945
    },
    {
      "entropy": 5.733774995803833,
      "epoch": 5.048295614084349,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002755299626010302,
      "loss": 4.8933,
      "mean_token_accuracy": 0.23485625684261321,
      "num_tokens": 100336643.0,
      "step": 44950
    },
    {
      "entropy": 5.734191131591797,
      "epoch": 5.048857190992306,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002754898139851565,
      "loss": 4.9297,
      "mean_token_accuracy": 0.23231476545333862,
      "num_tokens": 100348409.0,
      "step": 44955
    },
    {
      "entropy": 5.72011170387268,
      "epoch": 5.049418767900264,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00027544966535368695,
      "loss": 4.9573,
      "mean_token_accuracy": 0.2356201484799385,
      "num_tokens": 100360115.0,
      "step": 44960
    },
    {
      "entropy": 5.672696876525879,
      "epoch": 5.049980344808222,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002754095167078999,
      "loss": 4.8659,
      "mean_token_accuracy": 0.23529339730739593,
      "num_tokens": 100369618.0,
      "step": 44965
    },
    {
      "entropy": 5.669945192337036,
      "epoch": 5.050541921716179,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002753693680490736,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2364954948425293,
      "num_tokens": 100379470.0,
      "step": 44970
    },
    {
      "entropy": 5.636484098434448,
      "epoch": 5.051103498624137,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002753292193784865,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23490376770496368,
      "num_tokens": 100390490.0,
      "step": 44975
    },
    {
      "entropy": 5.690299797058105,
      "epoch": 5.051665075532094,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027528907069741696,
      "loss": 4.8507,
      "mean_token_accuracy": 0.23709446489810942,
      "num_tokens": 100402817.0,
      "step": 44980
    },
    {
      "entropy": 5.8003129959106445,
      "epoch": 5.052226652440051,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002752489220071433,
      "loss": 4.977,
      "mean_token_accuracy": 0.22809768915176393,
      "num_tokens": 100413411.0,
      "step": 44985
    },
    {
      "entropy": 5.630863857269287,
      "epoch": 5.0527882293480095,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002752087733089438,
      "loss": 4.8763,
      "mean_token_accuracy": 0.23523022830486298,
      "num_tokens": 100425118.0,
      "step": 44990
    },
    {
      "entropy": 5.77785587310791,
      "epoch": 5.053349806255967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002751686246040969,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22903341352939605,
      "num_tokens": 100436544.0,
      "step": 44995
    },
    {
      "entropy": 5.754489946365356,
      "epoch": 5.053911383163924,
      "grad_norm": 1.25,
      "learning_rate": 0.000275128475893881,
      "loss": 4.8939,
      "mean_token_accuracy": 0.2374609261751175,
      "num_tokens": 100448315.0,
      "step": 45000
    },
    {
      "epoch": 5.053911383163924,
      "eval_entropy": 5.440148228986502,
      "eval_loss": 5.0493340492248535,
      "eval_mean_token_accuracy": 0.23331864904790064,
      "eval_num_tokens": 100448315.0,
      "eval_runtime": 18.427,
      "eval_samples_per_second": 1683.021,
      "eval_steps_per_second": 210.398,
      "step": 45000
    },
    {
      "entropy": 5.644769477844238,
      "epoch": 5.054472960071882,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027508832717957433,
      "loss": 4.8765,
      "mean_token_accuracy": 0.24057786762714387,
      "num_tokens": 100459548.0,
      "step": 45005
    },
    {
      "entropy": 5.786318254470825,
      "epoch": 5.055034536979839,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002750481784624553,
      "loss": 4.9001,
      "mean_token_accuracy": 0.24056164771318436,
      "num_tokens": 100468687.0,
      "step": 45010
    },
    {
      "entropy": 5.6843681812286375,
      "epoch": 5.0555961138877965,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027500802974380224,
      "loss": 4.8744,
      "mean_token_accuracy": 0.23431380093097687,
      "num_tokens": 100479607.0,
      "step": 45015
    },
    {
      "entropy": 5.725704097747803,
      "epoch": 5.056157690795755,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002749678810248935,
      "loss": 4.9732,
      "mean_token_accuracy": 0.23629752844572066,
      "num_tokens": 100492245.0,
      "step": 45020
    },
    {
      "entropy": 5.717809009552002,
      "epoch": 5.056719267703712,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002749277323070075,
      "loss": 4.8272,
      "mean_token_accuracy": 0.23717208355665206,
      "num_tokens": 100503973.0,
      "step": 45025
    },
    {
      "entropy": 5.707981538772583,
      "epoch": 5.05728084461167,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027488758359142245,
      "loss": 4.8759,
      "mean_token_accuracy": 0.23878735899925232,
      "num_tokens": 100516003.0,
      "step": 45030
    },
    {
      "entropy": 5.7220165729522705,
      "epoch": 5.057842421519627,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002748474348794167,
      "loss": 4.9728,
      "mean_token_accuracy": 0.23120505809783937,
      "num_tokens": 100528130.0,
      "step": 45035
    },
    {
      "entropy": 5.68250617980957,
      "epoch": 5.058403998427584,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027480728617226884,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23426709175109864,
      "num_tokens": 100539532.0,
      "step": 45040
    },
    {
      "entropy": 5.718720293045044,
      "epoch": 5.0589655753355425,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000274767137471257,
      "loss": 4.9923,
      "mean_token_accuracy": 0.2268755242228508,
      "num_tokens": 100550520.0,
      "step": 45045
    },
    {
      "entropy": 5.673422479629517,
      "epoch": 5.0595271522435,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027472698877765967,
      "loss": 4.7817,
      "mean_token_accuracy": 0.24016674906015395,
      "num_tokens": 100561596.0,
      "step": 45050
    },
    {
      "entropy": 5.696254396438599,
      "epoch": 5.060088729151457,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002746868400927551,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23773179054260254,
      "num_tokens": 100571238.0,
      "step": 45055
    },
    {
      "entropy": 5.708683586120605,
      "epoch": 5.060650306059415,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002746466914178216,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22940977811813354,
      "num_tokens": 100582687.0,
      "step": 45060
    },
    {
      "entropy": 5.704419898986816,
      "epoch": 5.061211882967372,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002746065427541376,
      "loss": 4.9078,
      "mean_token_accuracy": 0.23347044587135315,
      "num_tokens": 100594270.0,
      "step": 45065
    },
    {
      "entropy": 5.721216773986816,
      "epoch": 5.06177345987533,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027456639410298144,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22911058515310287,
      "num_tokens": 100605449.0,
      "step": 45070
    },
    {
      "entropy": 5.618840456008911,
      "epoch": 5.062335036783288,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002745262454656314,
      "loss": 4.7972,
      "mean_token_accuracy": 0.24050319343805313,
      "num_tokens": 100616985.0,
      "step": 45075
    },
    {
      "entropy": 5.635556554794311,
      "epoch": 5.062896613691245,
      "grad_norm": 1.296875,
      "learning_rate": 0.000274486096843366,
      "loss": 4.8639,
      "mean_token_accuracy": 0.24128866344690322,
      "num_tokens": 100628088.0,
      "step": 45080
    },
    {
      "entropy": 5.786492872238159,
      "epoch": 5.063458190599203,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027444594823746344,
      "loss": 5.0101,
      "mean_token_accuracy": 0.2336941495537758,
      "num_tokens": 100639805.0,
      "step": 45085
    },
    {
      "entropy": 5.877184581756592,
      "epoch": 5.06401976750716,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002744057996492021,
      "loss": 5.0812,
      "mean_token_accuracy": 0.22209070920944213,
      "num_tokens": 100650437.0,
      "step": 45090
    },
    {
      "entropy": 5.692848968505859,
      "epoch": 5.064581344415117,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002743656510798603,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23084059059619905,
      "num_tokens": 100661466.0,
      "step": 45095
    },
    {
      "entropy": 5.745228481292725,
      "epoch": 5.065142921323075,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002743255025307165,
      "loss": 4.9097,
      "mean_token_accuracy": 0.24112665504217148,
      "num_tokens": 100671706.0,
      "step": 45100
    },
    {
      "entropy": 5.690116119384766,
      "epoch": 5.065704498231033,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027428535400304886,
      "loss": 4.8686,
      "mean_token_accuracy": 0.2369756355881691,
      "num_tokens": 100682592.0,
      "step": 45105
    },
    {
      "entropy": 5.706077241897583,
      "epoch": 5.06626607513899,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002742452054981358,
      "loss": 4.8868,
      "mean_token_accuracy": 0.23508809357881547,
      "num_tokens": 100693893.0,
      "step": 45110
    },
    {
      "entropy": 5.611273574829101,
      "epoch": 5.066827652046948,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002742050570172558,
      "loss": 4.8593,
      "mean_token_accuracy": 0.2391526684165001,
      "num_tokens": 100705455.0,
      "step": 45115
    },
    {
      "entropy": 5.684170579910278,
      "epoch": 5.067389228954905,
      "grad_norm": 1.28125,
      "learning_rate": 0.000274164908561687,
      "loss": 4.822,
      "mean_token_accuracy": 0.23794289082288742,
      "num_tokens": 100714945.0,
      "step": 45120
    },
    {
      "entropy": 5.723953485488892,
      "epoch": 5.067950805862863,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027412476013270786,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2286505326628685,
      "num_tokens": 100726299.0,
      "step": 45125
    },
    {
      "entropy": 5.769934892654419,
      "epoch": 5.0685123827708205,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027408461173159676,
      "loss": 5.026,
      "mean_token_accuracy": 0.22823247760534288,
      "num_tokens": 100739415.0,
      "step": 45130
    },
    {
      "entropy": 5.777725315093994,
      "epoch": 5.069073959678778,
      "grad_norm": 1.171875,
      "learning_rate": 0.00027404446335963195,
      "loss": 5.0284,
      "mean_token_accuracy": 0.2254694491624832,
      "num_tokens": 100751018.0,
      "step": 45135
    },
    {
      "entropy": 5.641719102859497,
      "epoch": 5.069635536586736,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002740043150180918,
      "loss": 4.8444,
      "mean_token_accuracy": 0.23650299608707429,
      "num_tokens": 100761425.0,
      "step": 45140
    },
    {
      "entropy": 5.725146722793579,
      "epoch": 5.070197113494693,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002739641667082546,
      "loss": 4.9249,
      "mean_token_accuracy": 0.23044222593307495,
      "num_tokens": 100772906.0,
      "step": 45145
    },
    {
      "entropy": 5.720306062698365,
      "epoch": 5.07075869040265,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027392401843139876,
      "loss": 4.9679,
      "mean_token_accuracy": 0.23019781559705735,
      "num_tokens": 100784106.0,
      "step": 45150
    },
    {
      "entropy": 5.72959041595459,
      "epoch": 5.071320267310608,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002738838701888026,
      "loss": 5.0172,
      "mean_token_accuracy": 0.22488956898450851,
      "num_tokens": 100795066.0,
      "step": 45155
    },
    {
      "entropy": 5.652118730545044,
      "epoch": 5.071881844218566,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002738437219817444,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2359546661376953,
      "num_tokens": 100806936.0,
      "step": 45160
    },
    {
      "entropy": 5.546273851394654,
      "epoch": 5.072443421126524,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027380357381150263,
      "loss": 4.768,
      "mean_token_accuracy": 0.23791221678256988,
      "num_tokens": 100817483.0,
      "step": 45165
    },
    {
      "entropy": 5.774194049835205,
      "epoch": 5.073004998034481,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002737634256793555,
      "loss": 5.018,
      "mean_token_accuracy": 0.22742094099521637,
      "num_tokens": 100828482.0,
      "step": 45170
    },
    {
      "entropy": 5.727821779251099,
      "epoch": 5.073566574942438,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027372327758658135,
      "loss": 4.9431,
      "mean_token_accuracy": 0.23386599570512773,
      "num_tokens": 100839745.0,
      "step": 45175
    },
    {
      "entropy": 5.682489109039307,
      "epoch": 5.074128151850396,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027368312953445845,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2327523335814476,
      "num_tokens": 100850441.0,
      "step": 45180
    },
    {
      "entropy": 5.618262147903442,
      "epoch": 5.0746897287583534,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002736429815242653,
      "loss": 4.8535,
      "mean_token_accuracy": 0.23466782569885253,
      "num_tokens": 100861428.0,
      "step": 45185
    },
    {
      "entropy": 5.686855173110962,
      "epoch": 5.075251305666311,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002736028335572802,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23583993911743165,
      "num_tokens": 100872874.0,
      "step": 45190
    },
    {
      "entropy": 5.751693058013916,
      "epoch": 5.075812882574269,
      "grad_norm": 1.25,
      "learning_rate": 0.00027356268563478135,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21852253526449203,
      "num_tokens": 100884150.0,
      "step": 45195
    },
    {
      "entropy": 5.721016788482666,
      "epoch": 5.076374459482226,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002735225377580472,
      "loss": 4.8816,
      "mean_token_accuracy": 0.23695822507143022,
      "num_tokens": 100896440.0,
      "step": 45200
    },
    {
      "entropy": 5.709068632125854,
      "epoch": 5.076936036390183,
      "grad_norm": 1.25,
      "learning_rate": 0.000273482389928356,
      "loss": 4.8793,
      "mean_token_accuracy": 0.24136842787265778,
      "num_tokens": 100906854.0,
      "step": 45205
    },
    {
      "entropy": 5.671851539611817,
      "epoch": 5.077497613298141,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002734422421469861,
      "loss": 4.87,
      "mean_token_accuracy": 0.24520527869462966,
      "num_tokens": 100918644.0,
      "step": 45210
    },
    {
      "entropy": 5.7214515209198,
      "epoch": 5.0780591902060985,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002734020944152157,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2358969882130623,
      "num_tokens": 100929313.0,
      "step": 45215
    },
    {
      "entropy": 5.7440611839294435,
      "epoch": 5.078620767114057,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002733619467343234,
      "loss": 4.9166,
      "mean_token_accuracy": 0.23331734240055085,
      "num_tokens": 100940406.0,
      "step": 45220
    },
    {
      "entropy": 5.672518396377564,
      "epoch": 5.079182344022014,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002733217991055873,
      "loss": 4.8318,
      "mean_token_accuracy": 0.23810425102710725,
      "num_tokens": 100951086.0,
      "step": 45225
    },
    {
      "entropy": 5.67860426902771,
      "epoch": 5.079743920929971,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002732816515302858,
      "loss": 4.961,
      "mean_token_accuracy": 0.23380398899316787,
      "num_tokens": 100961417.0,
      "step": 45230
    },
    {
      "entropy": 5.692643785476685,
      "epoch": 5.080305497837929,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002732415040096971,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2368110880255699,
      "num_tokens": 100973667.0,
      "step": 45235
    },
    {
      "entropy": 5.746442747116089,
      "epoch": 5.080867074745886,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027320135654509964,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2335373044013977,
      "num_tokens": 100984891.0,
      "step": 45240
    },
    {
      "entropy": 5.713233804702758,
      "epoch": 5.081428651653844,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002731612091377717,
      "loss": 4.9347,
      "mean_token_accuracy": 0.232599076628685,
      "num_tokens": 100997273.0,
      "step": 45245
    },
    {
      "entropy": 5.582094764709472,
      "epoch": 5.081990228561802,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027312106178899154,
      "loss": 4.7353,
      "mean_token_accuracy": 0.2526587024331093,
      "num_tokens": 101007341.0,
      "step": 45250
    },
    {
      "entropy": 5.737540435791016,
      "epoch": 5.082551805469759,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002730809145000374,
      "loss": 4.9553,
      "mean_token_accuracy": 0.23121078610420226,
      "num_tokens": 101017471.0,
      "step": 45255
    },
    {
      "entropy": 5.64499192237854,
      "epoch": 5.083113382377717,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002730407672721878,
      "loss": 4.8319,
      "mean_token_accuracy": 0.2405970498919487,
      "num_tokens": 101028726.0,
      "step": 45260
    },
    {
      "entropy": 5.855201196670532,
      "epoch": 5.083674959285674,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000273000620106721,
      "loss": 5.0444,
      "mean_token_accuracy": 0.22114395201206208,
      "num_tokens": 101040824.0,
      "step": 45265
    },
    {
      "entropy": 5.731151008605957,
      "epoch": 5.0842365361936315,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002729604730049151,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2348433405160904,
      "num_tokens": 101051849.0,
      "step": 45270
    },
    {
      "entropy": 5.6836799621582035,
      "epoch": 5.08479811310159,
      "grad_norm": 1.375,
      "learning_rate": 0.0002729203259680486,
      "loss": 4.9241,
      "mean_token_accuracy": 0.23537662625312805,
      "num_tokens": 101062667.0,
      "step": 45275
    },
    {
      "entropy": 5.695635795593262,
      "epoch": 5.085359690009547,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002728801789973998,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2336274728178978,
      "num_tokens": 101073679.0,
      "step": 45280
    },
    {
      "entropy": 5.793031644821167,
      "epoch": 5.085921266917504,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002728400320942468,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22986719459295274,
      "num_tokens": 101085394.0,
      "step": 45285
    },
    {
      "entropy": 5.839059925079345,
      "epoch": 5.086482843825462,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000272799885259868,
      "loss": 5.0073,
      "mean_token_accuracy": 0.22709112316370011,
      "num_tokens": 101096748.0,
      "step": 45290
    },
    {
      "entropy": 5.651520538330078,
      "epoch": 5.087044420733419,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002727597384955418,
      "loss": 4.7415,
      "mean_token_accuracy": 0.24699889868497849,
      "num_tokens": 101107704.0,
      "step": 45295
    },
    {
      "entropy": 5.7099710464477536,
      "epoch": 5.0876059976413766,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00027271959180254636,
      "loss": 4.8677,
      "mean_token_accuracy": 0.2355949893593788,
      "num_tokens": 101119890.0,
      "step": 45300
    },
    {
      "entropy": 5.680596590042114,
      "epoch": 5.088167574549335,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027267944518215997,
      "loss": 4.8673,
      "mean_token_accuracy": 0.23373135626316072,
      "num_tokens": 101130172.0,
      "step": 45305
    },
    {
      "entropy": 5.678371572494507,
      "epoch": 5.088729151457292,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027263929863566104,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2440643936395645,
      "num_tokens": 101141374.0,
      "step": 45310
    },
    {
      "entropy": 5.716498279571534,
      "epoch": 5.08929072836525,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002725991521643277,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22646607160568238,
      "num_tokens": 101155192.0,
      "step": 45315
    },
    {
      "entropy": 5.763027858734131,
      "epoch": 5.089852305273207,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002725590057694383,
      "loss": 4.9338,
      "mean_token_accuracy": 0.23772047460079193,
      "num_tokens": 101168438.0,
      "step": 45320
    },
    {
      "entropy": 5.748255968093872,
      "epoch": 5.090413882181164,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027251885945227093,
      "loss": 4.9148,
      "mean_token_accuracy": 0.23570625185966493,
      "num_tokens": 101180609.0,
      "step": 45325
    },
    {
      "entropy": 5.764480400085449,
      "epoch": 5.0909754590891225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027247871321410425,
      "loss": 4.869,
      "mean_token_accuracy": 0.23844571262598038,
      "num_tokens": 101191211.0,
      "step": 45330
    },
    {
      "entropy": 5.676105165481568,
      "epoch": 5.09153703599708,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027243856705621623,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23427218794822693,
      "num_tokens": 101202706.0,
      "step": 45335
    },
    {
      "entropy": 5.6791894912719725,
      "epoch": 5.092098612905037,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027239842097988526,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23463848680257798,
      "num_tokens": 101213890.0,
      "step": 45340
    },
    {
      "entropy": 5.792019987106324,
      "epoch": 5.092660189812995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002723582749863896,
      "loss": 5.0177,
      "mean_token_accuracy": 0.22426789700984956,
      "num_tokens": 101225946.0,
      "step": 45345
    },
    {
      "entropy": 5.781707239151001,
      "epoch": 5.093221766720952,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00027231812907700746,
      "loss": 4.9385,
      "mean_token_accuracy": 0.22973111569881438,
      "num_tokens": 101236269.0,
      "step": 45350
    },
    {
      "entropy": 5.728690242767334,
      "epoch": 5.09378334362891,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002722779832530171,
      "loss": 4.9194,
      "mean_token_accuracy": 0.23850345462560654,
      "num_tokens": 101247641.0,
      "step": 45355
    },
    {
      "entropy": 5.686880302429199,
      "epoch": 5.094344920536868,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027223783751569687,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22777061015367508,
      "num_tokens": 101259271.0,
      "step": 45360
    },
    {
      "entropy": 5.6074230670928955,
      "epoch": 5.094906497444825,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00027219769186632493,
      "loss": 4.7915,
      "mean_token_accuracy": 0.24276457875967025,
      "num_tokens": 101269709.0,
      "step": 45365
    },
    {
      "entropy": 5.716828489303589,
      "epoch": 5.095468074352783,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002721575463061796,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2236262410879135,
      "num_tokens": 101280559.0,
      "step": 45370
    },
    {
      "entropy": 5.766753101348877,
      "epoch": 5.09602965126074,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027211740083653903,
      "loss": 4.9266,
      "mean_token_accuracy": 0.23432591706514358,
      "num_tokens": 101292356.0,
      "step": 45375
    },
    {
      "entropy": 5.7241274356842045,
      "epoch": 5.096591228168697,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002720772554586816,
      "loss": 4.8928,
      "mean_token_accuracy": 0.23781895339488984,
      "num_tokens": 101303881.0,
      "step": 45380
    },
    {
      "entropy": 5.698309707641601,
      "epoch": 5.0971528050766555,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002720371101738855,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2435091495513916,
      "num_tokens": 101315310.0,
      "step": 45385
    },
    {
      "entropy": 5.636500024795533,
      "epoch": 5.097714381984613,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00027199696498342893,
      "loss": 4.8374,
      "mean_token_accuracy": 0.24113102853298188,
      "num_tokens": 101327085.0,
      "step": 45390
    },
    {
      "entropy": 5.653819036483765,
      "epoch": 5.09827595889257,
      "grad_norm": 1.25,
      "learning_rate": 0.0002719568198885901,
      "loss": 4.869,
      "mean_token_accuracy": 0.23687816709280013,
      "num_tokens": 101340473.0,
      "step": 45395
    },
    {
      "entropy": 5.727573442459106,
      "epoch": 5.098837535800528,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002719166748906473,
      "loss": 4.8795,
      "mean_token_accuracy": 0.24219955056905745,
      "num_tokens": 101353434.0,
      "step": 45400
    },
    {
      "entropy": 5.668012619018555,
      "epoch": 5.099399112708485,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027187652999087887,
      "loss": 4.8532,
      "mean_token_accuracy": 0.2320242166519165,
      "num_tokens": 101363782.0,
      "step": 45405
    },
    {
      "entropy": 5.712268877029419,
      "epoch": 5.099960689616443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027183638519056284,
      "loss": 4.9304,
      "mean_token_accuracy": 0.2316008523106575,
      "num_tokens": 101374985.0,
      "step": 45410
    },
    {
      "entropy": 5.6718967914581295,
      "epoch": 5.1005222665244005,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002717962404909776,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23953409492969513,
      "num_tokens": 101386916.0,
      "step": 45415
    },
    {
      "entropy": 5.66069598197937,
      "epoch": 5.101083843432358,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027175609589340135,
      "loss": 4.8097,
      "mean_token_accuracy": 0.23567781895399093,
      "num_tokens": 101397656.0,
      "step": 45420
    },
    {
      "entropy": 5.658000469207764,
      "epoch": 5.101645420340316,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002717159513991122,
      "loss": 4.8413,
      "mean_token_accuracy": 0.24060064256191255,
      "num_tokens": 101408725.0,
      "step": 45425
    },
    {
      "entropy": 5.595779848098755,
      "epoch": 5.102206997248273,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00027167580700938844,
      "loss": 4.7998,
      "mean_token_accuracy": 0.24589419960975648,
      "num_tokens": 101419580.0,
      "step": 45430
    },
    {
      "entropy": 5.698503684997559,
      "epoch": 5.10276857415623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027163566272550826,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22995340079069138,
      "num_tokens": 101430735.0,
      "step": 45435
    },
    {
      "entropy": 5.690603113174438,
      "epoch": 5.103330151064188,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027159551854874986,
      "loss": 4.8743,
      "mean_token_accuracy": 0.2357979655265808,
      "num_tokens": 101440544.0,
      "step": 45440
    },
    {
      "entropy": 5.759085369110108,
      "epoch": 5.103891727972146,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002715553744803915,
      "loss": 4.9408,
      "mean_token_accuracy": 0.23103067725896836,
      "num_tokens": 101450793.0,
      "step": 45445
    },
    {
      "entropy": 5.730669927597046,
      "epoch": 5.104453304880104,
      "grad_norm": 1.25,
      "learning_rate": 0.0002715152305217114,
      "loss": 5.0283,
      "mean_token_accuracy": 0.23068908452987671,
      "num_tokens": 101463593.0,
      "step": 45450
    },
    {
      "entropy": 5.667651653289795,
      "epoch": 5.105014881788061,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002714750866739877,
      "loss": 4.8668,
      "mean_token_accuracy": 0.2320212870836258,
      "num_tokens": 101474334.0,
      "step": 45455
    },
    {
      "entropy": 5.703938341140747,
      "epoch": 5.105576458696018,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002714349429384986,
      "loss": 4.9391,
      "mean_token_accuracy": 0.2324763461947441,
      "num_tokens": 101485356.0,
      "step": 45460
    },
    {
      "entropy": 5.689621496200561,
      "epoch": 5.106138035603976,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027139479931652226,
      "loss": 4.848,
      "mean_token_accuracy": 0.2367057040333748,
      "num_tokens": 101495939.0,
      "step": 45465
    },
    {
      "entropy": 5.696870803833008,
      "epoch": 5.1066996125119335,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002713546558093368,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2317931979894638,
      "num_tokens": 101505563.0,
      "step": 45470
    },
    {
      "entropy": 5.585991764068604,
      "epoch": 5.107261189419891,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002713145124182207,
      "loss": 4.784,
      "mean_token_accuracy": 0.24574816226959229,
      "num_tokens": 101516829.0,
      "step": 45475
    },
    {
      "entropy": 5.706165313720703,
      "epoch": 5.107822766327849,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00027127436914445194,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2319951355457306,
      "num_tokens": 101527427.0,
      "step": 45480
    },
    {
      "entropy": 5.7284986019134525,
      "epoch": 5.108384343235806,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002712342259893086,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23769186288118363,
      "num_tokens": 101537991.0,
      "step": 45485
    },
    {
      "entropy": 5.692518138885498,
      "epoch": 5.108945920143763,
      "grad_norm": 1.234375,
      "learning_rate": 0.00027119408295406906,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2239145502448082,
      "num_tokens": 101549485.0,
      "step": 45490
    },
    {
      "entropy": 5.738473320007325,
      "epoch": 5.109507497051721,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002711539400400114,
      "loss": 4.9526,
      "mean_token_accuracy": 0.22957140654325486,
      "num_tokens": 101560719.0,
      "step": 45495
    },
    {
      "entropy": 5.737746238708496,
      "epoch": 5.110069073959679,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027111379724841373,
      "loss": 5.0218,
      "mean_token_accuracy": 0.22286913841962813,
      "num_tokens": 101572118.0,
      "step": 45500
    },
    {
      "entropy": 5.805760860443115,
      "epoch": 5.110630650867637,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027107365458055424,
      "loss": 4.9492,
      "mean_token_accuracy": 0.23686417937278748,
      "num_tokens": 101582050.0,
      "step": 45505
    },
    {
      "entropy": 5.690562343597412,
      "epoch": 5.111192227775594,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002710335120377112,
      "loss": 4.7592,
      "mean_token_accuracy": 0.238247150182724,
      "num_tokens": 101592529.0,
      "step": 45510
    },
    {
      "entropy": 5.705385017395019,
      "epoch": 5.111753804683551,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027099336962116253,
      "loss": 4.9144,
      "mean_token_accuracy": 0.23154906183481216,
      "num_tokens": 101603106.0,
      "step": 45515
    },
    {
      "entropy": 5.741144514083862,
      "epoch": 5.112315381591509,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027095322733218665,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22692548334598542,
      "num_tokens": 101614171.0,
      "step": 45520
    },
    {
      "entropy": 5.7780884265899655,
      "epoch": 5.112876958499466,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002709130851720616,
      "loss": 4.9826,
      "mean_token_accuracy": 0.22624462842941284,
      "num_tokens": 101626311.0,
      "step": 45525
    },
    {
      "entropy": 5.633467006683349,
      "epoch": 5.113438535407424,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027087294314206544,
      "loss": 4.845,
      "mean_token_accuracy": 0.2393480733036995,
      "num_tokens": 101637923.0,
      "step": 45530
    },
    {
      "entropy": 5.770192956924438,
      "epoch": 5.114000112315382,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002708328012434764,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2222980424761772,
      "num_tokens": 101647943.0,
      "step": 45535
    },
    {
      "entropy": 5.679735803604126,
      "epoch": 5.114561689223339,
      "grad_norm": 1.25,
      "learning_rate": 0.0002707926594775725,
      "loss": 4.9119,
      "mean_token_accuracy": 0.23105159103870393,
      "num_tokens": 101660095.0,
      "step": 45540
    },
    {
      "entropy": 5.630959367752075,
      "epoch": 5.115123266131297,
      "grad_norm": 1.328125,
      "learning_rate": 0.00027075251784563196,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23851377367973328,
      "num_tokens": 101670092.0,
      "step": 45545
    },
    {
      "entropy": 5.684062862396241,
      "epoch": 5.115684843039254,
      "grad_norm": 1.234375,
      "learning_rate": 0.000270712376348933,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22917410284280776,
      "num_tokens": 101681692.0,
      "step": 45550
    },
    {
      "entropy": 5.714618778228759,
      "epoch": 5.1162464199472115,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002706722349887536,
      "loss": 4.893,
      "mean_token_accuracy": 0.2330215260386467,
      "num_tokens": 101692569.0,
      "step": 45555
    },
    {
      "entropy": 5.809184837341308,
      "epoch": 5.11680799685517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002706320937663719,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2336348071694374,
      "num_tokens": 101704185.0,
      "step": 45560
    },
    {
      "entropy": 5.777782297134399,
      "epoch": 5.117369573763127,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002705919526830661,
      "loss": 4.9639,
      "mean_token_accuracy": 0.23020407408475876,
      "num_tokens": 101715500.0,
      "step": 45565
    },
    {
      "entropy": 5.779625606536865,
      "epoch": 5.117931150671084,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027055181174011415,
      "loss": 4.9785,
      "mean_token_accuracy": 0.23713594377040864,
      "num_tokens": 101727562.0,
      "step": 45570
    },
    {
      "entropy": 5.6891251564025875,
      "epoch": 5.118492727579042,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00027051167093879424,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23981639593839646,
      "num_tokens": 101739269.0,
      "step": 45575
    },
    {
      "entropy": 5.7065185546875,
      "epoch": 5.119054304486999,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002704715302803846,
      "loss": 4.9646,
      "mean_token_accuracy": 0.228560109436512,
      "num_tokens": 101750189.0,
      "step": 45580
    },
    {
      "entropy": 5.740209293365479,
      "epoch": 5.119615881394957,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027043138976616304,
      "loss": 4.9412,
      "mean_token_accuracy": 0.23162226378917694,
      "num_tokens": 101760824.0,
      "step": 45585
    },
    {
      "entropy": 5.657158994674683,
      "epoch": 5.120177458302915,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00027039124939740787,
      "loss": 4.7874,
      "mean_token_accuracy": 0.24440808147192,
      "num_tokens": 101772647.0,
      "step": 45590
    },
    {
      "entropy": 5.700880622863769,
      "epoch": 5.120739035210872,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027035110917539713,
      "loss": 4.9676,
      "mean_token_accuracy": 0.23346189111471177,
      "num_tokens": 101783020.0,
      "step": 45595
    },
    {
      "entropy": 5.711303663253784,
      "epoch": 5.12130061211883,
      "grad_norm": 1.265625,
      "learning_rate": 0.00027031096910140877,
      "loss": 4.8775,
      "mean_token_accuracy": 0.23781560510396957,
      "num_tokens": 101793372.0,
      "step": 45600
    },
    {
      "entropy": 5.681576585769653,
      "epoch": 5.121862189026787,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027027082917672107,
      "loss": 4.8915,
      "mean_token_accuracy": 0.23951225429773332,
      "num_tokens": 101805077.0,
      "step": 45605
    },
    {
      "entropy": 5.697350931167603,
      "epoch": 5.1224237659347445,
      "grad_norm": 1.21875,
      "learning_rate": 0.00027023068940261203,
      "loss": 4.9319,
      "mean_token_accuracy": 0.23409998863935472,
      "num_tokens": 101816334.0,
      "step": 45610
    },
    {
      "entropy": 5.7646932125091555,
      "epoch": 5.122985342842703,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002701905497803595,
      "loss": 5.042,
      "mean_token_accuracy": 0.22646797597408294,
      "num_tokens": 101827812.0,
      "step": 45615
    },
    {
      "entropy": 5.664176607131958,
      "epoch": 5.12354691975066,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002701504103112419,
      "loss": 4.7851,
      "mean_token_accuracy": 0.24320604354143144,
      "num_tokens": 101839400.0,
      "step": 45620
    },
    {
      "entropy": 5.766156482696533,
      "epoch": 5.124108496658617,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027011027099653706,
      "loss": 4.9774,
      "mean_token_accuracy": 0.2357589438557625,
      "num_tokens": 101851948.0,
      "step": 45625
    },
    {
      "entropy": 5.685453414916992,
      "epoch": 5.124670073566575,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002700701318375232,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22913980185985566,
      "num_tokens": 101862985.0,
      "step": 45630
    },
    {
      "entropy": 5.688057947158813,
      "epoch": 5.125231650474532,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002700299928354781,
      "loss": 4.8787,
      "mean_token_accuracy": 0.23550950288772582,
      "num_tokens": 101873247.0,
      "step": 45635
    },
    {
      "entropy": 5.718344068527221,
      "epoch": 5.12579322738249,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026998985399168005,
      "loss": 4.9565,
      "mean_token_accuracy": 0.23410502225160598,
      "num_tokens": 101883686.0,
      "step": 45640
    },
    {
      "entropy": 5.771374702453613,
      "epoch": 5.126354804290448,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026994971530740696,
      "loss": 4.9205,
      "mean_token_accuracy": 0.2305506706237793,
      "num_tokens": 101894905.0,
      "step": 45645
    },
    {
      "entropy": 5.691358184814453,
      "epoch": 5.126916381198405,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026990957678393684,
      "loss": 4.9121,
      "mean_token_accuracy": 0.233327154815197,
      "num_tokens": 101906671.0,
      "step": 45650
    },
    {
      "entropy": 5.616622638702393,
      "epoch": 5.127477958106363,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026986943842254783,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23786109536886216,
      "num_tokens": 101917440.0,
      "step": 45655
    },
    {
      "entropy": 5.718140554428101,
      "epoch": 5.12803953501432,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002698293002245179,
      "loss": 4.9297,
      "mean_token_accuracy": 0.23412442505359649,
      "num_tokens": 101929177.0,
      "step": 45660
    },
    {
      "entropy": 5.772738456726074,
      "epoch": 5.128601111922277,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026978916219112505,
      "loss": 4.9518,
      "mean_token_accuracy": 0.23179252445697784,
      "num_tokens": 101940034.0,
      "step": 45665
    },
    {
      "entropy": 5.813388872146606,
      "epoch": 5.1291626888302355,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026974902432364727,
      "loss": 4.999,
      "mean_token_accuracy": 0.22755587846040726,
      "num_tokens": 101952108.0,
      "step": 45670
    },
    {
      "entropy": 5.774025297164917,
      "epoch": 5.129724265738193,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026970888662336263,
      "loss": 4.9813,
      "mean_token_accuracy": 0.22875573635101318,
      "num_tokens": 101963617.0,
      "step": 45675
    },
    {
      "entropy": 5.72850079536438,
      "epoch": 5.130285842646151,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002696687490915491,
      "loss": 4.902,
      "mean_token_accuracy": 0.24225885272026063,
      "num_tokens": 101975565.0,
      "step": 45680
    },
    {
      "entropy": 5.672870445251465,
      "epoch": 5.130847419554108,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026962861172948456,
      "loss": 4.9143,
      "mean_token_accuracy": 0.24102582782506943,
      "num_tokens": 101988048.0,
      "step": 45685
    },
    {
      "entropy": 5.640239143371582,
      "epoch": 5.131408996462065,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026958847453844713,
      "loss": 4.8572,
      "mean_token_accuracy": 0.2328009083867073,
      "num_tokens": 101999045.0,
      "step": 45690
    },
    {
      "entropy": 5.7465081214904785,
      "epoch": 5.131970573370023,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026954833751971486,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22609367370605468,
      "num_tokens": 102009679.0,
      "step": 45695
    },
    {
      "entropy": 5.7038599967956545,
      "epoch": 5.132532150277981,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002695082006745656,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22934699505567552,
      "num_tokens": 102021378.0,
      "step": 45700
    },
    {
      "entropy": 5.740504169464112,
      "epoch": 5.133093727185938,
      "grad_norm": 1.25,
      "learning_rate": 0.0002694680640042774,
      "loss": 4.9158,
      "mean_token_accuracy": 0.23946751654148102,
      "num_tokens": 102031037.0,
      "step": 45705
    },
    {
      "entropy": 5.7120380878448485,
      "epoch": 5.133655304093896,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026942792751012814,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23439816981554032,
      "num_tokens": 102041767.0,
      "step": 45710
    },
    {
      "entropy": 5.669504594802857,
      "epoch": 5.134216881001853,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026938779119339585,
      "loss": 4.893,
      "mean_token_accuracy": 0.23526906073093415,
      "num_tokens": 102052792.0,
      "step": 45715
    },
    {
      "entropy": 5.70314130783081,
      "epoch": 5.13477845790981,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002693476550553584,
      "loss": 4.8744,
      "mean_token_accuracy": 0.24168023020029067,
      "num_tokens": 102064238.0,
      "step": 45720
    },
    {
      "entropy": 5.723247528076172,
      "epoch": 5.1353400348177685,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026930751909729384,
      "loss": 4.9175,
      "mean_token_accuracy": 0.23610434532165528,
      "num_tokens": 102075557.0,
      "step": 45725
    },
    {
      "entropy": 5.649086761474609,
      "epoch": 5.135901611725726,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026926738332048013,
      "loss": 4.8202,
      "mean_token_accuracy": 0.24382200986146926,
      "num_tokens": 102086691.0,
      "step": 45730
    },
    {
      "entropy": 5.689279556274414,
      "epoch": 5.136463188633684,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026922724772619516,
      "loss": 4.8697,
      "mean_token_accuracy": 0.23845753073692322,
      "num_tokens": 102097215.0,
      "step": 45735
    },
    {
      "entropy": 5.71323971748352,
      "epoch": 5.137024765541641,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002691871123157168,
      "loss": 4.9375,
      "mean_token_accuracy": 0.2304776668548584,
      "num_tokens": 102108408.0,
      "step": 45740
    },
    {
      "entropy": 5.749732160568238,
      "epoch": 5.137586342449598,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002691469770903231,
      "loss": 4.9552,
      "mean_token_accuracy": 0.23617395162582397,
      "num_tokens": 102120594.0,
      "step": 45745
    },
    {
      "entropy": 5.694986629486084,
      "epoch": 5.138147919357556,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002691068420512919,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22834033966064454,
      "num_tokens": 102132306.0,
      "step": 45750
    },
    {
      "entropy": 5.739891195297242,
      "epoch": 5.1387094962655135,
      "grad_norm": 1.234375,
      "learning_rate": 0.000269066707199901,
      "loss": 4.9716,
      "mean_token_accuracy": 0.23433895111083985,
      "num_tokens": 102143271.0,
      "step": 45755
    },
    {
      "entropy": 5.696656084060669,
      "epoch": 5.139271073173471,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026902657253742855,
      "loss": 4.9564,
      "mean_token_accuracy": 0.2330433577299118,
      "num_tokens": 102154657.0,
      "step": 45760
    },
    {
      "entropy": 5.704431056976318,
      "epoch": 5.139832650081429,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002689864380651524,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2377222180366516,
      "num_tokens": 102167135.0,
      "step": 45765
    },
    {
      "entropy": 5.632419490814209,
      "epoch": 5.140394226989386,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002689463037843504,
      "loss": 4.8196,
      "mean_token_accuracy": 0.24302261918783188,
      "num_tokens": 102178485.0,
      "step": 45770
    },
    {
      "entropy": 5.695957517623901,
      "epoch": 5.140955803897343,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002689061696963004,
      "loss": 4.9347,
      "mean_token_accuracy": 0.23294197469949723,
      "num_tokens": 102189378.0,
      "step": 45775
    },
    {
      "entropy": 5.689749765396118,
      "epoch": 5.141517380805301,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002688660358022803,
      "loss": 4.9325,
      "mean_token_accuracy": 0.23481471389532088,
      "num_tokens": 102199864.0,
      "step": 45780
    },
    {
      "entropy": 5.626828718185425,
      "epoch": 5.142078957713259,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000268825902103568,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23464129865169525,
      "num_tokens": 102210699.0,
      "step": 45785
    },
    {
      "entropy": 5.725682306289673,
      "epoch": 5.142640534621217,
      "grad_norm": 1.25,
      "learning_rate": 0.0002687857686014414,
      "loss": 4.9578,
      "mean_token_accuracy": 0.23494011759757996,
      "num_tokens": 102221724.0,
      "step": 45790
    },
    {
      "entropy": 5.650874757766724,
      "epoch": 5.143202111529174,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002687456352971783,
      "loss": 4.7154,
      "mean_token_accuracy": 0.2514701008796692,
      "num_tokens": 102232023.0,
      "step": 45795
    },
    {
      "entropy": 5.653011178970337,
      "epoch": 5.143763688437131,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002687055021920566,
      "loss": 4.8601,
      "mean_token_accuracy": 0.24083492457866668,
      "num_tokens": 102242526.0,
      "step": 45800
    },
    {
      "entropy": 5.618646097183228,
      "epoch": 5.144325265345089,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002686653692873542,
      "loss": 4.9035,
      "mean_token_accuracy": 0.23321566879749298,
      "num_tokens": 102253271.0,
      "step": 45805
    },
    {
      "entropy": 5.725617599487305,
      "epoch": 5.1448868422530465,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00026862523658434883,
      "loss": 4.9226,
      "mean_token_accuracy": 0.23556668013334275,
      "num_tokens": 102263089.0,
      "step": 45810
    },
    {
      "entropy": 5.6714122772216795,
      "epoch": 5.145448419161004,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002685851040843185,
      "loss": 4.8885,
      "mean_token_accuracy": 0.24039302468299867,
      "num_tokens": 102273312.0,
      "step": 45815
    },
    {
      "entropy": 5.740534782409668,
      "epoch": 5.146009996068962,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026854497178854087,
      "loss": 4.9576,
      "mean_token_accuracy": 0.23082481771707536,
      "num_tokens": 102283940.0,
      "step": 45820
    },
    {
      "entropy": 5.771324825286865,
      "epoch": 5.146571572976919,
      "grad_norm": 1.25,
      "learning_rate": 0.0002685048396982939,
      "loss": 4.9219,
      "mean_token_accuracy": 0.23427843898534775,
      "num_tokens": 102295666.0,
      "step": 45825
    },
    {
      "entropy": 5.698532152175903,
      "epoch": 5.147133149884877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026846470781485517,
      "loss": 4.9151,
      "mean_token_accuracy": 0.23355509340763092,
      "num_tokens": 102306463.0,
      "step": 45830
    },
    {
      "entropy": 5.654435110092163,
      "epoch": 5.147694726792834,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002684245761395028,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23945495188236238,
      "num_tokens": 102317443.0,
      "step": 45835
    },
    {
      "entropy": 5.708714962005615,
      "epoch": 5.148256303700792,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026838444467351447,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2402994766831398,
      "num_tokens": 102327924.0,
      "step": 45840
    },
    {
      "entropy": 5.8164963722229,
      "epoch": 5.14881788060875,
      "grad_norm": 1.125,
      "learning_rate": 0.00026834431341816805,
      "loss": 5.0149,
      "mean_token_accuracy": 0.226939357817173,
      "num_tokens": 102339364.0,
      "step": 45845
    },
    {
      "entropy": 5.735497665405274,
      "epoch": 5.149379457516707,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026830418237474126,
      "loss": 4.9632,
      "mean_token_accuracy": 0.23042761534452438,
      "num_tokens": 102350947.0,
      "step": 45850
    },
    {
      "entropy": 5.769375371932983,
      "epoch": 5.149941034424664,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002682640515445118,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2285381332039833,
      "num_tokens": 102362412.0,
      "step": 45855
    },
    {
      "entropy": 5.709850549697876,
      "epoch": 5.150502611332622,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002682239209287577,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23365160673856736,
      "num_tokens": 102373630.0,
      "step": 45860
    },
    {
      "entropy": 5.653848314285279,
      "epoch": 5.151064188240579,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026818379052875646,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22755072116851807,
      "num_tokens": 102385081.0,
      "step": 45865
    },
    {
      "entropy": 5.693227529525757,
      "epoch": 5.1516257651485375,
      "grad_norm": 1.140625,
      "learning_rate": 0.000268143660345786,
      "loss": 4.8008,
      "mean_token_accuracy": 0.24023230820894242,
      "num_tokens": 102396519.0,
      "step": 45870
    },
    {
      "entropy": 5.727885961532593,
      "epoch": 5.152187342056495,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002681035303811241,
      "loss": 4.9551,
      "mean_token_accuracy": 0.22875937670469285,
      "num_tokens": 102406326.0,
      "step": 45875
    },
    {
      "entropy": 5.7949724197387695,
      "epoch": 5.152748918964452,
      "grad_norm": 1.359375,
      "learning_rate": 0.00026806340063604845,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2268880397081375,
      "num_tokens": 102416965.0,
      "step": 45880
    },
    {
      "entropy": 5.703940200805664,
      "epoch": 5.15331049587241,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026802327111183686,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23290757089853287,
      "num_tokens": 102428414.0,
      "step": 45885
    },
    {
      "entropy": 5.693345785140991,
      "epoch": 5.153872072780367,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026798314180976696,
      "loss": 4.8635,
      "mean_token_accuracy": 0.24075448215007783,
      "num_tokens": 102438665.0,
      "step": 45890
    },
    {
      "entropy": 5.730081415176391,
      "epoch": 5.1544336496883245,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026794301273111654,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22911631762981416,
      "num_tokens": 102449828.0,
      "step": 45895
    },
    {
      "entropy": 5.750223445892334,
      "epoch": 5.154995226596283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002679028838771633,
      "loss": 4.9455,
      "mean_token_accuracy": 0.23472265750169755,
      "num_tokens": 102461007.0,
      "step": 45900
    },
    {
      "entropy": 5.66032042503357,
      "epoch": 5.15555680350424,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026786275524918503,
      "loss": 4.8025,
      "mean_token_accuracy": 0.24397041499614716,
      "num_tokens": 102472372.0,
      "step": 45905
    },
    {
      "entropy": 5.651403379440308,
      "epoch": 5.156118380412197,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002678226268484594,
      "loss": 4.7915,
      "mean_token_accuracy": 0.23449200987815857,
      "num_tokens": 102482361.0,
      "step": 45910
    },
    {
      "entropy": 5.578908491134643,
      "epoch": 5.156679957320155,
      "grad_norm": 1.25,
      "learning_rate": 0.0002677824986762641,
      "loss": 4.7589,
      "mean_token_accuracy": 0.23906071186065675,
      "num_tokens": 102493058.0,
      "step": 45915
    },
    {
      "entropy": 5.751864719390869,
      "epoch": 5.157241534228112,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002677423707338769,
      "loss": 4.9759,
      "mean_token_accuracy": 0.23615313768386842,
      "num_tokens": 102505242.0,
      "step": 45920
    },
    {
      "entropy": 5.757642412185669,
      "epoch": 5.1578031111360705,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002677022430225754,
      "loss": 4.8849,
      "mean_token_accuracy": 0.24142929017543793,
      "num_tokens": 102517580.0,
      "step": 45925
    },
    {
      "entropy": 5.769930219650268,
      "epoch": 5.158364688044028,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026766211554363727,
      "loss": 4.9635,
      "mean_token_accuracy": 0.23166800141334534,
      "num_tokens": 102527336.0,
      "step": 45930
    },
    {
      "entropy": 5.717424440383911,
      "epoch": 5.158926264951985,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002676219882983403,
      "loss": 4.9993,
      "mean_token_accuracy": 0.22449183464050293,
      "num_tokens": 102538511.0,
      "step": 45935
    },
    {
      "entropy": 5.751841878890991,
      "epoch": 5.159487841859943,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026758186128796204,
      "loss": 4.8981,
      "mean_token_accuracy": 0.23569625169038771,
      "num_tokens": 102549981.0,
      "step": 45940
    },
    {
      "entropy": 5.807143878936768,
      "epoch": 5.1600494187679,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026754173451378014,
      "loss": 4.995,
      "mean_token_accuracy": 0.23079929798841475,
      "num_tokens": 102560504.0,
      "step": 45945
    },
    {
      "entropy": 5.74498724937439,
      "epoch": 5.1606109956758575,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026750160797707245,
      "loss": 4.9516,
      "mean_token_accuracy": 0.23630113750696183,
      "num_tokens": 102571136.0,
      "step": 45950
    },
    {
      "entropy": 5.718779373168945,
      "epoch": 5.161172572583816,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002674614816791164,
      "loss": 4.8726,
      "mean_token_accuracy": 0.242006254196167,
      "num_tokens": 102582879.0,
      "step": 45955
    },
    {
      "entropy": 5.749547290802002,
      "epoch": 5.161734149491773,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026742135562118967,
      "loss": 4.9322,
      "mean_token_accuracy": 0.23588382750749587,
      "num_tokens": 102593171.0,
      "step": 45960
    },
    {
      "entropy": 5.78606104850769,
      "epoch": 5.16229572639973,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0002673812298045699,
      "loss": 5.0172,
      "mean_token_accuracy": 0.2300393044948578,
      "num_tokens": 102603816.0,
      "step": 45965
    },
    {
      "entropy": 5.723579454421997,
      "epoch": 5.162857303307688,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026734110423053483,
      "loss": 4.8968,
      "mean_token_accuracy": 0.2374560996890068,
      "num_tokens": 102614360.0,
      "step": 45970
    },
    {
      "entropy": 5.7748924732208256,
      "epoch": 5.163418880215645,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026730097890036175,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2248435452580452,
      "num_tokens": 102625320.0,
      "step": 45975
    },
    {
      "entropy": 5.706417894363403,
      "epoch": 5.163980457123603,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026726085381532867,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23482818603515626,
      "num_tokens": 102635728.0,
      "step": 45980
    },
    {
      "entropy": 5.6320586681365965,
      "epoch": 5.164542034031561,
      "grad_norm": 1.3125,
      "learning_rate": 0.000267220728976713,
      "loss": 4.802,
      "mean_token_accuracy": 0.23265672624111175,
      "num_tokens": 102645880.0,
      "step": 45985
    },
    {
      "entropy": 5.684435844421387,
      "epoch": 5.165103610939518,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002671806043857923,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22885632067918776,
      "num_tokens": 102657854.0,
      "step": 45990
    },
    {
      "entropy": 5.750489377975464,
      "epoch": 5.165665187847476,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002671404800438442,
      "loss": 4.9401,
      "mean_token_accuracy": 0.23412407785654069,
      "num_tokens": 102669637.0,
      "step": 45995
    },
    {
      "entropy": 5.747105121612549,
      "epoch": 5.166226764755433,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002671003559521463,
      "loss": 4.8312,
      "mean_token_accuracy": 0.24346354603767395,
      "num_tokens": 102680520.0,
      "step": 46000
    },
    {
      "entropy": 5.674376678466797,
      "epoch": 5.16678834166339,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026706023211197613,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23044788986444473,
      "num_tokens": 102691633.0,
      "step": 46005
    },
    {
      "entropy": 5.7172308444976805,
      "epoch": 5.1673499185713485,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002670201085246111,
      "loss": 4.9414,
      "mean_token_accuracy": 0.22915240824222566,
      "num_tokens": 102703171.0,
      "step": 46010
    },
    {
      "entropy": 5.735824346542358,
      "epoch": 5.167911495479306,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026697998519132905,
      "loss": 4.933,
      "mean_token_accuracy": 0.2379453808069229,
      "num_tokens": 102714636.0,
      "step": 46015
    },
    {
      "entropy": 5.719607019424439,
      "epoch": 5.168473072387264,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026693986211340736,
      "loss": 5.0576,
      "mean_token_accuracy": 0.23332138508558273,
      "num_tokens": 102725554.0,
      "step": 46020
    },
    {
      "entropy": 5.704901504516601,
      "epoch": 5.169034649295221,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026689973929212363,
      "loss": 4.9715,
      "mean_token_accuracy": 0.23012625724077224,
      "num_tokens": 102736705.0,
      "step": 46025
    },
    {
      "entropy": 5.819681644439697,
      "epoch": 5.169596226203178,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026685961672875524,
      "loss": 5.0158,
      "mean_token_accuracy": 0.22361911535263063,
      "num_tokens": 102749187.0,
      "step": 46030
    },
    {
      "entropy": 5.720456266403199,
      "epoch": 5.170157803111136,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002668194944245799,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22729996740818023,
      "num_tokens": 102760095.0,
      "step": 46035
    },
    {
      "entropy": 5.730439376831055,
      "epoch": 5.170719380019094,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000266779372380875,
      "loss": 4.9181,
      "mean_token_accuracy": 0.22780706584453583,
      "num_tokens": 102771086.0,
      "step": 46040
    },
    {
      "entropy": 5.745031070709229,
      "epoch": 5.171280956927051,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000266739250598918,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22709188163280486,
      "num_tokens": 102782024.0,
      "step": 46045
    },
    {
      "entropy": 5.808866882324219,
      "epoch": 5.171842533835009,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026669912907998657,
      "loss": 4.9779,
      "mean_token_accuracy": 0.23083736896514892,
      "num_tokens": 102793654.0,
      "step": 46050
    },
    {
      "entropy": 5.817464542388916,
      "epoch": 5.172404110742966,
      "grad_norm": 1.3125,
      "learning_rate": 0.000266659007825358,
      "loss": 5.0579,
      "mean_token_accuracy": 0.22906692922115326,
      "num_tokens": 102804908.0,
      "step": 46055
    },
    {
      "entropy": 5.6171668529510494,
      "epoch": 5.172965687650924,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026661888683631,
      "loss": 4.8243,
      "mean_token_accuracy": 0.240178944170475,
      "num_tokens": 102815399.0,
      "step": 46060
    },
    {
      "entropy": 5.651945018768311,
      "epoch": 5.1735272645588815,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026657876611411974,
      "loss": 4.8384,
      "mean_token_accuracy": 0.23956994116306304,
      "num_tokens": 102825734.0,
      "step": 46065
    },
    {
      "entropy": 5.699292325973511,
      "epoch": 5.174088841466839,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026653864566006494,
      "loss": 4.7462,
      "mean_token_accuracy": 0.24517602026462554,
      "num_tokens": 102836251.0,
      "step": 46070
    },
    {
      "entropy": 5.7649940013885494,
      "epoch": 5.174650418374797,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002664985254754229,
      "loss": 5.018,
      "mean_token_accuracy": 0.2187100514769554,
      "num_tokens": 102847626.0,
      "step": 46075
    },
    {
      "entropy": 5.7176268100738525,
      "epoch": 5.175211995282754,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026645840556147103,
      "loss": 4.8309,
      "mean_token_accuracy": 0.24229300171136856,
      "num_tokens": 102859424.0,
      "step": 46080
    },
    {
      "entropy": 5.721153116226196,
      "epoch": 5.175773572190711,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002664182859194869,
      "loss": 4.8847,
      "mean_token_accuracy": 0.23721171915531158,
      "num_tokens": 102870466.0,
      "step": 46085
    },
    {
      "entropy": 5.698514366149903,
      "epoch": 5.176335149098669,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002663781665507479,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22639879286289216,
      "num_tokens": 102882181.0,
      "step": 46090
    },
    {
      "entropy": 5.672818183898926,
      "epoch": 5.1768967260066265,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026633804745653146,
      "loss": 4.9447,
      "mean_token_accuracy": 0.2358270615339279,
      "num_tokens": 102892562.0,
      "step": 46095
    },
    {
      "entropy": 5.625584268569947,
      "epoch": 5.177458302914584,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026629792863811484,
      "loss": 4.8187,
      "mean_token_accuracy": 0.24426167458295822,
      "num_tokens": 102904380.0,
      "step": 46100
    },
    {
      "entropy": 5.629299688339233,
      "epoch": 5.178019879822542,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002662578100967756,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22840608805418014,
      "num_tokens": 102916018.0,
      "step": 46105
    },
    {
      "entropy": 5.712538719177246,
      "epoch": 5.178581456730499,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002662176918337911,
      "loss": 4.9499,
      "mean_token_accuracy": 0.2222796380519867,
      "num_tokens": 102927885.0,
      "step": 46110
    },
    {
      "entropy": 5.699170398712158,
      "epoch": 5.179143033638457,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002661775738504387,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23958997130393983,
      "num_tokens": 102938343.0,
      "step": 46115
    },
    {
      "entropy": 5.655062246322632,
      "epoch": 5.179704610546414,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026613745614799574,
      "loss": 4.8972,
      "mean_token_accuracy": 0.23282146453857422,
      "num_tokens": 102949341.0,
      "step": 46120
    },
    {
      "entropy": 5.760246467590332,
      "epoch": 5.180266187454372,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002660973387277396,
      "loss": 4.9448,
      "mean_token_accuracy": 0.2400175988674164,
      "num_tokens": 102959722.0,
      "step": 46125
    },
    {
      "entropy": 5.769207096099853,
      "epoch": 5.18082776436233,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002660572215909477,
      "loss": 4.9522,
      "mean_token_accuracy": 0.23184603303670884,
      "num_tokens": 102971423.0,
      "step": 46130
    },
    {
      "entropy": 5.706975030899048,
      "epoch": 5.181389341270287,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002660171047388973,
      "loss": 4.899,
      "mean_token_accuracy": 0.2441621169447899,
      "num_tokens": 102983218.0,
      "step": 46135
    },
    {
      "entropy": 5.730048751831054,
      "epoch": 5.181950918178244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026597698817286576,
      "loss": 4.9294,
      "mean_token_accuracy": 0.23083430975675584,
      "num_tokens": 102995287.0,
      "step": 46140
    },
    {
      "entropy": 5.691871118545532,
      "epoch": 5.182512495086202,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026593687189413044,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2377895936369896,
      "num_tokens": 103006578.0,
      "step": 46145
    },
    {
      "entropy": 5.760332870483398,
      "epoch": 5.1830740719941595,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026589675590396865,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22562014311552048,
      "num_tokens": 103017959.0,
      "step": 46150
    },
    {
      "entropy": 5.742167234420776,
      "epoch": 5.183635648902117,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002658566402036576,
      "loss": 4.9202,
      "mean_token_accuracy": 0.23148497194051743,
      "num_tokens": 103029931.0,
      "step": 46155
    },
    {
      "entropy": 5.707079935073852,
      "epoch": 5.184197225810075,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002658165247944747,
      "loss": 4.8371,
      "mean_token_accuracy": 0.24063932597637178,
      "num_tokens": 103041603.0,
      "step": 46160
    },
    {
      "entropy": 5.656423330307007,
      "epoch": 5.184758802718032,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026577640967769727,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23392304480075837,
      "num_tokens": 103053353.0,
      "step": 46165
    },
    {
      "entropy": 5.621143484115601,
      "epoch": 5.18532037962599,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026573629485460257,
      "loss": 4.8019,
      "mean_token_accuracy": 0.24093992710113527,
      "num_tokens": 103064295.0,
      "step": 46170
    },
    {
      "entropy": 5.698610591888428,
      "epoch": 5.185881956533947,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026569618032646773,
      "loss": 4.9621,
      "mean_token_accuracy": 0.23703776597976683,
      "num_tokens": 103075582.0,
      "step": 46175
    },
    {
      "entropy": 5.674321031570434,
      "epoch": 5.186443533441905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026565606609457016,
      "loss": 4.911,
      "mean_token_accuracy": 0.23035358488559723,
      "num_tokens": 103087969.0,
      "step": 46180
    },
    {
      "entropy": 5.723114013671875,
      "epoch": 5.187005110349863,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002656159521601871,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22953645884990692,
      "num_tokens": 103099505.0,
      "step": 46185
    },
    {
      "entropy": 5.777262115478516,
      "epoch": 5.18756668725782,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002655758385245956,
      "loss": 4.9651,
      "mean_token_accuracy": 0.22979339957237244,
      "num_tokens": 103110706.0,
      "step": 46190
    },
    {
      "entropy": 5.6778120517730715,
      "epoch": 5.188128264165777,
      "grad_norm": 1.25,
      "learning_rate": 0.0002655357251890732,
      "loss": 4.823,
      "mean_token_accuracy": 0.24173504263162612,
      "num_tokens": 103121129.0,
      "step": 46195
    },
    {
      "entropy": 5.653455066680908,
      "epoch": 5.188689841073735,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026549561215489695,
      "loss": 4.8954,
      "mean_token_accuracy": 0.2305653616786003,
      "num_tokens": 103130706.0,
      "step": 46200
    },
    {
      "entropy": 5.720905017852783,
      "epoch": 5.189251417981692,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002654554994233442,
      "loss": 4.9646,
      "mean_token_accuracy": 0.22789174020290376,
      "num_tokens": 103140441.0,
      "step": 46205
    },
    {
      "entropy": 5.615092182159424,
      "epoch": 5.1898129948896505,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002654153869956919,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23618113100528718,
      "num_tokens": 103151245.0,
      "step": 46210
    },
    {
      "entropy": 5.706813001632691,
      "epoch": 5.190374571797608,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002653752748732175,
      "loss": 4.9021,
      "mean_token_accuracy": 0.23921293169260024,
      "num_tokens": 103161985.0,
      "step": 46215
    },
    {
      "entropy": 5.745518922805786,
      "epoch": 5.190936148705565,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000265335163057198,
      "loss": 4.9303,
      "mean_token_accuracy": 0.23609597235918045,
      "num_tokens": 103174471.0,
      "step": 46220
    },
    {
      "entropy": 5.683405590057373,
      "epoch": 5.191497725613523,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002652950515489107,
      "loss": 4.874,
      "mean_token_accuracy": 0.23864724934101106,
      "num_tokens": 103184853.0,
      "step": 46225
    },
    {
      "entropy": 5.760870409011841,
      "epoch": 5.19205930252148,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002652549403496327,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2261437028646469,
      "num_tokens": 103195909.0,
      "step": 46230
    },
    {
      "entropy": 5.776816415786743,
      "epoch": 5.1926208794294375,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026521482946064123,
      "loss": 4.984,
      "mean_token_accuracy": 0.2298436388373375,
      "num_tokens": 103206557.0,
      "step": 46235
    },
    {
      "entropy": 5.659218406677246,
      "epoch": 5.193182456337396,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002651747188832134,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2375745192170143,
      "num_tokens": 103216887.0,
      "step": 46240
    },
    {
      "entropy": 5.695422887802124,
      "epoch": 5.193744033245353,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002651346086186262,
      "loss": 4.9318,
      "mean_token_accuracy": 0.23443031162023545,
      "num_tokens": 103228879.0,
      "step": 46245
    },
    {
      "entropy": 5.683432579040527,
      "epoch": 5.194305610153311,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000265094498668157,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22958140820264816,
      "num_tokens": 103240564.0,
      "step": 46250
    },
    {
      "entropy": 5.678846120834351,
      "epoch": 5.194867187061268,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026505438903308285,
      "loss": 4.879,
      "mean_token_accuracy": 0.23555289655923844,
      "num_tokens": 103251297.0,
      "step": 46255
    },
    {
      "entropy": 5.712587690353393,
      "epoch": 5.195428763969225,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026501427971468056,
      "loss": 4.8532,
      "mean_token_accuracy": 0.24277032613754274,
      "num_tokens": 103262041.0,
      "step": 46260
    },
    {
      "entropy": 5.680443048477173,
      "epoch": 5.1959903408771835,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002649741707142277,
      "loss": 4.8607,
      "mean_token_accuracy": 0.24280264973640442,
      "num_tokens": 103273024.0,
      "step": 46265
    },
    {
      "entropy": 5.713138484954834,
      "epoch": 5.196551917785141,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026493406203300106,
      "loss": 4.9024,
      "mean_token_accuracy": 0.23265138119459153,
      "num_tokens": 103284511.0,
      "step": 46270
    },
    {
      "entropy": 5.746639680862427,
      "epoch": 5.197113494693098,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026489395367227785,
      "loss": 4.9129,
      "mean_token_accuracy": 0.2336493507027626,
      "num_tokens": 103296787.0,
      "step": 46275
    },
    {
      "entropy": 5.735622262954712,
      "epoch": 5.197675071601056,
      "grad_norm": 1.203125,
      "learning_rate": 0.000264853845633335,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2378787189722061,
      "num_tokens": 103308422.0,
      "step": 46280
    },
    {
      "entropy": 5.657946825027466,
      "epoch": 5.198236648509013,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002648137379174497,
      "loss": 4.8361,
      "mean_token_accuracy": 0.23600769340991973,
      "num_tokens": 103319592.0,
      "step": 46285
    },
    {
      "entropy": 5.677166509628296,
      "epoch": 5.1987982254169705,
      "grad_norm": 1.125,
      "learning_rate": 0.0002647736305258989,
      "loss": 4.8644,
      "mean_token_accuracy": 0.2351193457841873,
      "num_tokens": 103330702.0,
      "step": 46290
    },
    {
      "entropy": 5.635950136184692,
      "epoch": 5.199359802324929,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002647335234599597,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23477444648742676,
      "num_tokens": 103342339.0,
      "step": 46295
    },
    {
      "entropy": 5.643259477615357,
      "epoch": 5.199921379232886,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026469341672090906,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2397024154663086,
      "num_tokens": 103353604.0,
      "step": 46300
    },
    {
      "entropy": 5.6850964546203615,
      "epoch": 5.200482956140844,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026465331031002404,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2385489210486412,
      "num_tokens": 103363713.0,
      "step": 46305
    },
    {
      "entropy": 5.782031488418579,
      "epoch": 5.201044533048801,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00026461320422858166,
      "loss": 4.9657,
      "mean_token_accuracy": 0.2269342765212059,
      "num_tokens": 103374966.0,
      "step": 46310
    },
    {
      "entropy": 5.721109867095947,
      "epoch": 5.201606109956758,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002645730984778589,
      "loss": 4.9137,
      "mean_token_accuracy": 0.23176535069942475,
      "num_tokens": 103387025.0,
      "step": 46315
    },
    {
      "entropy": 5.771420812606811,
      "epoch": 5.202167686864716,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002645329930591327,
      "loss": 4.953,
      "mean_token_accuracy": 0.22407152354717255,
      "num_tokens": 103397947.0,
      "step": 46320
    },
    {
      "entropy": 5.7260191440582275,
      "epoch": 5.202729263772674,
      "grad_norm": 1.296875,
      "learning_rate": 0.00026449288797368006,
      "loss": 4.9532,
      "mean_token_accuracy": 0.23368007838726043,
      "num_tokens": 103409874.0,
      "step": 46325
    },
    {
      "entropy": 5.658824872970581,
      "epoch": 5.203290840680631,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026445278322277796,
      "loss": 4.9382,
      "mean_token_accuracy": 0.2259092003107071,
      "num_tokens": 103423484.0,
      "step": 46330
    },
    {
      "entropy": 5.713621664047241,
      "epoch": 5.203852417588589,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002644126788077032,
      "loss": 4.8725,
      "mean_token_accuracy": 0.2351704254746437,
      "num_tokens": 103434933.0,
      "step": 46335
    },
    {
      "entropy": 5.743420839309692,
      "epoch": 5.204413994496546,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026437257472973296,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2270999163389206,
      "num_tokens": 103447111.0,
      "step": 46340
    },
    {
      "entropy": 5.706616497039795,
      "epoch": 5.204975571404503,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026433247099014406,
      "loss": 4.8331,
      "mean_token_accuracy": 0.24988146126270294,
      "num_tokens": 103457627.0,
      "step": 46345
    },
    {
      "entropy": 5.700667858123779,
      "epoch": 5.2055371483124615,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002642923675902135,
      "loss": 4.9045,
      "mean_token_accuracy": 0.2365787297487259,
      "num_tokens": 103468047.0,
      "step": 46350
    },
    {
      "entropy": 5.637009525299073,
      "epoch": 5.206098725220419,
      "grad_norm": 1.203125,
      "learning_rate": 0.000264252264531218,
      "loss": 4.842,
      "mean_token_accuracy": 0.23366844207048415,
      "num_tokens": 103479671.0,
      "step": 46355
    },
    {
      "entropy": 5.646584749221802,
      "epoch": 5.206660302128377,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026421216181443457,
      "loss": 4.8541,
      "mean_token_accuracy": 0.23232119530439377,
      "num_tokens": 103490565.0,
      "step": 46360
    },
    {
      "entropy": 5.688727998733521,
      "epoch": 5.207221879036334,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002641720594411401,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2377782255411148,
      "num_tokens": 103500915.0,
      "step": 46365
    },
    {
      "entropy": 5.7218985080719,
      "epoch": 5.207783455944291,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026413195741261145,
      "loss": 4.9259,
      "mean_token_accuracy": 0.23770544081926345,
      "num_tokens": 103511739.0,
      "step": 46370
    },
    {
      "entropy": 5.751352739334107,
      "epoch": 5.208345032852249,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002640918557301256,
      "loss": 4.9787,
      "mean_token_accuracy": 0.22600287646055223,
      "num_tokens": 103521957.0,
      "step": 46375
    },
    {
      "entropy": 5.70605959892273,
      "epoch": 5.208906609760207,
      "grad_norm": 1.25,
      "learning_rate": 0.0002640517543949592,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2341214746236801,
      "num_tokens": 103532511.0,
      "step": 46380
    },
    {
      "entropy": 5.725134563446045,
      "epoch": 5.209468186668164,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002640116534083892,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2365953266620636,
      "num_tokens": 103544484.0,
      "step": 46385
    },
    {
      "entropy": 5.733850908279419,
      "epoch": 5.210029763576122,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026397155277169246,
      "loss": 4.8941,
      "mean_token_accuracy": 0.23364219218492507,
      "num_tokens": 103555020.0,
      "step": 46390
    },
    {
      "entropy": 5.700289916992188,
      "epoch": 5.210591340484079,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002639314524861457,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2340398371219635,
      "num_tokens": 103565331.0,
      "step": 46395
    },
    {
      "entropy": 5.736643362045288,
      "epoch": 5.211152917392037,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002638913525530258,
      "loss": 4.9431,
      "mean_token_accuracy": 0.23898895382881163,
      "num_tokens": 103576970.0,
      "step": 46400
    },
    {
      "entropy": 5.723450422286987,
      "epoch": 5.2117144942999944,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026385125297360946,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2272755205631256,
      "num_tokens": 103587981.0,
      "step": 46405
    },
    {
      "entropy": 5.726024627685547,
      "epoch": 5.212276071207952,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00026381115374917363,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23687417954206466,
      "num_tokens": 103599508.0,
      "step": 46410
    },
    {
      "entropy": 5.705495548248291,
      "epoch": 5.21283764811591,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026377105488099503,
      "loss": 4.8646,
      "mean_token_accuracy": 0.23149629235267638,
      "num_tokens": 103609982.0,
      "step": 46415
    },
    {
      "entropy": 5.6680608749389645,
      "epoch": 5.213399225023867,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026373095637035037,
      "loss": 4.858,
      "mean_token_accuracy": 0.24121971875429155,
      "num_tokens": 103620419.0,
      "step": 46420
    },
    {
      "entropy": 5.654499340057373,
      "epoch": 5.213960801931824,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026369085821851645,
      "loss": 4.8366,
      "mean_token_accuracy": 0.23781156539916992,
      "num_tokens": 103630512.0,
      "step": 46425
    },
    {
      "entropy": 5.722693252563476,
      "epoch": 5.214522378839782,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026365076042677,
      "loss": 4.9596,
      "mean_token_accuracy": 0.23885490298271178,
      "num_tokens": 103641596.0,
      "step": 46430
    },
    {
      "entropy": 5.7370086193084715,
      "epoch": 5.2150839557477395,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002636106629963877,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2319421276450157,
      "num_tokens": 103654291.0,
      "step": 46435
    },
    {
      "entropy": 5.595844411849976,
      "epoch": 5.215645532655698,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002635705659286463,
      "loss": 4.7357,
      "mean_token_accuracy": 0.24985006898641587,
      "num_tokens": 103665843.0,
      "step": 46440
    },
    {
      "entropy": 5.791073369979858,
      "epoch": 5.216207109563655,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026353046922482256,
      "loss": 5.0422,
      "mean_token_accuracy": 0.2274833843111992,
      "num_tokens": 103676682.0,
      "step": 46445
    },
    {
      "entropy": 5.7281828880310055,
      "epoch": 5.216768686471612,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002634903728861931,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22583779841661453,
      "num_tokens": 103687189.0,
      "step": 46450
    },
    {
      "entropy": 5.706825494766235,
      "epoch": 5.21733026337957,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002634502769140346,
      "loss": 4.8851,
      "mean_token_accuracy": 0.23738446682691575,
      "num_tokens": 103698332.0,
      "step": 46455
    },
    {
      "entropy": 5.70835862159729,
      "epoch": 5.217891840287527,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026341018130962384,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23658292144536971,
      "num_tokens": 103708652.0,
      "step": 46460
    },
    {
      "entropy": 5.742325782775879,
      "epoch": 5.218453417195485,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026337008607423733,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22965288460254668,
      "num_tokens": 103719275.0,
      "step": 46465
    },
    {
      "entropy": 5.693460845947266,
      "epoch": 5.219014994103443,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00026332999120915184,
      "loss": 4.8797,
      "mean_token_accuracy": 0.23737130165100098,
      "num_tokens": 103730865.0,
      "step": 46470
    },
    {
      "entropy": 5.7236639022827145,
      "epoch": 5.2195765710114,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026328989671564395,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2223804250359535,
      "num_tokens": 103743784.0,
      "step": 46475
    },
    {
      "entropy": 5.757291793823242,
      "epoch": 5.220138147919357,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002632498025949902,
      "loss": 5.0004,
      "mean_token_accuracy": 0.23383030742406846,
      "num_tokens": 103755580.0,
      "step": 46480
    },
    {
      "entropy": 5.75605845451355,
      "epoch": 5.220699724827315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002632097088484674,
      "loss": 5.0242,
      "mean_token_accuracy": 0.22613821774721146,
      "num_tokens": 103767296.0,
      "step": 46485
    },
    {
      "entropy": 5.721654462814331,
      "epoch": 5.2212613017352725,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026316961547735203,
      "loss": 4.8915,
      "mean_token_accuracy": 0.23705091178417206,
      "num_tokens": 103778176.0,
      "step": 46490
    },
    {
      "entropy": 5.789253902435303,
      "epoch": 5.221822878643231,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026312952248292067,
      "loss": 4.9905,
      "mean_token_accuracy": 0.23623390644788742,
      "num_tokens": 103791499.0,
      "step": 46495
    },
    {
      "entropy": 5.737691164016724,
      "epoch": 5.222384455551188,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002630894298664499,
      "loss": 4.9506,
      "mean_token_accuracy": 0.23704157620668412,
      "num_tokens": 103803102.0,
      "step": 46500
    },
    {
      "entropy": 5.722194004058838,
      "epoch": 5.222946032459145,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026304933762921637,
      "loss": 4.9245,
      "mean_token_accuracy": 0.23088283389806746,
      "num_tokens": 103814273.0,
      "step": 46505
    },
    {
      "entropy": 5.6595673084259035,
      "epoch": 5.223507609367103,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002630092457724965,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24300118982791902,
      "num_tokens": 103823713.0,
      "step": 46510
    },
    {
      "entropy": 5.6934239864349365,
      "epoch": 5.22406918627506,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002629691542975669,
      "loss": 4.8107,
      "mean_token_accuracy": 0.23802754431962966,
      "num_tokens": 103834244.0,
      "step": 46515
    },
    {
      "entropy": 5.599305582046509,
      "epoch": 5.224630763183018,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002629290632057041,
      "loss": 4.7457,
      "mean_token_accuracy": 0.24577538073062896,
      "num_tokens": 103844561.0,
      "step": 46520
    },
    {
      "entropy": 5.718812799453735,
      "epoch": 5.225192340090976,
      "grad_norm": 1.125,
      "learning_rate": 0.0002628889724981846,
      "loss": 4.9737,
      "mean_token_accuracy": 0.22801725417375565,
      "num_tokens": 103856812.0,
      "step": 46525
    },
    {
      "entropy": 5.7868493556976315,
      "epoch": 5.225753916998933,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002628488821762849,
      "loss": 5.01,
      "mean_token_accuracy": 0.22253558784723282,
      "num_tokens": 103870062.0,
      "step": 46530
    },
    {
      "entropy": 5.73596887588501,
      "epoch": 5.22631549390689,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026280879224128165,
      "loss": 4.9155,
      "mean_token_accuracy": 0.2284786358475685,
      "num_tokens": 103880763.0,
      "step": 46535
    },
    {
      "entropy": 5.731964254379273,
      "epoch": 5.226877070814848,
      "grad_norm": 1.1875,
      "learning_rate": 0.000262768702694451,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22889842838048935,
      "num_tokens": 103892528.0,
      "step": 46540
    },
    {
      "entropy": 5.695834398269653,
      "epoch": 5.227438647722805,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002627286135370697,
      "loss": 4.8923,
      "mean_token_accuracy": 0.23289866745471954,
      "num_tokens": 103903589.0,
      "step": 46545
    },
    {
      "entropy": 5.672017574310303,
      "epoch": 5.2280002246307635,
      "grad_norm": 1.234375,
      "learning_rate": 0.000262688524770414,
      "loss": 4.8979,
      "mean_token_accuracy": 0.23186187893152238,
      "num_tokens": 103914192.0,
      "step": 46550
    },
    {
      "entropy": 5.6944739818573,
      "epoch": 5.228561801538721,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026264843639576055,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2360879510641098,
      "num_tokens": 103924649.0,
      "step": 46555
    },
    {
      "entropy": 5.650540494918824,
      "epoch": 5.229123378446678,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002626083484143856,
      "loss": 4.8283,
      "mean_token_accuracy": 0.2368161976337433,
      "num_tokens": 103935834.0,
      "step": 46560
    },
    {
      "entropy": 5.606198120117187,
      "epoch": 5.229684955354636,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002625682608275657,
      "loss": 4.8375,
      "mean_token_accuracy": 0.2395588770508766,
      "num_tokens": 103947007.0,
      "step": 46565
    },
    {
      "entropy": 5.676851081848144,
      "epoch": 5.230246532262593,
      "grad_norm": 1.25,
      "learning_rate": 0.0002625281736365772,
      "loss": 4.9339,
      "mean_token_accuracy": 0.2319536551833153,
      "num_tokens": 103958424.0,
      "step": 46570
    },
    {
      "entropy": 5.731931352615357,
      "epoch": 5.2308081091705505,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026248808684269646,
      "loss": 4.942,
      "mean_token_accuracy": 0.23658826053142548,
      "num_tokens": 103969289.0,
      "step": 46575
    },
    {
      "entropy": 5.819073629379273,
      "epoch": 5.231369686078509,
      "grad_norm": 1.125,
      "learning_rate": 0.0002624480004472,
      "loss": 4.9763,
      "mean_token_accuracy": 0.2305781900882721,
      "num_tokens": 103981081.0,
      "step": 46580
    },
    {
      "entropy": 5.778150129318237,
      "epoch": 5.231931262986466,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026240791445136384,
      "loss": 4.964,
      "mean_token_accuracy": 0.2261814296245575,
      "num_tokens": 103990904.0,
      "step": 46585
    },
    {
      "entropy": 5.698744106292724,
      "epoch": 5.232492839894424,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002623678288564647,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2310882478952408,
      "num_tokens": 104001535.0,
      "step": 46590
    },
    {
      "entropy": 5.656168603897095,
      "epoch": 5.233054416802381,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002623277436637788,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23111891150474548,
      "num_tokens": 104012840.0,
      "step": 46595
    },
    {
      "entropy": 5.671757411956787,
      "epoch": 5.233615993710338,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002622876588745825,
      "loss": 4.8112,
      "mean_token_accuracy": 0.23951104134321213,
      "num_tokens": 104024035.0,
      "step": 46600
    },
    {
      "entropy": 5.70433874130249,
      "epoch": 5.2341775706182965,
      "grad_norm": 1.15625,
      "learning_rate": 0.000262247574490152,
      "loss": 4.863,
      "mean_token_accuracy": 0.23776088804006576,
      "num_tokens": 104035391.0,
      "step": 46605
    },
    {
      "entropy": 5.7389442920684814,
      "epoch": 5.234739147526254,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002622074905117637,
      "loss": 4.9088,
      "mean_token_accuracy": 0.23714740723371505,
      "num_tokens": 104046645.0,
      "step": 46610
    },
    {
      "entropy": 5.684802913665772,
      "epoch": 5.235300724434211,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026216740694069387,
      "loss": 4.8358,
      "mean_token_accuracy": 0.2388899579644203,
      "num_tokens": 104057706.0,
      "step": 46615
    },
    {
      "entropy": 5.684814882278443,
      "epoch": 5.235862301342169,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002621273237782188,
      "loss": 4.8904,
      "mean_token_accuracy": 0.23336266428232194,
      "num_tokens": 104068076.0,
      "step": 46620
    },
    {
      "entropy": 5.654889249801636,
      "epoch": 5.236423878250126,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002620872410256146,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22925695776939392,
      "num_tokens": 104078116.0,
      "step": 46625
    },
    {
      "entropy": 5.634376192092896,
      "epoch": 5.236985455158084,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002620471586841577,
      "loss": 4.738,
      "mean_token_accuracy": 0.24970148205757142,
      "num_tokens": 104089543.0,
      "step": 46630
    },
    {
      "entropy": 5.656867647171021,
      "epoch": 5.2375470320660416,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002620070767551244,
      "loss": 4.8565,
      "mean_token_accuracy": 0.23708769381046296,
      "num_tokens": 104100824.0,
      "step": 46635
    },
    {
      "entropy": 5.679096794128418,
      "epoch": 5.238108608973999,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002619669952397908,
      "loss": 4.9155,
      "mean_token_accuracy": 0.23604880422353744,
      "num_tokens": 104112410.0,
      "step": 46640
    },
    {
      "entropy": 5.679787731170654,
      "epoch": 5.238670185881957,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002619269141394331,
      "loss": 4.9191,
      "mean_token_accuracy": 0.23474751263856888,
      "num_tokens": 104124299.0,
      "step": 46645
    },
    {
      "entropy": 5.8295189380645756,
      "epoch": 5.239231762789914,
      "grad_norm": 1.25,
      "learning_rate": 0.0002618868334553275,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22818006873130797,
      "num_tokens": 104136266.0,
      "step": 46650
    },
    {
      "entropy": 5.737268686294556,
      "epoch": 5.239793339697871,
      "grad_norm": 1.125,
      "learning_rate": 0.00026184675318875023,
      "loss": 4.924,
      "mean_token_accuracy": 0.22579824924468994,
      "num_tokens": 104146449.0,
      "step": 46655
    },
    {
      "entropy": 5.656716918945312,
      "epoch": 5.240354916605829,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026180667334097733,
      "loss": 4.7934,
      "mean_token_accuracy": 0.2350084289908409,
      "num_tokens": 104156959.0,
      "step": 46660
    },
    {
      "entropy": 5.643295860290527,
      "epoch": 5.240916493513787,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002617665939132852,
      "loss": 4.8385,
      "mean_token_accuracy": 0.231210421025753,
      "num_tokens": 104167834.0,
      "step": 46665
    },
    {
      "entropy": 5.747311592102051,
      "epoch": 5.241478070421744,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002617265149069498,
      "loss": 4.9668,
      "mean_token_accuracy": 0.22997458130121232,
      "num_tokens": 104180035.0,
      "step": 46670
    },
    {
      "entropy": 5.831507349014283,
      "epoch": 5.242039647329702,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002616864363232474,
      "loss": 5.1,
      "mean_token_accuracy": 0.23224658966064454,
      "num_tokens": 104192960.0,
      "step": 46675
    },
    {
      "entropy": 5.767740297317505,
      "epoch": 5.242601224237659,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026164635816345395,
      "loss": 4.9492,
      "mean_token_accuracy": 0.22928874492645263,
      "num_tokens": 104204120.0,
      "step": 46680
    },
    {
      "entropy": 5.687240934371948,
      "epoch": 5.243162801145617,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002616062804288457,
      "loss": 4.8866,
      "mean_token_accuracy": 0.2338129460811615,
      "num_tokens": 104215222.0,
      "step": 46685
    },
    {
      "entropy": 5.716815567016601,
      "epoch": 5.2437243780535745,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002615662031206986,
      "loss": 4.9034,
      "mean_token_accuracy": 0.23190943896770477,
      "num_tokens": 104227208.0,
      "step": 46690
    },
    {
      "entropy": 5.756999826431274,
      "epoch": 5.244285954961532,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002615261262402887,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22833803445100784,
      "num_tokens": 104236836.0,
      "step": 46695
    },
    {
      "entropy": 5.70440444946289,
      "epoch": 5.24484753186949,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026148604978889237,
      "loss": 4.9079,
      "mean_token_accuracy": 0.23654953986406327,
      "num_tokens": 104248950.0,
      "step": 46700
    },
    {
      "entropy": 5.717470264434814,
      "epoch": 5.245409108777447,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026144597376778533,
      "loss": 4.9099,
      "mean_token_accuracy": 0.23259898275136948,
      "num_tokens": 104260179.0,
      "step": 46705
    },
    {
      "entropy": 5.722206020355225,
      "epoch": 5.245970685685404,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002614058981782438,
      "loss": 4.9922,
      "mean_token_accuracy": 0.23176783621311187,
      "num_tokens": 104273704.0,
      "step": 46710
    },
    {
      "entropy": 5.697960138320923,
      "epoch": 5.246532262593362,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026136582302154377,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23872482925653457,
      "num_tokens": 104283851.0,
      "step": 46715
    },
    {
      "entropy": 5.714468097686767,
      "epoch": 5.24709383950132,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026132574829896123,
      "loss": 4.8872,
      "mean_token_accuracy": 0.24405730664730071,
      "num_tokens": 104296064.0,
      "step": 46720
    },
    {
      "entropy": 5.774474620819092,
      "epoch": 5.247655416409277,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002612856740117721,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2223445937037468,
      "num_tokens": 104309016.0,
      "step": 46725
    },
    {
      "entropy": 5.715777349472046,
      "epoch": 5.248216993317235,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026124560016125236,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2339767888188362,
      "num_tokens": 104320598.0,
      "step": 46730
    },
    {
      "entropy": 5.781919622421265,
      "epoch": 5.248778570225192,
      "grad_norm": 1.25,
      "learning_rate": 0.0002612055267486782,
      "loss": 4.9449,
      "mean_token_accuracy": 0.22600328624248506,
      "num_tokens": 104332055.0,
      "step": 46735
    },
    {
      "entropy": 5.731698656082154,
      "epoch": 5.24934014713315,
      "grad_norm": 1.25,
      "learning_rate": 0.00026116545377532536,
      "loss": 4.9337,
      "mean_token_accuracy": 0.23557552993297576,
      "num_tokens": 104343879.0,
      "step": 46740
    },
    {
      "entropy": 5.679244136810302,
      "epoch": 5.249901724041107,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002611253812424698,
      "loss": 4.9136,
      "mean_token_accuracy": 0.23155736029148102,
      "num_tokens": 104354653.0,
      "step": 46745
    },
    {
      "entropy": 5.773394393920898,
      "epoch": 5.250463300949065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002610853091513876,
      "loss": 4.9615,
      "mean_token_accuracy": 0.23332659900188446,
      "num_tokens": 104365273.0,
      "step": 46750
    },
    {
      "entropy": 5.771021938323974,
      "epoch": 5.251024877857023,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026104523750335445,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22518273144960405,
      "num_tokens": 104375990.0,
      "step": 46755
    },
    {
      "entropy": 5.707470226287842,
      "epoch": 5.25158645476498,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002610051662996464,
      "loss": 4.8152,
      "mean_token_accuracy": 0.24070746898651124,
      "num_tokens": 104385975.0,
      "step": 46760
    },
    {
      "entropy": 5.7656293392181395,
      "epoch": 5.252148031672937,
      "grad_norm": 1.28125,
      "learning_rate": 0.00026096509554153914,
      "loss": 4.9288,
      "mean_token_accuracy": 0.23560314029455184,
      "num_tokens": 104396053.0,
      "step": 46765
    },
    {
      "entropy": 5.671411037445068,
      "epoch": 5.252709608580895,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002609250252303088,
      "loss": 4.8083,
      "mean_token_accuracy": 0.2367091953754425,
      "num_tokens": 104407874.0,
      "step": 46770
    },
    {
      "entropy": 5.68572473526001,
      "epoch": 5.2532711854888525,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002608849553672311,
      "loss": 4.893,
      "mean_token_accuracy": 0.24311309009790422,
      "num_tokens": 104418527.0,
      "step": 46775
    },
    {
      "entropy": 5.691398286819458,
      "epoch": 5.253832762396811,
      "grad_norm": 1.21875,
      "learning_rate": 0.00026084488595358197,
      "loss": 4.9148,
      "mean_token_accuracy": 0.23814602345228195,
      "num_tokens": 104430490.0,
      "step": 46780
    },
    {
      "entropy": 5.739304113388061,
      "epoch": 5.254394339304768,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002608048169906371,
      "loss": 4.8324,
      "mean_token_accuracy": 0.2397286206483841,
      "num_tokens": 104441767.0,
      "step": 46785
    },
    {
      "entropy": 5.670536470413208,
      "epoch": 5.254955916212725,
      "grad_norm": 1.234375,
      "learning_rate": 0.00026076474847967236,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23749190717935562,
      "num_tokens": 104452720.0,
      "step": 46790
    },
    {
      "entropy": 5.744123888015747,
      "epoch": 5.255517493120683,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00026072468042196356,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22940697371959687,
      "num_tokens": 104463829.0,
      "step": 46795
    },
    {
      "entropy": 5.737128210067749,
      "epoch": 5.25607907002864,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026068461281878645,
      "loss": 4.8796,
      "mean_token_accuracy": 0.23840344101190566,
      "num_tokens": 104474830.0,
      "step": 46800
    },
    {
      "entropy": 5.699039793014526,
      "epoch": 5.256640646936598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00026064454567141683,
      "loss": 4.857,
      "mean_token_accuracy": 0.2395595297217369,
      "num_tokens": 104485518.0,
      "step": 46805
    },
    {
      "entropy": 5.784093284606934,
      "epoch": 5.257202223844556,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002606044789811304,
      "loss": 4.9486,
      "mean_token_accuracy": 0.22958145141601563,
      "num_tokens": 104495562.0,
      "step": 46810
    },
    {
      "entropy": 5.720361042022705,
      "epoch": 5.257763800752513,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000260564412749203,
      "loss": 4.8629,
      "mean_token_accuracy": 0.2394092261791229,
      "num_tokens": 104506057.0,
      "step": 46815
    },
    {
      "entropy": 5.65921425819397,
      "epoch": 5.258325377660471,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026052434697691035,
      "loss": 4.865,
      "mean_token_accuracy": 0.24033822119235992,
      "num_tokens": 104518269.0,
      "step": 46820
    },
    {
      "entropy": 5.668212938308716,
      "epoch": 5.258886954568428,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000260484281665528,
      "loss": 4.835,
      "mean_token_accuracy": 0.2355603724718094,
      "num_tokens": 104528469.0,
      "step": 46825
    },
    {
      "entropy": 5.705700540542603,
      "epoch": 5.2594485314763855,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002604442168163318,
      "loss": 4.9452,
      "mean_token_accuracy": 0.22918744385242462,
      "num_tokens": 104539079.0,
      "step": 46830
    },
    {
      "entropy": 5.679761171340942,
      "epoch": 5.260010108384344,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002604041524305974,
      "loss": 5.006,
      "mean_token_accuracy": 0.23387213200330734,
      "num_tokens": 104551235.0,
      "step": 46835
    },
    {
      "entropy": 5.828889465332031,
      "epoch": 5.260571685292301,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002603640885096004,
      "loss": 5.0031,
      "mean_token_accuracy": 0.23094717413187027,
      "num_tokens": 104563911.0,
      "step": 46840
    },
    {
      "entropy": 5.730592346191406,
      "epoch": 5.261133262200258,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002603240250546165,
      "loss": 4.9418,
      "mean_token_accuracy": 0.23304537385702134,
      "num_tokens": 104574877.0,
      "step": 46845
    },
    {
      "entropy": 5.751327323913574,
      "epoch": 5.261694839108216,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00026028396206692134,
      "loss": 4.9554,
      "mean_token_accuracy": 0.2306463211774826,
      "num_tokens": 104585209.0,
      "step": 46850
    },
    {
      "entropy": 5.667649984359741,
      "epoch": 5.262256416016173,
      "grad_norm": 1.25,
      "learning_rate": 0.00026024389954779054,
      "loss": 4.8676,
      "mean_token_accuracy": 0.24120347052812577,
      "num_tokens": 104596213.0,
      "step": 46855
    },
    {
      "entropy": 5.6951391220092775,
      "epoch": 5.2628179929241306,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002602038374984997,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2338988959789276,
      "num_tokens": 104606813.0,
      "step": 46860
    },
    {
      "entropy": 5.712830591201782,
      "epoch": 5.263379569832089,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00026016377592032446,
      "loss": 4.9481,
      "mean_token_accuracy": 0.2340875133872032,
      "num_tokens": 104619040.0,
      "step": 46865
    },
    {
      "entropy": 5.567136907577515,
      "epoch": 5.263941146740046,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002601237148145403,
      "loss": 4.7577,
      "mean_token_accuracy": 0.23958449363708495,
      "num_tokens": 104629879.0,
      "step": 46870
    },
    {
      "entropy": 5.6658491611480715,
      "epoch": 5.264502723648004,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026008365418242284,
      "loss": 4.874,
      "mean_token_accuracy": 0.24240439385175705,
      "num_tokens": 104641408.0,
      "step": 46875
    },
    {
      "entropy": 5.818747186660767,
      "epoch": 5.265064300555961,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002600435940252476,
      "loss": 5.0182,
      "mean_token_accuracy": 0.22971633970737457,
      "num_tokens": 104653537.0,
      "step": 46880
    },
    {
      "entropy": 5.653452825546265,
      "epoch": 5.265625877463918,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002600035343442902,
      "loss": 4.8134,
      "mean_token_accuracy": 0.24214009791612626,
      "num_tokens": 104665499.0,
      "step": 46885
    },
    {
      "entropy": 5.691402959823608,
      "epoch": 5.2661874543718765,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025996347514082603,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2316884160041809,
      "num_tokens": 104676964.0,
      "step": 46890
    },
    {
      "entropy": 5.803315496444702,
      "epoch": 5.266749031279834,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025992341641613065,
      "loss": 4.9591,
      "mean_token_accuracy": 0.22898137122392653,
      "num_tokens": 104688853.0,
      "step": 46895
    },
    {
      "entropy": 5.703245449066162,
      "epoch": 5.267310608187791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002598833581714795,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23699203729629517,
      "num_tokens": 104699723.0,
      "step": 46900
    },
    {
      "entropy": 5.728063106536865,
      "epoch": 5.267872185095749,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025984330040814817,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23874409198760987,
      "num_tokens": 104711370.0,
      "step": 46905
    },
    {
      "entropy": 5.766997814178467,
      "epoch": 5.268433762003706,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002598032431274119,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2269150972366333,
      "num_tokens": 104722939.0,
      "step": 46910
    },
    {
      "entropy": 5.7617058753967285,
      "epoch": 5.2689953389116635,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025976318633054636,
      "loss": 4.9857,
      "mean_token_accuracy": 0.22788286060094834,
      "num_tokens": 104735433.0,
      "step": 46915
    },
    {
      "entropy": 5.749786376953125,
      "epoch": 5.269556915819622,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002597231300188269,
      "loss": 4.8965,
      "mean_token_accuracy": 0.23426633924245835,
      "num_tokens": 104748325.0,
      "step": 46920
    },
    {
      "entropy": 5.741112852096558,
      "epoch": 5.270118492727579,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025968307419352884,
      "loss": 4.9293,
      "mean_token_accuracy": 0.23375632762908935,
      "num_tokens": 104758745.0,
      "step": 46925
    },
    {
      "entropy": 5.65616888999939,
      "epoch": 5.270680069635537,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025964301885592766,
      "loss": 4.8105,
      "mean_token_accuracy": 0.2386339247226715,
      "num_tokens": 104769424.0,
      "step": 46930
    },
    {
      "entropy": 5.628004789352417,
      "epoch": 5.271241646543494,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002596029640072988,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2390105202794075,
      "num_tokens": 104779913.0,
      "step": 46935
    },
    {
      "entropy": 5.6797833919525145,
      "epoch": 5.271803223451451,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002595629096489174,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2287687763571739,
      "num_tokens": 104793066.0,
      "step": 46940
    },
    {
      "entropy": 5.677030277252197,
      "epoch": 5.2723648003594095,
      "grad_norm": 1.234375,
      "learning_rate": 0.000259522855782059,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23822158426046372,
      "num_tokens": 104803641.0,
      "step": 46945
    },
    {
      "entropy": 5.80878472328186,
      "epoch": 5.272926377267367,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002594828024079989,
      "loss": 5.0031,
      "mean_token_accuracy": 0.22206757217645645,
      "num_tokens": 104816025.0,
      "step": 46950
    },
    {
      "entropy": 5.767509126663208,
      "epoch": 5.273487954175324,
      "grad_norm": 1.1875,
      "learning_rate": 0.00025944274952801233,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2305646851658821,
      "num_tokens": 104828188.0,
      "step": 46955
    },
    {
      "entropy": 5.6963459014892575,
      "epoch": 5.274049531083282,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002594026971433747,
      "loss": 4.9347,
      "mean_token_accuracy": 0.23392205834388732,
      "num_tokens": 104838276.0,
      "step": 46960
    },
    {
      "entropy": 5.704325532913208,
      "epoch": 5.274611107991239,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025936264525536126,
      "loss": 4.9468,
      "mean_token_accuracy": 0.2312423035502434,
      "num_tokens": 104848295.0,
      "step": 46965
    },
    {
      "entropy": 5.8009946823120115,
      "epoch": 5.275172684899197,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002593225938652472,
      "loss": 5.0291,
      "mean_token_accuracy": 0.22586823403835296,
      "num_tokens": 104857857.0,
      "step": 46970
    },
    {
      "entropy": 5.723586750030518,
      "epoch": 5.2757342618071545,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002592825429743079,
      "loss": 4.9101,
      "mean_token_accuracy": 0.23260480612516404,
      "num_tokens": 104867819.0,
      "step": 46975
    },
    {
      "entropy": 5.819567537307739,
      "epoch": 5.276295838715112,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002592424925838185,
      "loss": 4.9442,
      "mean_token_accuracy": 0.22813328802585603,
      "num_tokens": 104879034.0,
      "step": 46980
    },
    {
      "entropy": 5.791859197616577,
      "epoch": 5.27685741562307,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002592024426950541,
      "loss": 5.0273,
      "mean_token_accuracy": 0.22337529212236404,
      "num_tokens": 104890245.0,
      "step": 46985
    },
    {
      "entropy": 5.726378488540649,
      "epoch": 5.277418992531027,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025916239330929026,
      "loss": 4.9319,
      "mean_token_accuracy": 0.233552385866642,
      "num_tokens": 104900805.0,
      "step": 46990
    },
    {
      "entropy": 5.663439655303955,
      "epoch": 5.277980569438984,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025912234442780185,
      "loss": 4.8269,
      "mean_token_accuracy": 0.24152521640062333,
      "num_tokens": 104911636.0,
      "step": 46995
    },
    {
      "entropy": 5.736020517349243,
      "epoch": 5.278542146346942,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025908229605186423,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2373928427696228,
      "num_tokens": 104922253.0,
      "step": 47000
    },
    {
      "entropy": 5.64994101524353,
      "epoch": 5.2791037232549,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025904224818275247,
      "loss": 4.8781,
      "mean_token_accuracy": 0.23556830435991288,
      "num_tokens": 104933274.0,
      "step": 47005
    },
    {
      "entropy": 5.781205844879151,
      "epoch": 5.279665300162858,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025900220082174165,
      "loss": 4.9559,
      "mean_token_accuracy": 0.22441641688346864,
      "num_tokens": 104943931.0,
      "step": 47010
    },
    {
      "entropy": 5.6641031265258786,
      "epoch": 5.280226877070815,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025896215397010696,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2296583354473114,
      "num_tokens": 104955080.0,
      "step": 47015
    },
    {
      "entropy": 5.709355974197388,
      "epoch": 5.280788453978772,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002589221076291236,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23897408694028854,
      "num_tokens": 104967232.0,
      "step": 47020
    },
    {
      "entropy": 5.711672782897949,
      "epoch": 5.28135003088673,
      "grad_norm": 1.1875,
      "learning_rate": 0.00025888206180006647,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22826375365257262,
      "num_tokens": 104979462.0,
      "step": 47025
    },
    {
      "entropy": 5.718045043945312,
      "epoch": 5.2819116077946875,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002588420164842108,
      "loss": 4.9548,
      "mean_token_accuracy": 0.229871167242527,
      "num_tokens": 104990567.0,
      "step": 47030
    },
    {
      "entropy": 5.627068758010864,
      "epoch": 5.282473184702645,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002588019716828316,
      "loss": 4.7589,
      "mean_token_accuracy": 0.24248310327529907,
      "num_tokens": 105001965.0,
      "step": 47035
    },
    {
      "entropy": 5.690421533584595,
      "epoch": 5.283034761610603,
      "grad_norm": 1.1875,
      "learning_rate": 0.00025876192739720383,
      "loss": 4.796,
      "mean_token_accuracy": 0.24416870325803758,
      "num_tokens": 105013349.0,
      "step": 47040
    },
    {
      "entropy": 5.68878207206726,
      "epoch": 5.28359633851856,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025872188362860264,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2353678047657013,
      "num_tokens": 105025009.0,
      "step": 47045
    },
    {
      "entropy": 5.63876166343689,
      "epoch": 5.284157915426517,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00025868184037830305,
      "loss": 4.8201,
      "mean_token_accuracy": 0.24005085825920106,
      "num_tokens": 105035474.0,
      "step": 47050
    },
    {
      "entropy": 5.765946722030639,
      "epoch": 5.284719492334475,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002586417976475798,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22833588421344758,
      "num_tokens": 105047310.0,
      "step": 47055
    },
    {
      "entropy": 5.639092445373535,
      "epoch": 5.285281069242433,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025860175543770817,
      "loss": 4.7837,
      "mean_token_accuracy": 0.2372434601187706,
      "num_tokens": 105057198.0,
      "step": 47060
    },
    {
      "entropy": 5.753183317184448,
      "epoch": 5.285842646150391,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002585617137499631,
      "loss": 4.9102,
      "mean_token_accuracy": 0.23283104300498964,
      "num_tokens": 105067885.0,
      "step": 47065
    },
    {
      "entropy": 5.738375425338745,
      "epoch": 5.286404223058348,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002585216725856193,
      "loss": 4.9381,
      "mean_token_accuracy": 0.2312258169054985,
      "num_tokens": 105078917.0,
      "step": 47070
    },
    {
      "entropy": 5.763571310043335,
      "epoch": 5.286965799966305,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002584816319459519,
      "loss": 5.0032,
      "mean_token_accuracy": 0.22617995440959932,
      "num_tokens": 105089425.0,
      "step": 47075
    },
    {
      "entropy": 5.791400337219239,
      "epoch": 5.287527376874263,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002584415918322358,
      "loss": 4.9259,
      "mean_token_accuracy": 0.23167065680027008,
      "num_tokens": 105100043.0,
      "step": 47080
    },
    {
      "entropy": 5.676818037033081,
      "epoch": 5.28808895378222,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002584015522457458,
      "loss": 4.8272,
      "mean_token_accuracy": 0.23403735607862472,
      "num_tokens": 105110317.0,
      "step": 47085
    },
    {
      "entropy": 5.70330958366394,
      "epoch": 5.288650530690178,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025836151318775675,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22828383296728133,
      "num_tokens": 105120688.0,
      "step": 47090
    },
    {
      "entropy": 5.813273000717163,
      "epoch": 5.289212107598136,
      "grad_norm": 1.25,
      "learning_rate": 0.00025832147465954365,
      "loss": 5.0079,
      "mean_token_accuracy": 0.23032391220331191,
      "num_tokens": 105130883.0,
      "step": 47095
    },
    {
      "entropy": 5.785958480834961,
      "epoch": 5.289773684506093,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025828143666238124,
      "loss": 4.9769,
      "mean_token_accuracy": 0.2283341869711876,
      "num_tokens": 105141203.0,
      "step": 47100
    },
    {
      "entropy": 5.728460073471069,
      "epoch": 5.29033526141405,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025824139919754436,
      "loss": 4.8965,
      "mean_token_accuracy": 0.23676841259002684,
      "num_tokens": 105152419.0,
      "step": 47105
    },
    {
      "entropy": 5.815078544616699,
      "epoch": 5.290896838322008,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002582013622663079,
      "loss": 5.0368,
      "mean_token_accuracy": 0.22982011437416078,
      "num_tokens": 105164242.0,
      "step": 47110
    },
    {
      "entropy": 5.775970506668091,
      "epoch": 5.2914584152299655,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002581613258699465,
      "loss": 4.955,
      "mean_token_accuracy": 0.237685863673687,
      "num_tokens": 105174008.0,
      "step": 47115
    },
    {
      "entropy": 5.654160451889038,
      "epoch": 5.292019992137924,
      "grad_norm": 1.3125,
      "learning_rate": 0.000258121290009735,
      "loss": 4.8937,
      "mean_token_accuracy": 0.2343290388584137,
      "num_tokens": 105184880.0,
      "step": 47120
    },
    {
      "entropy": 5.750075387954712,
      "epoch": 5.292581569045881,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002580812546869481,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22606091648340226,
      "num_tokens": 105196492.0,
      "step": 47125
    },
    {
      "entropy": 5.6472255229949955,
      "epoch": 5.293143145953838,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002580412199028607,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23153224736452102,
      "num_tokens": 105206227.0,
      "step": 47130
    },
    {
      "entropy": 5.8103804111480715,
      "epoch": 5.293704722861796,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025800118565874754,
      "loss": 5.0726,
      "mean_token_accuracy": 0.22290928810834884,
      "num_tokens": 105217877.0,
      "step": 47135
    },
    {
      "entropy": 5.739455986022949,
      "epoch": 5.294266299769753,
      "grad_norm": 1.25,
      "learning_rate": 0.0002579611519558831,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23775921314954757,
      "num_tokens": 105229124.0,
      "step": 47140
    },
    {
      "entropy": 5.652945852279663,
      "epoch": 5.294827876677711,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025792111879554224,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2466011330485344,
      "num_tokens": 105240952.0,
      "step": 47145
    },
    {
      "entropy": 5.709973239898682,
      "epoch": 5.295389453585669,
      "grad_norm": 1.25,
      "learning_rate": 0.0002578810861789996,
      "loss": 4.8989,
      "mean_token_accuracy": 0.235058930516243,
      "num_tokens": 105252594.0,
      "step": 47150
    },
    {
      "entropy": 5.772776651382446,
      "epoch": 5.295951030493626,
      "grad_norm": 1.25,
      "learning_rate": 0.0002578410541075298,
      "loss": 4.9887,
      "mean_token_accuracy": 0.22692185491323472,
      "num_tokens": 105263457.0,
      "step": 47155
    },
    {
      "entropy": 5.7099486827850345,
      "epoch": 5.296512607401584,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025780102258240745,
      "loss": 4.8981,
      "mean_token_accuracy": 0.23678000420331954,
      "num_tokens": 105275455.0,
      "step": 47160
    },
    {
      "entropy": 5.74954252243042,
      "epoch": 5.297074184309541,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002577609916049072,
      "loss": 4.9328,
      "mean_token_accuracy": 0.23507894277572633,
      "num_tokens": 105285643.0,
      "step": 47165
    },
    {
      "entropy": 5.689897966384888,
      "epoch": 5.2976357612174985,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002577209611763037,
      "loss": 4.9388,
      "mean_token_accuracy": 0.232782743871212,
      "num_tokens": 105297322.0,
      "step": 47170
    },
    {
      "entropy": 5.655672931671143,
      "epoch": 5.298197338125457,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002576809312978715,
      "loss": 4.8185,
      "mean_token_accuracy": 0.23408450186252594,
      "num_tokens": 105307977.0,
      "step": 47175
    },
    {
      "entropy": 5.693756103515625,
      "epoch": 5.298758915033414,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002576409019708852,
      "loss": 4.8896,
      "mean_token_accuracy": 0.23697670996189119,
      "num_tokens": 105318461.0,
      "step": 47180
    },
    {
      "entropy": 5.81259069442749,
      "epoch": 5.299320491941371,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002576008731966193,
      "loss": 5.0342,
      "mean_token_accuracy": 0.2222766950726509,
      "num_tokens": 105329494.0,
      "step": 47185
    },
    {
      "entropy": 5.5908924579620365,
      "epoch": 5.299882068849329,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025756084497634827,
      "loss": 4.7453,
      "mean_token_accuracy": 0.24312693178653716,
      "num_tokens": 105341475.0,
      "step": 47190
    },
    {
      "entropy": 5.701360177993775,
      "epoch": 5.300443645757286,
      "grad_norm": 1.375,
      "learning_rate": 0.0002575208173113468,
      "loss": 4.8259,
      "mean_token_accuracy": 0.2428544655442238,
      "num_tokens": 105352075.0,
      "step": 47195
    },
    {
      "entropy": 5.725534152984619,
      "epoch": 5.301005222665244,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002574807902028891,
      "loss": 4.9457,
      "mean_token_accuracy": 0.23156039863824845,
      "num_tokens": 105363801.0,
      "step": 47200
    },
    {
      "entropy": 5.761869812011719,
      "epoch": 5.301566799573202,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025744076365225,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2371249318122864,
      "num_tokens": 105374821.0,
      "step": 47205
    },
    {
      "entropy": 5.706854820251465,
      "epoch": 5.302128376481159,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025740073766070376,
      "loss": 4.8776,
      "mean_token_accuracy": 0.23734974712133408,
      "num_tokens": 105385890.0,
      "step": 47210
    },
    {
      "entropy": 5.759768104553222,
      "epoch": 5.302689953389117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002573607122295249,
      "loss": 4.9581,
      "mean_token_accuracy": 0.2284274846315384,
      "num_tokens": 105396943.0,
      "step": 47215
    },
    {
      "entropy": 5.77320442199707,
      "epoch": 5.303251530297074,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002573206873599878,
      "loss": 4.9708,
      "mean_token_accuracy": 0.23272179514169694,
      "num_tokens": 105407223.0,
      "step": 47220
    },
    {
      "entropy": 5.77117428779602,
      "epoch": 5.303813107205031,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025728066305336685,
      "loss": 4.9139,
      "mean_token_accuracy": 0.23412801325321198,
      "num_tokens": 105417714.0,
      "step": 47225
    },
    {
      "entropy": 5.836246585845947,
      "epoch": 5.3043746841129895,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002572406393109364,
      "loss": 5.0555,
      "mean_token_accuracy": 0.23143706768751143,
      "num_tokens": 105429035.0,
      "step": 47230
    },
    {
      "entropy": 5.7203937530517575,
      "epoch": 5.304936261020947,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002572006161339709,
      "loss": 4.9026,
      "mean_token_accuracy": 0.2278515338897705,
      "num_tokens": 105440347.0,
      "step": 47235
    },
    {
      "entropy": 5.784732294082642,
      "epoch": 5.305497837928904,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002571605935237448,
      "loss": 4.9935,
      "mean_token_accuracy": 0.2287690207362175,
      "num_tokens": 105451265.0,
      "step": 47240
    },
    {
      "entropy": 5.772875833511352,
      "epoch": 5.306059414836862,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002571205714815322,
      "loss": 4.9855,
      "mean_token_accuracy": 0.22541827261447905,
      "num_tokens": 105462034.0,
      "step": 47245
    },
    {
      "entropy": 5.758577489852906,
      "epoch": 5.306620991744819,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002570805500086076,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2358751878142357,
      "num_tokens": 105472733.0,
      "step": 47250
    },
    {
      "entropy": 5.775497961044311,
      "epoch": 5.307182568652777,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002570405291062452,
      "loss": 4.9891,
      "mean_token_accuracy": 0.22675827294588088,
      "num_tokens": 105483152.0,
      "step": 47255
    },
    {
      "entropy": 5.662153148651123,
      "epoch": 5.307744145560735,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002570005087757193,
      "loss": 4.8119,
      "mean_token_accuracy": 0.23844778090715407,
      "num_tokens": 105493541.0,
      "step": 47260
    },
    {
      "entropy": 5.702059364318847,
      "epoch": 5.308305722468692,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002569604890183043,
      "loss": 4.9187,
      "mean_token_accuracy": 0.23330777138471603,
      "num_tokens": 105505550.0,
      "step": 47265
    },
    {
      "entropy": 5.67128005027771,
      "epoch": 5.30886729937665,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002569204698352741,
      "loss": 4.8918,
      "mean_token_accuracy": 0.23874042481184005,
      "num_tokens": 105518460.0,
      "step": 47270
    },
    {
      "entropy": 5.714705276489258,
      "epoch": 5.309428876284607,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002568804512279033,
      "loss": 4.8749,
      "mean_token_accuracy": 0.2361736297607422,
      "num_tokens": 105529074.0,
      "step": 47275
    },
    {
      "entropy": 5.730164337158203,
      "epoch": 5.309990453192564,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025684043319746595,
      "loss": 4.906,
      "mean_token_accuracy": 0.23400490283966063,
      "num_tokens": 105540501.0,
      "step": 47280
    },
    {
      "entropy": 5.7318226337432865,
      "epoch": 5.3105520301005225,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025680041574523626,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23468708992004395,
      "num_tokens": 105551158.0,
      "step": 47285
    },
    {
      "entropy": 5.700511646270752,
      "epoch": 5.31111360700848,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002567603988724884,
      "loss": 4.8687,
      "mean_token_accuracy": 0.24742306172847747,
      "num_tokens": 105563219.0,
      "step": 47290
    },
    {
      "entropy": 5.69121789932251,
      "epoch": 5.311675183916437,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002567203825804966,
      "loss": 4.9002,
      "mean_token_accuracy": 0.23944982290267944,
      "num_tokens": 105574846.0,
      "step": 47295
    },
    {
      "entropy": 5.667195081710815,
      "epoch": 5.312236760824395,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002566803668705347,
      "loss": 4.8589,
      "mean_token_accuracy": 0.240269273519516,
      "num_tokens": 105585848.0,
      "step": 47300
    },
    {
      "entropy": 5.725594806671142,
      "epoch": 5.312798337732352,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002566403517438772,
      "loss": 4.9396,
      "mean_token_accuracy": 0.23916372209787368,
      "num_tokens": 105596757.0,
      "step": 47305
    },
    {
      "entropy": 5.722304391860962,
      "epoch": 5.31335991464031,
      "grad_norm": 1.25,
      "learning_rate": 0.00025660033720179795,
      "loss": 5.0463,
      "mean_token_accuracy": 0.22647486478090287,
      "num_tokens": 105609363.0,
      "step": 47310
    },
    {
      "entropy": 5.732486438751221,
      "epoch": 5.3139214915482675,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025656032324557107,
      "loss": 4.9065,
      "mean_token_accuracy": 0.23412539064884186,
      "num_tokens": 105620909.0,
      "step": 47315
    },
    {
      "entropy": 5.674446439743042,
      "epoch": 5.314483068456225,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00025652030987647066,
      "loss": 4.7856,
      "mean_token_accuracy": 0.24000355154275893,
      "num_tokens": 105631134.0,
      "step": 47320
    },
    {
      "entropy": 5.63794059753418,
      "epoch": 5.315044645364183,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002564802970957708,
      "loss": 4.8108,
      "mean_token_accuracy": 0.24534085839986802,
      "num_tokens": 105641979.0,
      "step": 47325
    },
    {
      "entropy": 5.703265619277954,
      "epoch": 5.31560622227214,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025644028490474544,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2323181539773941,
      "num_tokens": 105652320.0,
      "step": 47330
    },
    {
      "entropy": 5.72889895439148,
      "epoch": 5.316167799180098,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025640027330466866,
      "loss": 4.8992,
      "mean_token_accuracy": 0.23065903633832932,
      "num_tokens": 105663278.0,
      "step": 47335
    },
    {
      "entropy": 5.744380617141724,
      "epoch": 5.316729376088055,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025636026229681433,
      "loss": 4.9482,
      "mean_token_accuracy": 0.2350664645433426,
      "num_tokens": 105673930.0,
      "step": 47340
    },
    {
      "entropy": 5.674177169799805,
      "epoch": 5.317290952996013,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002563202518824564,
      "loss": 4.8394,
      "mean_token_accuracy": 0.24018379151821137,
      "num_tokens": 105684617.0,
      "step": 47345
    },
    {
      "entropy": 5.672280597686767,
      "epoch": 5.317852529903971,
      "grad_norm": 1.25,
      "learning_rate": 0.000256280242062869,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2370416060090065,
      "num_tokens": 105695278.0,
      "step": 47350
    },
    {
      "entropy": 5.729849767684937,
      "epoch": 5.318414106811928,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002562402328393259,
      "loss": 4.8877,
      "mean_token_accuracy": 0.23127377182245254,
      "num_tokens": 105706026.0,
      "step": 47355
    },
    {
      "entropy": 5.692736291885376,
      "epoch": 5.318975683719885,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025620022421310115,
      "loss": 4.8551,
      "mean_token_accuracy": 0.23917495161294938,
      "num_tokens": 105717141.0,
      "step": 47360
    },
    {
      "entropy": 5.719398546218872,
      "epoch": 5.319537260627843,
      "grad_norm": 1.15625,
      "learning_rate": 0.00025616021618546847,
      "loss": 4.9022,
      "mean_token_accuracy": 0.23727618604898454,
      "num_tokens": 105729240.0,
      "step": 47365
    },
    {
      "entropy": 5.7379233837127686,
      "epoch": 5.3200988375358005,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025612020875770186,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2336053192615509,
      "num_tokens": 105741344.0,
      "step": 47370
    },
    {
      "entropy": 5.747870874404907,
      "epoch": 5.320660414443758,
      "grad_norm": 1.25,
      "learning_rate": 0.0002560802019310751,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23976625204086305,
      "num_tokens": 105752750.0,
      "step": 47375
    },
    {
      "entropy": 5.7346573829650875,
      "epoch": 5.321221991351716,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025604019570686203,
      "loss": 4.9568,
      "mean_token_accuracy": 0.23237101435661317,
      "num_tokens": 105764677.0,
      "step": 47380
    },
    {
      "entropy": 5.733879899978637,
      "epoch": 5.321783568259673,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002560001900863365,
      "loss": 4.8749,
      "mean_token_accuracy": 0.23162132799625396,
      "num_tokens": 105775649.0,
      "step": 47385
    },
    {
      "entropy": 5.741329336166382,
      "epoch": 5.322345145167631,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002559601850707723,
      "loss": 4.912,
      "mean_token_accuracy": 0.23761836141347886,
      "num_tokens": 105785605.0,
      "step": 47390
    },
    {
      "entropy": 5.715231275558471,
      "epoch": 5.322906722075588,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002559201806614432,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23182040452957153,
      "num_tokens": 105795887.0,
      "step": 47395
    },
    {
      "entropy": 5.679343748092651,
      "epoch": 5.323468298983546,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025588017685962297,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23798947632312775,
      "num_tokens": 105806129.0,
      "step": 47400
    },
    {
      "entropy": 5.76726975440979,
      "epoch": 5.324029875891504,
      "grad_norm": 1.234375,
      "learning_rate": 0.00025584017366658525,
      "loss": 4.9631,
      "mean_token_accuracy": 0.23230984658002854,
      "num_tokens": 105817048.0,
      "step": 47405
    },
    {
      "entropy": 5.754914331436157,
      "epoch": 5.324591452799461,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025580017108360386,
      "loss": 4.9849,
      "mean_token_accuracy": 0.23191641122102738,
      "num_tokens": 105829847.0,
      "step": 47410
    },
    {
      "entropy": 5.732714605331421,
      "epoch": 5.325153029707418,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002557601691119524,
      "loss": 4.8921,
      "mean_token_accuracy": 0.23985091596841812,
      "num_tokens": 105839797.0,
      "step": 47415
    },
    {
      "entropy": 5.734473848342896,
      "epoch": 5.325714606615376,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00025572016775290475,
      "loss": 4.9306,
      "mean_token_accuracy": 0.22880012094974517,
      "num_tokens": 105851004.0,
      "step": 47420
    },
    {
      "entropy": 5.729868173599243,
      "epoch": 5.326276183523333,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025568016700773435,
      "loss": 4.9661,
      "mean_token_accuracy": 0.22373055070638656,
      "num_tokens": 105861558.0,
      "step": 47425
    },
    {
      "entropy": 5.688717889785766,
      "epoch": 5.326837760431291,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025564016687771495,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23857473880052565,
      "num_tokens": 105872705.0,
      "step": 47430
    },
    {
      "entropy": 5.699254131317138,
      "epoch": 5.327399337339249,
      "grad_norm": 1.1875,
      "learning_rate": 0.00025560016736412015,
      "loss": 4.9441,
      "mean_token_accuracy": 0.2310459330677986,
      "num_tokens": 105884426.0,
      "step": 47435
    },
    {
      "entropy": 5.686929512023926,
      "epoch": 5.327960914247206,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025556016846822353,
      "loss": 4.8946,
      "mean_token_accuracy": 0.23327834159135818,
      "num_tokens": 105896141.0,
      "step": 47440
    },
    {
      "entropy": 5.71047739982605,
      "epoch": 5.328522491155164,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025552017019129874,
      "loss": 4.837,
      "mean_token_accuracy": 0.2428850457072258,
      "num_tokens": 105907216.0,
      "step": 47445
    },
    {
      "entropy": 5.8045783042907715,
      "epoch": 5.329084068063121,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025548017253461923,
      "loss": 4.9798,
      "mean_token_accuracy": 0.23948068171739578,
      "num_tokens": 105918011.0,
      "step": 47450
    },
    {
      "entropy": 5.780403184890747,
      "epoch": 5.3296456449710785,
      "grad_norm": 1.25,
      "learning_rate": 0.0002554401754994586,
      "loss": 4.9812,
      "mean_token_accuracy": 0.23039890825748444,
      "num_tokens": 105929173.0,
      "step": 47455
    },
    {
      "entropy": 5.758951997756958,
      "epoch": 5.330207221879037,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002554001790870904,
      "loss": 4.9543,
      "mean_token_accuracy": 0.23346073925495148,
      "num_tokens": 105940825.0,
      "step": 47460
    },
    {
      "entropy": 5.71659026145935,
      "epoch": 5.330768798786994,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00025536018329878814,
      "loss": 4.8677,
      "mean_token_accuracy": 0.23805305659770964,
      "num_tokens": 105951335.0,
      "step": 47465
    },
    {
      "entropy": 5.74775538444519,
      "epoch": 5.331330375694951,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002553201881358252,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22708771228790284,
      "num_tokens": 105963529.0,
      "step": 47470
    },
    {
      "entropy": 5.837436532974243,
      "epoch": 5.331891952602909,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025528019359947514,
      "loss": 5.0113,
      "mean_token_accuracy": 0.22629703283309938,
      "num_tokens": 105975386.0,
      "step": 47475
    },
    {
      "entropy": 5.705038928985596,
      "epoch": 5.332453529510866,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002552401996910113,
      "loss": 4.8539,
      "mean_token_accuracy": 0.2331700697541237,
      "num_tokens": 105985710.0,
      "step": 47480
    },
    {
      "entropy": 5.645787477493286,
      "epoch": 5.333015106418824,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002552002064117072,
      "loss": 4.8169,
      "mean_token_accuracy": 0.24145326316356658,
      "num_tokens": 105997216.0,
      "step": 47485
    },
    {
      "entropy": 5.732470417022705,
      "epoch": 5.333576683326782,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002551602137628361,
      "loss": 4.8769,
      "mean_token_accuracy": 0.24308751225471498,
      "num_tokens": 106008111.0,
      "step": 47490
    },
    {
      "entropy": 5.7421201229095455,
      "epoch": 5.334138260234739,
      "grad_norm": 1.34375,
      "learning_rate": 0.00025512022174567164,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2370323970913887,
      "num_tokens": 106018732.0,
      "step": 47495
    },
    {
      "entropy": 5.698074054718018,
      "epoch": 5.334699837142697,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025508023036148697,
      "loss": 4.839,
      "mean_token_accuracy": 0.2333671972155571,
      "num_tokens": 106029740.0,
      "step": 47500
    },
    {
      "entropy": 5.663387680053711,
      "epoch": 5.335261414050654,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025504023961155553,
      "loss": 4.8398,
      "mean_token_accuracy": 0.24645345509052277,
      "num_tokens": 106040553.0,
      "step": 47505
    },
    {
      "entropy": 5.701514387130738,
      "epoch": 5.3358229909586115,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025500024949715055,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2326935589313507,
      "num_tokens": 106051667.0,
      "step": 47510
    },
    {
      "entropy": 5.801189994812011,
      "epoch": 5.33638456786657,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025496026001954543,
      "loss": 5.0074,
      "mean_token_accuracy": 0.23040899336338044,
      "num_tokens": 106062492.0,
      "step": 47515
    },
    {
      "entropy": 5.783686685562134,
      "epoch": 5.336946144774527,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025492027118001345,
      "loss": 4.9433,
      "mean_token_accuracy": 0.23096464276313783,
      "num_tokens": 106072645.0,
      "step": 47520
    },
    {
      "entropy": 5.670742559432983,
      "epoch": 5.337507721682485,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002548802829798277,
      "loss": 4.8351,
      "mean_token_accuracy": 0.2448546513915062,
      "num_tokens": 106084115.0,
      "step": 47525
    },
    {
      "entropy": 5.685857915878296,
      "epoch": 5.338069298590442,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002548402954202616,
      "loss": 4.9134,
      "mean_token_accuracy": 0.23409562706947326,
      "num_tokens": 106094093.0,
      "step": 47530
    },
    {
      "entropy": 5.705794858932495,
      "epoch": 5.338630875498399,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002548003085025883,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23781046122312546,
      "num_tokens": 106104452.0,
      "step": 47535
    },
    {
      "entropy": 5.696465635299683,
      "epoch": 5.339192452406357,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025476032222808106,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23099120259284972,
      "num_tokens": 106114540.0,
      "step": 47540
    },
    {
      "entropy": 5.717884922027588,
      "epoch": 5.339754029314315,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025472033659801297,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22979109585285187,
      "num_tokens": 106124981.0,
      "step": 47545
    },
    {
      "entropy": 5.695574998855591,
      "epoch": 5.340315606222272,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002546803516136572,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2272832676768303,
      "num_tokens": 106135037.0,
      "step": 47550
    },
    {
      "entropy": 5.776706171035767,
      "epoch": 5.34087718313023,
      "grad_norm": 1.265625,
      "learning_rate": 0.00025464036727628695,
      "loss": 4.9647,
      "mean_token_accuracy": 0.23099718987941742,
      "num_tokens": 106145515.0,
      "step": 47555
    },
    {
      "entropy": 5.755941104888916,
      "epoch": 5.341438760038187,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025460038358717527,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23284514546394347,
      "num_tokens": 106157246.0,
      "step": 47560
    },
    {
      "entropy": 5.735108995437622,
      "epoch": 5.342000336946144,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002545604005475953,
      "loss": 4.9501,
      "mean_token_accuracy": 0.23018169403076172,
      "num_tokens": 106167889.0,
      "step": 47565
    },
    {
      "entropy": 5.6685045719146725,
      "epoch": 5.3425619138541025,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002545204181588201,
      "loss": 4.9072,
      "mean_token_accuracy": 0.23516298681497574,
      "num_tokens": 106178330.0,
      "step": 47570
    },
    {
      "entropy": 5.7250611782073975,
      "epoch": 5.34312349076206,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025448043642212274,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2329896494746208,
      "num_tokens": 106188995.0,
      "step": 47575
    },
    {
      "entropy": 5.765840625762939,
      "epoch": 5.343685067670018,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0002544404553387763,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22969738841056825,
      "num_tokens": 106199271.0,
      "step": 47580
    },
    {
      "entropy": 5.68995943069458,
      "epoch": 5.344246644577975,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025440047491005367,
      "loss": 4.9218,
      "mean_token_accuracy": 0.23530919551849366,
      "num_tokens": 106211495.0,
      "step": 47585
    },
    {
      "entropy": 5.68708758354187,
      "epoch": 5.344808221485932,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025436049513722795,
      "loss": 4.8561,
      "mean_token_accuracy": 0.2386656329035759,
      "num_tokens": 106223437.0,
      "step": 47590
    },
    {
      "entropy": 5.7605414390563965,
      "epoch": 5.34536979839389,
      "grad_norm": 1.328125,
      "learning_rate": 0.00025432051602157203,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22778613418340682,
      "num_tokens": 106233456.0,
      "step": 47595
    },
    {
      "entropy": 5.706081342697144,
      "epoch": 5.345931375301848,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002542805375643589,
      "loss": 4.8135,
      "mean_token_accuracy": 0.235844224691391,
      "num_tokens": 106244559.0,
      "step": 47600
    },
    {
      "entropy": 5.6544084548950195,
      "epoch": 5.346492952209805,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002542405597668615,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23656601160764695,
      "num_tokens": 106254427.0,
      "step": 47605
    },
    {
      "entropy": 5.668998956680298,
      "epoch": 5.347054529117763,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025420058263035276,
      "loss": 4.8921,
      "mean_token_accuracy": 0.23831116259098054,
      "num_tokens": 106264616.0,
      "step": 47610
    },
    {
      "entropy": 5.701438617706299,
      "epoch": 5.34761610602572,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002541606061561055,
      "loss": 4.8446,
      "mean_token_accuracy": 0.24107838571071624,
      "num_tokens": 106275103.0,
      "step": 47615
    },
    {
      "entropy": 5.779726457595825,
      "epoch": 5.348177682933677,
      "grad_norm": 1.25,
      "learning_rate": 0.0002541206303453927,
      "loss": 4.9896,
      "mean_token_accuracy": 0.23061943501234056,
      "num_tokens": 106286004.0,
      "step": 47620
    },
    {
      "entropy": 5.77326807975769,
      "epoch": 5.3487392598416355,
      "grad_norm": 1.3125,
      "learning_rate": 0.000254080655199487,
      "loss": 4.9287,
      "mean_token_accuracy": 0.23508456945419312,
      "num_tokens": 106297161.0,
      "step": 47625
    },
    {
      "entropy": 5.784700393676758,
      "epoch": 5.349300836749593,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00025404068071966133,
      "loss": 4.993,
      "mean_token_accuracy": 0.2236977681517601,
      "num_tokens": 106309330.0,
      "step": 47630
    },
    {
      "entropy": 5.760316944122314,
      "epoch": 5.349862413657551,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025400070690718865,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22902546375989913,
      "num_tokens": 106319801.0,
      "step": 47635
    },
    {
      "entropy": 5.7285459518432615,
      "epoch": 5.350423990565508,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025396073376334155,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22510057389736177,
      "num_tokens": 106331328.0,
      "step": 47640
    },
    {
      "entropy": 5.7338909149169925,
      "epoch": 5.350985567473465,
      "grad_norm": 1.25,
      "learning_rate": 0.00025392076128939283,
      "loss": 4.906,
      "mean_token_accuracy": 0.235590597987175,
      "num_tokens": 106343740.0,
      "step": 47645
    },
    {
      "entropy": 5.7028340816497805,
      "epoch": 5.351547144381423,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002538807894866153,
      "loss": 4.8445,
      "mean_token_accuracy": 0.24889385849237441,
      "num_tokens": 106354620.0,
      "step": 47650
    },
    {
      "entropy": 5.676136493682861,
      "epoch": 5.3521087212893805,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025384081835628163,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2380377858877182,
      "num_tokens": 106366247.0,
      "step": 47655
    },
    {
      "entropy": 5.733401727676392,
      "epoch": 5.352670298197338,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002538008478996645,
      "loss": 4.9397,
      "mean_token_accuracy": 0.23456366062164308,
      "num_tokens": 106377045.0,
      "step": 47660
    },
    {
      "entropy": 5.645670366287232,
      "epoch": 5.353231875105296,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002537608781180366,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23542707413434982,
      "num_tokens": 106389070.0,
      "step": 47665
    },
    {
      "entropy": 5.672217226028442,
      "epoch": 5.353793452013253,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002537209090126704,
      "loss": 4.87,
      "mean_token_accuracy": 0.236531625688076,
      "num_tokens": 106401292.0,
      "step": 47670
    },
    {
      "entropy": 5.747935676574707,
      "epoch": 5.35435502892121,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025368094058483885,
      "loss": 4.8689,
      "mean_token_accuracy": 0.23539715260267258,
      "num_tokens": 106413045.0,
      "step": 47675
    },
    {
      "entropy": 5.727750492095947,
      "epoch": 5.354916605829168,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002536409728358145,
      "loss": 4.8566,
      "mean_token_accuracy": 0.24023361802101134,
      "num_tokens": 106424655.0,
      "step": 47680
    },
    {
      "entropy": 5.757383203506469,
      "epoch": 5.355478182737126,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002536010057668698,
      "loss": 4.9631,
      "mean_token_accuracy": 0.2255005419254303,
      "num_tokens": 106436544.0,
      "step": 47685
    },
    {
      "entropy": 5.718592357635498,
      "epoch": 5.356039759645084,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002535610393792773,
      "loss": 4.9036,
      "mean_token_accuracy": 0.23144062161445617,
      "num_tokens": 106446936.0,
      "step": 47690
    },
    {
      "entropy": 5.738528442382813,
      "epoch": 5.356601336553041,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025352107367430975,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2331264555454254,
      "num_tokens": 106457328.0,
      "step": 47695
    },
    {
      "entropy": 5.697639846801758,
      "epoch": 5.357162913460998,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00025348110865323946,
      "loss": 4.9225,
      "mean_token_accuracy": 0.23547909557819366,
      "num_tokens": 106468037.0,
      "step": 47700
    },
    {
      "entropy": 5.710516452789307,
      "epoch": 5.357724490368956,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002534411443173389,
      "loss": 4.9253,
      "mean_token_accuracy": 0.23757695853710176,
      "num_tokens": 106479406.0,
      "step": 47705
    },
    {
      "entropy": 5.597842121124268,
      "epoch": 5.3582860672769135,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002534011806678809,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23280397653579712,
      "num_tokens": 106491142.0,
      "step": 47710
    },
    {
      "entropy": 5.677484130859375,
      "epoch": 5.358847644184872,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002533612177061376,
      "loss": 4.9161,
      "mean_token_accuracy": 0.23450518846511842,
      "num_tokens": 106502743.0,
      "step": 47715
    },
    {
      "entropy": 5.706745433807373,
      "epoch": 5.359409221092829,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025332125543338146,
      "loss": 4.9309,
      "mean_token_accuracy": 0.23709176033735274,
      "num_tokens": 106513765.0,
      "step": 47720
    },
    {
      "entropy": 5.687739896774292,
      "epoch": 5.359970798000786,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000253281293850885,
      "loss": 4.7837,
      "mean_token_accuracy": 0.24709200263023376,
      "num_tokens": 106524154.0,
      "step": 47725
    },
    {
      "entropy": 5.7309540748596195,
      "epoch": 5.360532374908744,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025324133295992056,
      "loss": 4.9094,
      "mean_token_accuracy": 0.23557631373405458,
      "num_tokens": 106535946.0,
      "step": 47730
    },
    {
      "entropy": 5.769757413864136,
      "epoch": 5.361093951816701,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002532013727617605,
      "loss": 5.0016,
      "mean_token_accuracy": 0.23021924048662185,
      "num_tokens": 106546352.0,
      "step": 47735
    },
    {
      "entropy": 5.68513069152832,
      "epoch": 5.361655528724659,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002531614132576772,
      "loss": 4.8239,
      "mean_token_accuracy": 0.24235369712114335,
      "num_tokens": 106557640.0,
      "step": 47740
    },
    {
      "entropy": 5.737710046768188,
      "epoch": 5.362217105632617,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000253121454448943,
      "loss": 4.895,
      "mean_token_accuracy": 0.22700136601924897,
      "num_tokens": 106567762.0,
      "step": 47745
    },
    {
      "entropy": 5.746862173080444,
      "epoch": 5.362778682540574,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025308149633683014,
      "loss": 4.8882,
      "mean_token_accuracy": 0.2457549214363098,
      "num_tokens": 106579589.0,
      "step": 47750
    },
    {
      "entropy": 5.7114705562591555,
      "epoch": 5.363340259448531,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000253041538922611,
      "loss": 4.872,
      "mean_token_accuracy": 0.23409072309732437,
      "num_tokens": 106590122.0,
      "step": 47755
    },
    {
      "entropy": 5.743349647521972,
      "epoch": 5.363901836356489,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025300158220755776,
      "loss": 4.9745,
      "mean_token_accuracy": 0.23107359260320665,
      "num_tokens": 106601269.0,
      "step": 47760
    },
    {
      "entropy": 5.7794561862945555,
      "epoch": 5.364463413264446,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025296162619294267,
      "loss": 5.0622,
      "mean_token_accuracy": 0.22380002290010453,
      "num_tokens": 106612462.0,
      "step": 47765
    },
    {
      "entropy": 5.753435850143433,
      "epoch": 5.3650249901724045,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002529216708800379,
      "loss": 4.901,
      "mean_token_accuracy": 0.23700395077466965,
      "num_tokens": 106623333.0,
      "step": 47770
    },
    {
      "entropy": 5.696902418136597,
      "epoch": 5.365586567080362,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00025288171627011564,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2395428255200386,
      "num_tokens": 106634505.0,
      "step": 47775
    },
    {
      "entropy": 5.779445028305053,
      "epoch": 5.366148143988319,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00025284176236444817,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2302863359451294,
      "num_tokens": 106645764.0,
      "step": 47780
    },
    {
      "entropy": 5.657665157318116,
      "epoch": 5.366709720896277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025280180916430764,
      "loss": 4.8513,
      "mean_token_accuracy": 0.23779519498348237,
      "num_tokens": 106657484.0,
      "step": 47785
    },
    {
      "entropy": 5.690886831283569,
      "epoch": 5.367271297804234,
      "grad_norm": 1.28125,
      "learning_rate": 0.00025276185667096605,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23164938688278197,
      "num_tokens": 106669038.0,
      "step": 47790
    },
    {
      "entropy": 5.764738845825195,
      "epoch": 5.3678328747121915,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025272190488569564,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2358582079410553,
      "num_tokens": 106679447.0,
      "step": 47795
    },
    {
      "entropy": 5.719118690490722,
      "epoch": 5.36839445162015,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025268195380976835,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2306215986609459,
      "num_tokens": 106689596.0,
      "step": 47800
    },
    {
      "entropy": 5.727885961532593,
      "epoch": 5.368956028528107,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002526420034444564,
      "loss": 4.8997,
      "mean_token_accuracy": 0.24120342880487441,
      "num_tokens": 106701825.0,
      "step": 47805
    },
    {
      "entropy": 5.811853837966919,
      "epoch": 5.369517605436064,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00025260205379103157,
      "loss": 5.0826,
      "mean_token_accuracy": 0.2201310873031616,
      "num_tokens": 106713451.0,
      "step": 47810
    },
    {
      "entropy": 5.772068071365356,
      "epoch": 5.370079182344022,
      "grad_norm": 1.359375,
      "learning_rate": 0.00025256210485076616,
      "loss": 5.0126,
      "mean_token_accuracy": 0.22417904287576676,
      "num_tokens": 106724732.0,
      "step": 47815
    },
    {
      "entropy": 5.78039436340332,
      "epoch": 5.370640759251979,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000252522156624932,
      "loss": 4.9486,
      "mean_token_accuracy": 0.23789778500795364,
      "num_tokens": 106736533.0,
      "step": 47820
    },
    {
      "entropy": 5.670916128158569,
      "epoch": 5.3712023361599375,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025248220911480105,
      "loss": 4.8359,
      "mean_token_accuracy": 0.2410484954714775,
      "num_tokens": 106747454.0,
      "step": 47825
    },
    {
      "entropy": 5.625749492645264,
      "epoch": 5.371763913067895,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00025244226232164533,
      "loss": 4.8458,
      "mean_token_accuracy": 0.24029953628778458,
      "num_tokens": 106757833.0,
      "step": 47830
    },
    {
      "entropy": 5.613366985321045,
      "epoch": 5.372325489975852,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025240231624673674,
      "loss": 4.8539,
      "mean_token_accuracy": 0.2308957278728485,
      "num_tokens": 106768646.0,
      "step": 47835
    },
    {
      "entropy": 5.7298253059387205,
      "epoch": 5.37288706688381,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002523623708913472,
      "loss": 4.8981,
      "mean_token_accuracy": 0.23493292480707167,
      "num_tokens": 106780434.0,
      "step": 47840
    },
    {
      "entropy": 5.721533012390137,
      "epoch": 5.373448643791767,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002523224262567485,
      "loss": 4.9253,
      "mean_token_accuracy": 0.24091764390468598,
      "num_tokens": 106792372.0,
      "step": 47845
    },
    {
      "entropy": 5.736039924621582,
      "epoch": 5.3740102206997244,
      "grad_norm": 1.46875,
      "learning_rate": 0.00025228248234421247,
      "loss": 4.892,
      "mean_token_accuracy": 0.22860318869352342,
      "num_tokens": 106802251.0,
      "step": 47850
    },
    {
      "entropy": 5.719500732421875,
      "epoch": 5.3745717976076826,
      "grad_norm": 1.375,
      "learning_rate": 0.0002522425391550111,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2368605688214302,
      "num_tokens": 106813818.0,
      "step": 47855
    },
    {
      "entropy": 5.757336854934692,
      "epoch": 5.37513337451564,
      "grad_norm": 1.296875,
      "learning_rate": 0.00025220259669041613,
      "loss": 4.99,
      "mean_token_accuracy": 0.22678103893995286,
      "num_tokens": 106824096.0,
      "step": 47860
    },
    {
      "entropy": 5.79515962600708,
      "epoch": 5.375694951423597,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002521626549516993,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22406664192676545,
      "num_tokens": 106836037.0,
      "step": 47865
    },
    {
      "entropy": 5.815630722045898,
      "epoch": 5.376256528331555,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002521227139401324,
      "loss": 5.0134,
      "mean_token_accuracy": 0.22859730273485185,
      "num_tokens": 106846370.0,
      "step": 47870
    },
    {
      "entropy": 5.7804638862609865,
      "epoch": 5.376818105239512,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025208277365698715,
      "loss": 4.9939,
      "mean_token_accuracy": 0.22992179840803145,
      "num_tokens": 106857945.0,
      "step": 47875
    },
    {
      "entropy": 5.725511121749878,
      "epoch": 5.37737968214747,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002520428341035353,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2383955791592598,
      "num_tokens": 106867899.0,
      "step": 47880
    },
    {
      "entropy": 5.6952790260314945,
      "epoch": 5.377941259055428,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002520028952810485,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22990721911191941,
      "num_tokens": 106879305.0,
      "step": 47885
    },
    {
      "entropy": 5.7475450992584225,
      "epoch": 5.378502835963385,
      "grad_norm": 1.21875,
      "learning_rate": 0.00025196295719079835,
      "loss": 4.983,
      "mean_token_accuracy": 0.2307627886533737,
      "num_tokens": 106891944.0,
      "step": 47890
    },
    {
      "entropy": 5.705444812774658,
      "epoch": 5.379064412871343,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002519230198340567,
      "loss": 4.9237,
      "mean_token_accuracy": 0.23706196248531342,
      "num_tokens": 106903420.0,
      "step": 47895
    },
    {
      "entropy": 5.751879024505615,
      "epoch": 5.3796259897793,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025188308321209504,
      "loss": 4.9168,
      "mean_token_accuracy": 0.23580024391412735,
      "num_tokens": 106914838.0,
      "step": 47900
    },
    {
      "entropy": 5.635126256942749,
      "epoch": 5.380187566687258,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025184314732618496,
      "loss": 4.8264,
      "mean_token_accuracy": 0.23823232501745223,
      "num_tokens": 106926519.0,
      "step": 47905
    },
    {
      "entropy": 5.718350267410278,
      "epoch": 5.3807491435952155,
      "grad_norm": 1.21875,
      "learning_rate": 0.000251803212177598,
      "loss": 4.9486,
      "mean_token_accuracy": 0.23330575227737427,
      "num_tokens": 106938303.0,
      "step": 47910
    },
    {
      "entropy": 5.724035739898682,
      "epoch": 5.381310720503173,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002517632777676058,
      "loss": 4.9398,
      "mean_token_accuracy": 0.24016852229833602,
      "num_tokens": 106950120.0,
      "step": 47915
    },
    {
      "entropy": 5.749152421951294,
      "epoch": 5.381872297411131,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00025172334409747976,
      "loss": 4.9559,
      "mean_token_accuracy": 0.2259376361966133,
      "num_tokens": 106960998.0,
      "step": 47920
    },
    {
      "entropy": 5.729631090164185,
      "epoch": 5.382433874319088,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002516834111684915,
      "loss": 4.9034,
      "mean_token_accuracy": 0.23481334000825882,
      "num_tokens": 106972640.0,
      "step": 47925
    },
    {
      "entropy": 5.726376676559449,
      "epoch": 5.382995451227045,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002516434789819126,
      "loss": 4.9182,
      "mean_token_accuracy": 0.23209938108921052,
      "num_tokens": 106985300.0,
      "step": 47930
    },
    {
      "entropy": 5.739586448669433,
      "epoch": 5.383557028135003,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002516035475390143,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22365637570619584,
      "num_tokens": 106996074.0,
      "step": 47935
    },
    {
      "entropy": 5.748926877975464,
      "epoch": 5.384118605042961,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00025156361684106814,
      "loss": 4.9356,
      "mean_token_accuracy": 0.23274981528520583,
      "num_tokens": 107006493.0,
      "step": 47940
    },
    {
      "entropy": 5.69588418006897,
      "epoch": 5.384680181950918,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002515236868893455,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2335769549012184,
      "num_tokens": 107018739.0,
      "step": 47945
    },
    {
      "entropy": 5.625814914703369,
      "epoch": 5.385241758858876,
      "grad_norm": 1.25,
      "learning_rate": 0.00025148375768511776,
      "loss": 4.793,
      "mean_token_accuracy": 0.2440428465604782,
      "num_tokens": 107028762.0,
      "step": 47950
    },
    {
      "entropy": 5.734732961654663,
      "epoch": 5.385803335766833,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002514438292296563,
      "loss": 4.9332,
      "mean_token_accuracy": 0.225775608420372,
      "num_tokens": 107039609.0,
      "step": 47955
    },
    {
      "entropy": 5.6326312065124515,
      "epoch": 5.386364912674791,
      "grad_norm": 1.25,
      "learning_rate": 0.00025140390152423247,
      "loss": 4.8305,
      "mean_token_accuracy": 0.2446603924036026,
      "num_tokens": 107051871.0,
      "step": 47960
    },
    {
      "entropy": 5.684198379516602,
      "epoch": 5.3869264895827484,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002513639745701176,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2347772851586342,
      "num_tokens": 107063560.0,
      "step": 47965
    },
    {
      "entropy": 5.752175951004029,
      "epoch": 5.387488066490706,
      "grad_norm": 1.3125,
      "learning_rate": 0.00025132404836858294,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2369634360074997,
      "num_tokens": 107074466.0,
      "step": 47970
    },
    {
      "entropy": 5.726258230209351,
      "epoch": 5.388049643398664,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002512841229208997,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2415146052837372,
      "num_tokens": 107085470.0,
      "step": 47975
    },
    {
      "entropy": 5.73289475440979,
      "epoch": 5.388611220306621,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002512441982283392,
      "loss": 4.9844,
      "mean_token_accuracy": 0.23666778951883316,
      "num_tokens": 107098179.0,
      "step": 47980
    },
    {
      "entropy": 5.6990828037261965,
      "epoch": 5.389172797214578,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002512042742921727,
      "loss": 4.9064,
      "mean_token_accuracy": 0.2377736195921898,
      "num_tokens": 107108941.0,
      "step": 47985
    },
    {
      "entropy": 5.734758090972901,
      "epoch": 5.389734374122536,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002511643511136712,
      "loss": 4.9747,
      "mean_token_accuracy": 0.22905554324388505,
      "num_tokens": 107120019.0,
      "step": 47990
    },
    {
      "entropy": 5.758144187927246,
      "epoch": 5.3902959510304935,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00025112442869410607,
      "loss": 4.9433,
      "mean_token_accuracy": 0.23153853118419648,
      "num_tokens": 107130254.0,
      "step": 47995
    },
    {
      "entropy": 5.655142021179199,
      "epoch": 5.390857527938451,
      "grad_norm": 1.203125,
      "learning_rate": 0.00025108450703474844,
      "loss": 4.8223,
      "mean_token_accuracy": 0.24310199320316314,
      "num_tokens": 107140625.0,
      "step": 48000
    },
    {
      "epoch": 5.390857527938451,
      "eval_entropy": 5.402058941573251,
      "eval_loss": 5.044894695281982,
      "eval_mean_token_accuracy": 0.23424660753644186,
      "eval_num_tokens": 107140625.0,
      "eval_runtime": 18.2445,
      "eval_samples_per_second": 1699.857,
      "eval_steps_per_second": 212.503,
      "step": 48000
    }
  ],
  "logging_steps": 5,
  "max_steps": 89030,
  "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": 2.489756188336128e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}